perceptions of risk factors and mitigating strategies for

University of Calgary

PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2016

Perceptions of Risk Factors and Mitigating Strategies

for Water Related Zoonotic Diseases on Small-Scale

Integrated Farms in Vietnam

Le, Quynh Ba

Le, Q. B. (2016). Perceptions of Risk Factors and Mitigating Strategies for Water Related Zoonotic

Diseases on Small-Scale Integrated Farms in Vietnam (Unpublished doctoral thesis). University of

Calgary, Calgary, AB. doi:10.11575/PRISM/27287

http://hdl.handle.net/11023/2862

doctoral thesis

University of Calgary graduate students retain copyright ownership and moral rights for their

thesis. You may use this material in any way that is permitted by the Copyright Act or through

licensing that has been assigned to the document. For uses that are not allowable under

copyright legislation or licensing, you are required to seek permission.

Downloaded from PRISM: https://prism.ucalgary.ca

UNIVERSITY OF CALGARY

Perceptions of Risk Factors and Mitigating Strategies for Water Related Zoonotic Diseases on

Small-Scale Integrated Farms in Vietnam

by

Quynh Ba Le

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

GRADUATE PROGRAM IN VETERINARY MEDICAL SCIENCES

CALGARY, ALBERTA

MARCH, 2016

© Quynh Ba Le 2016

ii

Abstract

Small-scale integrated (SSI) farming is a common agricultural model in Vietnam. However, SSI

farmers use contaminated water, have no access to hygienic latrines, and have very limited

awareness about water quality or environmental sanitation. There are potential interactions in the

context of this model that may result in transmission of water-related zoonotic diseases (WRZD).

The government of Vietnam plans to restructure the agriculture system by 2020. Study of the

model in the context of water and public health can contribute to shaping the future of Vietnam’s

agriculture system. I used a cross-sectional study to develop a socioeconomic profile of SSI

farmers, examine their perceptions of risk factors for transmitting WRZD, assess basic microbial

and related quality of water used on their farms, and explore factors that are associated with on-

farm water quality as well as SSI farmers’ engagement in strategies to reduce transmission of

WRZD. Data were collected from 600 SSI farms in two provinces in North and South Vietnam.

The typical profile of the participating SSI farmers was a 45 year old married individual with two

children, seven years of formal education, low income (c. $1200 p.a.), and nine years farming

experience. Most SSI farmers had basic awareness of avian influenza prevention, but very

limited awareness of WRZDs such as Escherichia coli (E. coli). Water on the majority of SSI

farms had unacceptable levels of E. coli and was significantly associated with SSI farmers’

characteristics and perceptions (e.g., years of farming, number of poultry on a farm, and

perceived self-efficacy in managing livestock). SSI farmers’ characteristics and perceptions (e.g.,

education, income, and perceived self-efficacy in managing livestock) influenced their

engagement in strategies to reduce transmission of WRZD. Future research and policies in

Vietnamese agriculture need to consider a transdisciplinary approach (e.g., EcoHealth) to

increase SSI farmers’ awareness of water public health and their engagement in mitigating

iii

strategies for WRZD transmission. Consideration for the health of farmers, animals, and the

environment should be interwoven with livestock production and be an equally important part of

a holistic integrated SSI farming model.

iv

Acknowledgements

There are many people - from rural communes in Vietnam to various research

organizations in both Vietnam and Canada – to whom I am grateful for their support throughout

my doctoral study program. Their support has helped me this far in my journey/adventure of

further education and career development moving beyond my human medicine and public health

background and into a fascinating, challenging, and holistic research area that involves not only

humans but also animals and their environment – i.e., EcoHealth/One Health/Global Health.

Ultimately, I am grateful for the tremendous guidance and mentorship from my primary

supervisor, Dr. David Hall, not only throughout my study in Canada and in the field study in

Vietnam but also my life in Canada. His experiences, patience, support, and faith in me has

empowered me to strive to become a transdisciplinary development researcher. I would not have

completed my study without my co-supervisor, Dr. Susan Cork, who provided advice,

leadership, and vision that helped me position my study and career. I am grateful for the advice

and support from my committee members Drs. Carl Ribble, Margaret Russell, and Sylvia

Checkley. Their expertise, knowledge, and extremely patient feedback have shaped my research

thinking and skills.

I am also grateful for the support from Dr. Jeff Davidson, who procured a seed grant that

enabled me to connect with local researchers and pilot my study tools. I highly appreciate the

help of staff members at UCVM, Dr. James Cross, Dr. John Matyas, Dr. Tara Christie, Lorraine

Toews, Katie Douglas, Ingrid Middleton, Kasia Judycki, Lisa Clough, Jess Nakaska, and my

fellow graduate students who helped me back on track. I would like to thank Chris Morley for

supporting my research and connecting to his network researcher in Vietnam.

v

Especially, I would like to extend my sincere thanks to the 600 farmers in Thai Binh and

An Giang who participated in my field data collection and enthusiastically shared their thoughts,

ideas, and concerns, which were the foundation for my research. A huge benefit for me was the

unique opportunity to work with and learn from the stakeholders and partners across disciplines

and sectors in Vietnam and Canada. The list of stakeholders, partners, and staff who helped me

during the field data collection is long and my apologies to those who I am not able to name in

the following: Drs. Vu Hoang Hoa, Tran Van On (Hanoi Water Resources University), Dr.

Nguyen Viet Hung, Dr. Pham Duc Phuc, Luu Quoc Tuan, Lam Thi Binh, Dang Van Chinh,

Nguyen Duy Tien, and Le Thanh Huyen (Hanoi School of Public Health), Drs. Nguyen Dinh

Minh, Nguyen Viet Khong, Nguyen Lan Anh, Pham Thi Ngoc (National Institute of Veterinary

Research), Dr. Phan Thi Van (Research Institute of Aqua-agriculture No 1), Dr. Nguyen Van

Duc and his team in Thai Binh, Drs. Huynh Van Nen and his team in An Giang, Dr. Pham

Quang Khai (NUSA Vietnam), and Kimberly Schiefer (Bluewater BioSciences Inc, Canada).

Their support and collaboration were extremely helpful for me to complete my data collection in

Vietnam. This dissertation was professionally copy edited by Kathleen McWilliams, with

permission from my supervisor and following the thesis editing regulations of the Faculty of

Graduate Studies, University of Calgary.

My work would not have been possible without enabling support from different funders

across Canada. I highly appreciate funding support for my research from IDRC Doctoral

Research Awards/International Development Research Centre (IDRC), Stars in Global

Health/Grand Challenges Canada (GCC), Seed Grant/Canada Excellence Research Chair

(CERC), and the Building Ecohealth Capacity in Asia (BECA) project that was implemented by

the University of Calgary and administered by Veterinarians Without Borders Canada. The

vi

UCVM Graduate Student Entrance Awards and the DEPH Studentship and Fellowship Funds

from the Department of Ecosystem and Public Health, Faculty of Veterinary Medicine were

crucial for me to start and maintain my study.

I thank my relatives and friends in Vietnam and Canada for their love, support,

understanding, and patience during my extended long commitment to PhD study. I am proud of

my grandparents, parents, brothers, and sisters for believing in me and supporting me

unconditionally. I am extremely lucky for the unconditional love and support from my wife and

my kids – Fish, Chip, and Bob – who have been always with me through my difficult times, kept

me optimistic and on my feet when I was about to fall apart.

Thank you to all of you without which this work would not have been possible!

vii

Dedication

“May be tomorrow I will settle down, until tomorrow I will just keep moving on!”

From my favorite TV series “Littlest Hobo”

viii

Table of Contents

Abstract ............................................................................................................................... ii Acknowledgements ............................................................................................................ iv

Dedication ......................................................................................................................... vii Table of Contents ............................................................................................................. viii List of Tables ................................................................................................................... xiv List of Figures and Illustrations .........................................................................................xx List of Symbols, Abbreviations and Nomenclature ......................................................... xxi

Glossary .......................................................................................................................... xxii

CHAPTER ONE: INTRODUCTION ................................................................................24

1.1. Overview of the Research ......................................................................................24 1.2. Background ............................................................................................................26

1.2.1. Small-Scale Integrated farming as an important agricultural model in

Vietnam ............................................................................................................26

1.2.2. Rural water and public health issues may challenge the SSI farming model.

28

1.2.3. SSI farmers’ perceptions about risk factors for transmitting WRZD may be

associated with their engagement in mitigating strategies to reduce transmitting

WRZD ..............................................................................................................29

1.3. Key Research Questions ........................................................................................30 1.4. Research Methods ..................................................................................................31

1.5. A Description of Chapters in the Dissertation .......................................................32

1.6. Connection Between Chapters ...............................................................................37

References ......................................................................................................................38

CHAPTER TWO: DEVELOPING A SOCIOECONOMIC PROFILE OF SMALL-SCALE

INTEGRATED (SSI) FARMS IN THE PROVINCES OF THAI BINH AND AN

GIANG IN VIETNAM .............................................................................................48 2.1. Introduction ............................................................................................................48

2.1.1. Overview of Small-Scale Integrated (SSI) farming in Vietnam .............48 2.1.2. Rationale for the research .......................................................................50

2.1.3. SSI farming and risk factors for transmission of HPAI ..........................52 2.1.4. Research questions and objectives ..........................................................53

2.2. Methods..................................................................................................................57

2.2.1. Research design.......................................................................................57

2.2.1.1. Methods and phases of the research ...................................................57 2.2.1.2. Research areas ....................................................................................58 2.1.2.3. Selection criteria ................................................................................59 2.2.1.4. Sampling design: ................................................................................61 2.2.1.5. Variables ............................................................................................62

2.2.2. Pilot study ...............................................................................................63 2.2.3. Data collection for the full research study ..............................................63 2.2.4. Analysis of the full field survey’s data ...................................................66

2.3. Results ....................................................................................................................68 2.3.1. Demographics .........................................................................................68

ix

2.3.1.1. Age, years of farming, years of attending school, and household size68 2.3.1.2. Gender and gender role ......................................................................69

2.3.2. On-farm production.................................................................................72 2.3.2.1. Fish production ..................................................................................72

2.3.2.2. Livestock production .........................................................................73 2.3.2.3. On-farm production other than fish and livestock production ...........73 2.3.2.4. Occupation other than farming among participating farmers ............73

2.3.3. Farm economics ......................................................................................75 2.3.3.1. Income of SSI farmers .......................................................................75

2.3.3.2. Contribution to total farm income from various kinds of agriculture

production ...............................................................................................76 2.4. Discussion ..............................................................................................................78

2.4.1. SSI farming practice and its profile in the provinces of Thai Binh and An

Giang ................................................................................................................78 2.4.2. SSI farming and EcoHealth.....................................................................79

2.5. Conclusions ......................................................................................................83 References ......................................................................................................................85

CHAPTER THREE: SMALL-SCALE INTEGRATED FARMERS’ PERCEPTIONS

ABOUT RISK FACTORS FOR TRANSMITTING WATER-RELATED ZOONOTIC

DISEASE (WRZD) IN THE PROVINCES OF THAI BINH AND AN GIANG IN

VIETNAM ..............................................................................................................100 3.1. Introduction ..........................................................................................................100

3.3.1. Rationale for studying farmers’ perceptions of risk factors for transmission

of WRZD .......................................................................................................101

3.1.2. Research question and objective ...........................................................102 3.2. Methods................................................................................................................103

3.3. Results ..................................................................................................................109 3.3.1. Perceptions of water quality in the villages ..........................................109 3.3.2. Perceptions of threat to health and/or wellbeing ...................................112

3.3.2.1. Perceived susceptibility to diseases .................................................113 3.3.2.2. Perceived severity of diseases ..........................................................114

3.3.3. Expectations ..........................................................................................117 3.3.3.1. Perceptions of barriers to taking mitigating actions ........................117

3.3.3.2. Perceptions of benefits of taking or not taking mitigating actions ..117 3.3.3.3. Perceptions about self-efficacy in performing mitigating actions ...121

3.3.4. Cues to actions ......................................................................................124 3.4. Discussion ............................................................................................................125 3.5. Conclusions ..........................................................................................................130 References ....................................................................................................................132

CHAPTER FOUR: WATER QUALITY AND PUBLIC HEALTH AMONG SMALL-

SCALE INTEGRATED (SSI) FARMS IN THE PROVINCES OF THAI BINH AND

AN GIANG IN VIETNAM ....................................................................................138 4.1. Introduction ..........................................................................................................138

4.1.1. Rationale ...............................................................................................138 4.1.2. Research questions and hypotheses ......................................................141

x

4.1.2.1. Research questions: ..........................................................................141 4.1.2.2. Hypotheses: ......................................................................................141

4.2. Methods................................................................................................................142 4.3. Results: .................................................................................................................147

4.3.1. Sources of water and frequencies of use in participating farms ...........147 4.3.2. Results of on-farm water quality assessed using the Vietnamese national

laboratory protocol .........................................................................................148 4.3.3. Quality of on-farm water assessed by community health/veterinary workers

(CHW/CVWs)................................................................................................149

4.3.4. Factors associated with the auality of on-farm water ...........................149 4.3.4.1. Description of variables used for modelling ....................................150 4.3.4.2. MLR Model - E. coli cfu in sources of water used for drinking (DV) vs. a

set of IVs ...............................................................................................150 4.3.4.3. MLR Model - E. coli cfu in water sources for domestic use (DV) vs. a set

of IVs.....................................................................................................157

4.3.4.4. SUR Model - E. coli cfu in sources of on-farm water used for domestic

purposes ................................................................................................158

4.4. Discussion ............................................................................................................164 4.4.1. Sources and frequency of use of on-farm water for drinking and domestic

purposes .........................................................................................................164

4.4.2. Low quality of on-farm water for both drinking and domestic purposes167 4.4.3. Associated factors of on-farm water quality .........................................171

4.4.4. Rural water and its regulatory framework on SSI farms in Vietnam....174 4.5. Conclusions ..........................................................................................................177

References ....................................................................................................................179

CHAPTER FIVE: FACTORS ASSOCIATED WITH SMALL-SCALE INTEGRATED (SSI)

FARMERS' ENGAGEMENT IN MITIGATING STRATEGIES TO REDUCE

TRANSMISSION OF WATER-RELATED ZOONOTIC DISEASES (WRZD) IN THE

PROVINCES OF THAI BINH AND AN GIANG IN VIETNAM ........................189

5.1. Introduction ..........................................................................................................189 5.1.1. Rationale ...............................................................................................189

5.1.2. Research questions and hypothesis .......................................................193 5.1.2.1. Research questions ...........................................................................193

5.1.2.2 Research hypothesis .........................................................................194 5.2. Methods................................................................................................................194

5.2.1. Methods used to identify SSI farmers’ engagement in mitigating strategies

to reduce the transmission of WRZD.............................................................194 5.2.2. Methods used to explore factors that are associated with SSI farmers’

engagement in mitigating strategies...............................................................196 5.3. Results ..................................................................................................................198

5.3.1. SSI farmers’ actions and engagement in mitigating strategies to reduce

WRZD transmission.......................................................................................199 5.3.1.1 Livestock management (LM) actions ..............................................199 5.3.1.2. Source water protection (SWP) actions ...........................................199 5.3.1.3. Water storage treatment and distribution (WSTD) actions ..............200 5.3.1.4. Point of use/household (POU) actions .............................................202

xi

5.3.1.5. SSI farmers’ engagement in mitigating strategies to reduce WRZD

transmission ..........................................................................................203 5.3.2. Factors associated with SSI farmers’ engagement in the mitigating strategies

to reduce transmission of WRZD ..................................................................204

5.3.2.1. Results of probit analyses of SSI farmers’ engagement in mitigating

strategies to reduce WRZD transmission ..............................................204 5.3.2.2. Marginal probability effects of the factors that are associated with SSI

farmers’ engagement in mitigating strategies to reduce WRZD transmission

205

5.4. Discussion ............................................................................................................212 5.4.1. SSI farmers’ choices of actions and their engagement in mitigating strategies

to reduce the transmission of WRZD.............................................................212

5.4.1.1. Wide differences in choices of specific actions to mitigate WRZD

transmission ..........................................................................................212 5.4.1.2. SSI farmers’ broad engagement in mitigating strategies to reduce WRZD

transmission ..........................................................................................214 5.4.2. Factors associated with SSI farmers’ engagement in mitigating strategies to

reduce transmission of WRZD .......................................................................214 5.4.2.1. Factors associated with SSI farmers’ engagement in LM as a mitigating

strategy to reduce WRZD transmission ................................................216

5.4.2.2. Factors associated with SSI farmers’ engagement in SWP as a mitigating


5.4.2.3. Factors associated with SSI farmers’ engagement in POU as a mitigating


5.4.2.4. Factors associated with SSI farmers’ engagement in WSTD as a

mitigating strategy to reduce WRZD transmission ...............................220

5.4.2.5. Common factors associated with SSI farmers’ engagement across four

mitigating strategies to reduce WRZD transmission ............................221 5.5. Conclusions ..........................................................................................................222

References ....................................................................................................................225

CHAPTER SIX: CONCLUSION ....................................................................................231

6.1. Overview ..............................................................................................................231 6.2. Summary of the Literature Review ......................................................................232

6.3. Research Questions, Hypotheses, and Methods ..................................................233 6.4. Contribution to Knowledge..................................................................................235

6.4.1. An understanding of SSI farming model in the context of rural water and

public health ...................................................................................................235 6.4.1.1. A socioeconomic profile of the SSI farming practice in Thai Binh and An

Giang .....................................................................................................236 6.4.1.2. SSI farmers’ limited perceptions of risk factors for transmission of

WRZD ...................................................................................................238 6.4.1.3. SSI farmers’ engagement in mitigating strategies to reduce transmission

of WRZD...............................................................................................239 6.4.1.4. Low quality of water used on SSI farms for both drinking and domestic

purposes ................................................................................................239

xii

6.4.2. An understanding of factors associated with on-farm water quality and SSI

farmers’ engagement in mitigating strategies to reduce the transmission of WRZD

240 6.4.2.1. Factors that are associated with on-farm water quality ...................240

6.4.2.2. Factors that are associated with SSI farmers’ engagement in mitigating

strategies to reduce the transmission of WRZD ....................................242 6.5. Limitations ...........................................................................................................242 6.6. Overall Recommendations ...................................................................................244 6.7. Conclusions/Final Thoughts ................................................................................247

References ....................................................................................................................248

APPENDIX A: THEORETICAL MODEL .....................................................................251

APPENDIX B: QUESTIONNAIRES USED DURING THE FIELD STUDY ..............252 B.1. Questionnaires in Vietnamsese: ...........................................................................252 B.2. Questionnaires in English: ...................................................................................269

APPENDIX C: CONSENT SCRIPTS .............................................................................285

C.1. Consent Scripts in Vietnamese .............................................................................285 C.2. Consent Scripts in English ...................................................................................290

APPENDIX D: ETHICS APPROVALS .........................................................................294 D.1. University of Calgary Ethics ID: REB13-0470 ...................................................294 D.2. Hanoi School of Public Health Ethics Approval: No. 151/2013/YTCC_HD3 ....294

D.3. Hanoi School of Public Health Ethics Approval: No. 008/2014/YTCC_HD3 ....294

APPENDIX E: CHAPTER TWO’S APPENDIX............................................................295

APPENDIX F: CHAPTER FOUR’S APPENDICES ......................................................300 F.1. Brief Ecology of Escherichia coli and WHO’s Recommendations for Assessing Rural

Water Quality .......................................................................................................300 F.2. Protocols Used for Assessing On-Farm Water Quality ........................................303

F.2.1. Protocol for Laboratories to Assess Basic Microbial and Related Water Quality

........................................................................................................................303

F.2.2. Protocol for Community Health/Veterinary Workers to Quantify E. coli ...305 F.3. Detailed Model Building Strategies for MLR and SUR .......................................307

F.3.1. Modeling building strategies ........................................................................307

F.3.2. Multiple linear regression (MLR) model .....................................................307 F.3.3. Seemingly Unrelated Regression (SUR) Model ..........................................309 F.3.4. Detailed Results of Sources of Water Used and Corresponding Frequencies of

Use among SSI Farmers in the Provinces of Thai Binh and An Giang .........315

F.4. Detailed Results of Basic Microbial and Related Indicators of On-Farm Water .316 F.4.1. Quality of On-Farm Water Used for Drinking .............................................316 F.4.2. Quality of On-Farm Water Used for Domestic Purposes ............................317 F.4.3. Detailed Results of Drinking Water Quality Assessed by CHW/CVW using

ColiplateTM .....................................................................................................318 F.5. Detailed Results of MLR Models for E. coli in Sources of Water On Farms ......319

xiii

F.6. Detailed Results of SUR Model for E. coli in Sources of Water On Farms ........331

APPENDIX G: CHAPTER FIVE’S APPENDICES .......................................................334 G.1. Summary Statistics of Variables Used in the Probit Models ...............................334 G.2. Summary Statistics of SSI Farmers’ Perceptions Related Variables Used in the Probit

Models..................................................................................................................335 Table G.2. Summary Statistics of SSI Farmers’ Perceptions Related Variables Used in Probit

Models to Explore Factors Associated with SSI Farmers’ Engagement in Mitigating Strategies to

Reduce WRZD Transmission ..........................................................................................335 References ....................................................................................................................336

xiv

List of Tables

Table 2.1. Risk Factors for Transmission of Highly Pathogenic Avian Influenza (HPAI) and

Their Possible Connection with Small-Scale Integrated (SSI) Farming .............................. 55

Table 2.2. Demographics of Small-Scale Integrated (SSI) Farm Study Participants in the

Provinces of Thai Binh and An Giang, Vietnam .................................................................. 68

Table 2.3. Types of School Attended by Small-Scale Integrated (SSI) Farmers in the

Provinces of Thai Binh and An Giang, Vietnam .................................................................. 69

Table 2.4. Gender of the Participating Farmers in the Provinces of Thai Binh and An Giang

in Vietnam and Their Relationship with Other Household Members .................................. 71

Table 2.6. Mean Quantity of Young Fish Produced Per Year on Small-Scale Integrated (SSI)

Farms in the Provinces of Thai Binh and An Giang, Vietnam, Stratified for Education

and Income ............................................................................................................................ 72

Table 2.7. Annual Livestock Production on Small-Scale Integrated Farms (SSI) in the Thai

Binh and An Giang Provinces in Vietnam ............................................................................ 73

Table 2.8. On-farm Production Other Than Fish and Livestock on Small-Scale Integrated

(SSI) Farms in the Thai Binh and An Giang Provinces in Vietnam ..................................... 74

Table 2.9. Occupations Other Than Farming Among Small-Scale Integrated (SSI) Farmers in

the Provinces of Thai Binh and An Giang, Vietnam ............................................................ 75

Table 2.10. Income of Small-Scale Integrated (SSI) Farmers in the Provinces of Thai Binh

and An Giang, Vietnam ........................................................................................................ 76

Table 2.11a. Farmers’ Perceptions of the Importance of Contribution to Total Farm Income

from Various Types of Agricultural Production on Small-Scale Integrated (SSI) Farms

in the Province of Thai Binh, Vietnam ................................................................................. 77

Table 2.11b. Farmers’ Perceptions of the Importance of Contribution to Total Farm Income

from Various Types of Agricultural Production on Small-Scale Integrated (SSI) Farms

in the Province of An Giang, Vietnam .................................................................................. 77

Table 3.1. Perceived General Condition of the Water Environment in the Villages of Small-

Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam .... 110

Table 3.2. Perceived Harm to Health When Using Different Sources of Water among Small-

Scale (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam ..................... 111

Table 3.3. Factors Influencing Small-Scale Integrated (SSI) Farmers’ Thoughts about Water

Quality in the Provinces of Thai Binh and An Giang, Vietnam ......................................... 112

xv

Table 3.4. Perceived Susceptibility to Highly Pathogenic Avian Influenza (HPAI), Diarrhoea,

Coliform Bacteria, and Parasites from Untreated Water Sources among Small-Scale

Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam ............. 115

Table 3.5. Perceived Severity of Diseases from Highly Pathogenic Avian Influenza (HPAI),

Diarrhoea, Coliform Bacteria, and Parasites among Small-Scale Integrated (SSI)

Farmers in the Provinces of Thai Binh and An Giang, Vietnam ........................................ 116

Table 3.6. Perceived Barriers to Taking Actions to Mitigate Transmission of WRZD among

Small-Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang,

Vietnam ............................................................................................................................... 118

Table 3.7. Perceived Benefits that Encourage Small-Scale (SSI) Farmers in the Provinces of

Thai Binh and An Giang, Vietnam, to Either Take or Not Take Actions to Mitigate

Transmission of WRZD ...................................................................................................... 119

Table 3.8. Perceptions of Water and Wastewater Value for Livestock Production among


Vietnam ............................................................................................................................... 120

Table 3.9. Levels of Importance for Water Use Issues in Reducing Transmission of WRZD

among Small-Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An

Giang, Vietnam ................................................................................................................... 122

Table 3.10. Perceived Self-efficacy to Perform Mitigating Actions among Small-Scale

Integrated Farmers in Thai Binh and An Giang, Vietnam .................................................. 123

Table 4.1. Mean Frequency of Use for On-Farm Water Used for Drinking and Domestic

Purposes in Both Provinces of Thai Binh and An Giang, Vietnam .................................... 151

Table 4.2. Comparison of Mean pH, Turbidity, and E. coli cfu of on-Farm Water Used for

Drinking, Assessed Using Samples Tested by National Laboratories, in the Provinces of

Thai Binh and An Giang, Vietnam ..................................................................................... 152

Table 4.3. Comparison of Mean pH, Turbidity, E. coli cfu of On-Farm Water Used for

Domestic Purposes, Assessed Using Samples Tested by National Laboratories, in the

Provinces of Thai Binh and An Giang, Vietnam ................................................................ 153

Table 4.4. Prevalence of E. coli in On-Farm Water Tests (using ColiplateTM) Used for

Drinking by Source in Thai Binh and An Giang, Vietnam ................................................. 154

Table 4.5. Summary Descriptive Statistics for the Dependent E. coli Variables Used in the

Regression Models .............................................................................................................. 154

Table 4.6a. Summary Statistics for the Independent Variables (IVs) Considered for Use in

Regression Models .............................................................................................................. 155

xvi

Table 4.6b. Summary Statistics for the Independent Variables (IVs) Considered for Use in

Regression Models (cont’d) ................................................................................................ 156

Table 4.7a. Summary of the Best MLR Model Fittings for E. coli cfu in Each of the Four

Sources of Drinking Water in the Provinces of Thai Binh and An Giang, Vietnam .......... 159

Table 4.7b. Summary of the Best MLR Model Fittings for E. coli cfu in Each of the Four

Sources of Drinking Water in the Provinces of Thai Binh and An Giang, Vietnam

(cont’d) ................................................................................................................................ 160

Table 4.8. Summary of the MLR Best Model Fittings for E. coli in Pond Water and River

Water for Domestic Purposes in the Provinces of Thai Binh and An Giang, Vietnam ...... 161

Table 4.9a. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints ......................................................................... 162

Table 4.9b. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints (cont’d) ........................................................... 162

Table 4.9c. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints (cont’d) ........................................................... 163

Table 5.1. Small-Scale Integrated (SSI) Farmers’ Mitigating Strategies and Their

Corresponding Actions in Reducing Risk Factors for Water-Related Zoonotic Disease

(WRZD) Transmission (*) .................................................................................................. 196

Table 5.2. Summary Statistics of Variables Indicating SSI Farmers’ Engagement in

Mitigating Strategies for WRZD Transmission .................................................................. 204

Table 5.3a. Maximum Likelihood Estimates of Factors that are Associated with Small-Scale

Integrated (SSI) Farmers’ Engagement in Mitigating Strategies of Water Related

Zoonotic Diseases (WRZD) Transmission Using Probit Analyses .................................... 206

Table 5.3b. Maximum Likelihood Estimates of Factors that are Associated with Small-Scale

Integrated (SSI) Farmers’ Engagement in Mitigating Strategies of Water Related

Zoonotic Diseases (WRZD) Transmission Using Probit Analyses (cont’d.) ..................... 207

Table 5.3c. Maximum Likelihood Estimates of Factors that are Associated with SSI Farmers’

Engagement in Mitigating Strategies of Water Related Zoonotic Diseases (WRZD)

Transmission using Probit Analyses (cont’d.) .................................................................... 208

Table 5.4. Predicted Frequency of Mitigating Strategies ........................................................... 209

Table 5.5a. Marginal Effects (at Means) of Factors Associated with SSI Farmers’

Engagement in Mitigating Strategies for WRZD Transmission Using Probit Analyses .... 210

xvii

Table 5.5b. Marginal Effects (at Means) of Factors that are Associated with Small-Scale

Integrated (SSI) Farmers’ Engagement in Mitigating Strategies of WRZD Transmission

Using Probit Analyses ......................................................................................................... 211

Table E.1. Annual Fish Production on Small-Scale (SSI) Farms in the Provinces of Thai Binh

and An Giang in Vietnam ................................................................................................... 296

Table E.2. Quantity and Kilograms of Fish Produced per Year on Small-Scale Integrated

(SSI) Farms in the Provinces of Thai Binh and An Giang, Vietnam (without outliers) ..... 297

Table E.3. Quantity of Young Fish Produced per Year on Small-Scale Integrated (SSI) Farms

in the Provinces of Thai Binh and An Giang in Vietnam Stratified over Education and

Income (with outliers) ......................................................................................................... 299

Table F.1a. Pairwise Correlation between Independent Variables used for Regression Models 311

Table F.1b. Pairwise Correlation between Independent Variables used for Regression Model

(cont’d) ................................................................................................................................ 312

Table F.1c. Pairwise Correlation between Independent Variables used for Regression Models

(cont’d) ................................................................................................................................ 313

Table F.1d. Pairwise Correlation between Independent Variables used for Regression Models

(cont’d) ................................................................................................................................ 314

Table F.2. Mean Frequency of Use of On-Farm Water for Drinking in the Provinces of Thai

Binh and An Giang, Vietnam .............................................................................................. 315

Table F.3. Mean Frequency of Use of On-Farm Water for Domestic Purposes in the


Table F.4. Mean pH, Turbidity, and E. coli cfu of On-Farm Water Used for Drinking in the


Table F.5. Mean pH, Turbidity, and E. coli cfu of On-Farm Water Used for Domestic

Purposes in the Provinces of Thai Binh and An Giang, Vietnam ....................................... 318

Table F.6. Mean MPN a of E. coli cfu Using ColiplateTM Tests in On-Farm Water for

Drinking by Sources and by Number of Incubation Day in the Provinces of Thai Binh

and An Giang, Vietnam ...................................................................................................... 318

Table F.7a. Summary of the Best MLR Model Fittings for E. coli cfu in Stored Rainwater for

Drinking by Selection Procedures in the Provinces of Thai Binh and An Giang, Vietnam 319

Table F.7b. Summary of the Best MLR Model Fittings for E. coli cfu in Stored Rainwater for

Drinking by Selection Procedures in the Provinces of Thai Binh and An Giang, Vietnam

(cont’d) ................................................................................................................................ 320

xviii

Table F.8a. Summary of the Best MLR Model Fittings for E. coli cfu in Pipe Water for

Drinking by Selection Procedures in the Provinces of Thai Binh and An Giang

Provinces, Vietnam ............................................................................................................. 321

Table F.8b. Summary of the Best MLR Model Fittings for E. coli cfu in Pipe Water for


(cont’d) ................................................................................................................................ 322

Table F.9. Summary of the Best MLR Model Fittings for E. coli cfu in Well Water for


Table F.10. Summary of the Best MLR Model Fittings for E. coli cfu in Well Water for


Table F.11a. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for

Drinking by Selection Procedures in the Provinces of Thai Binh and An Giang

Provinces, Vietnam ............................................................................................................. 325

Table F.11b. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for


(cont’d) ................................................................................................................................ 326

Table F.12a. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for

Domestic Purposes by Selection Procedures in the Provinces of Thai Binh and An

Giang, Vietnam ................................................................................................................... 327

Table F.12b. Summary of the Best MLR Model Fittings for E. coli cfu in River Water Used

for Domestic Purposes by Selection Procedures in the Provinces of Thai Binh and An

Giang, Vietnam (cont’d) ..................................................................................................... 328

Table F.13a. Summary of the Best MLR Model Fittings for E. coli in Pond Water for


Giang, Vietnam ................................................................................................................... 329

Table F.13b. Summary of the Best MLR Model Fittings for E. coli in Pond Water for


Giang, Vietnam (cont’d) ..................................................................................................... 330

Table F.14. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking .............................................................................................................................. 331


Drinking (cont'd) ................................................................................................................. 331

Table F.16. SUR Model Estimation of the Two Equations of E. coli cfu in On-Farm Water .... 332

xix

Table G.1. Summary Statistics of Demographic Variables Used in Probit Models of Factors

Associated with Small-Scale Integrated (SSI) Farmers’ Engagement in Mitigating

Strategies to Reduce WRZD Transmission ........................................................................ 334

xx

List of Figures and Illustrations

Figure 1.1. Water-Related Zoonotic Disease Interactions in the Water Environment ................. 25

Figure 1.2. Dissertation’s Outline ................................................................................................. 33

Figure 2.1. Chapter Two’s Outline ............................................................................................... 56

Figure 2.2. Research Areas – The Provinces of Thai Binh and An Giang in Vietnam ................ 60

Figure 3.1. Chapter Three’s Outline ........................................................................................... 104

Figure 3.2. Triggers/Cues to Taking Mitigating Actions against Transmission of Water

Related Zoonotic Diseases (WRZD)................................................................................... 124

Figure 4.1. Chapter Four’s Outline ............................................................................................. 143

Figure 5.1. Outline of Chapter Five ............................................................................................ 191

Figure 5.2. Livestock Management Actions ............................................................................... 201

Figure 5.3. Source Water Protection Actions ............................................................................. 201

Figure 5.4. Water Storage, Treatment, and Distribution Actions ............................................... 202

Figure 5.5. Point of Use/Household Actions .............................................................................. 203

Figure F.1. The Coliform Group ................................................................................................. 302

xxi

List of Symbols, Abbreviations and Nomenclature

Abbreviations Definition

AI Avian Influenza

AIV Avian Influenza Viruses

CHWs Community Health Workers

CVWs Community Veterinary Workers

CHWs/CVWs Community Health Workers and Community

Veterinary Workers

DVs Dependent Variables

E. coli Escherichia Coli

EPA United States Environmental Protection Agency

HPAI Highly Pathogenic Avian Influenza

IVs Independent Variables

MDG Millennium Development Goals

MoC Ministry of Construction

MoH Ministry of Health

MARD Ministry of Agriculture and Rural Development

MF Membrane Filtration

MPN Most Probable Numbers

MLR Multiple Linear Regression

NTU Nephelometric Turbidity Units

ISO International Organization for Standardization

OECD Organizations for Economic Cooperation and

Development

pH A numeric scale used to specify the acidity or

alkalinity of on-farm water

SSI Small-Scale Integrated

SUR Seemingly Unrelated Regression

TSA Trypone Soy Agar

UN United Nations

UNICEF United Nations Children’s Funds

WB World Bank

WHO World Health Organization

WRZD Water Related Zoonotic Diseases

xxii

Glossary

Association Association is a statistical dependence between 2 variables, and

indicates the degree to which an outcome is higher or lower in

those with or without exposure. An association can also be

referred to as a link, relationship or correlation (Oxford

Handbook of Epidemiology for Clinicians, 2012)

Cues/Triggers to Action Strategies to activate “readiness” condition (Health Belief

Model) or stimuli needed to trigger the decision-making

process to accept a recommended health action:

http://www.utwente.nl/cw/theorieenoverzicht/Theory%20Clust

ers/Health%20Communication/Health_Belief_Model/

Domestic water Domestic water is defined as “water used for domestic use but

not for direct drinking or processing food” (QCVN 02:

2009/BYT - National Technical Regulation on the Domestic

Water Quality, Department of Preventive Medicine and

Environment, Ministry of Health – Vietnam)

Drinking water Drinking water is defined as “water used for direct drinking or

processing food” (QCVN 01: 2009/BYT - National Technical

Regulation on the Drinking Water Quality, Department of

Preventive Medicine and Environment, Ministry of Health –

Vietnam)

Mitigation A mitigating strategy of water related zoonotic disease

(WRZD) transmission is a plan (e.g., a plan for managing

livestock) that includes one or more activities a small-scale

integrated (SSI) farmer could choose to reduce the transmission

of WRZD. An example of a mitigating strategy is good

livestock management which may include actions or tactics

(e.g., vaccinate domestic animals and/or treat animal waste at

source) that SSI farmers could choose to reduce WRZD

transmission (adapted from World Health Organization)

Perceived susceptibility One’s opinion of chances of getting a condition (Health Belief

Model):



Perceived severity One’s opinion on how serious a condition and its consequences

are condition (Health Belief Model):



Perceived benefits One’s belief in the efficacy of the advised action to reduce risk

or seriousness of an impact condition (Health Belief Model):



Perceived barriers One’s opinion of the tangible and psychological costs of the

advised action condition (Health Belief Model):

http://www.utwente.nl/cw/theorieenoverzicht/Theory%20Clusters/Health%20Communication/Health_Belief_Model/








xxiii



Risk factors Something that increases a person’ chance of getting a disease.

This definition assumes that there is a causal relationship rather

than simply a statistical association. Risk factors are often

described as modifiable (e.g., environmental exposures and

behaviours) or fixed (e.g. age or ethnicity) (Ward, Toledano,

Shadick, Davies, and Elliott, 2012)

Self-efficacy Confidence in one’s ability to take action condition (Health

Belief Model):



Small-scale integrated

model/farming

Small-scale integrated (SSI) farming is defined as farming

practices in which farmers combine some types of livestock and

either directly or indirectly depend on agriculture.

VAC model A traditional small-scale integrated farming model in which:

V= Vuon (i.e., Vegetable garden); A= Ao (i.e., Fish pond), C=

Chuong (i.e., Livestock housing)

Water related zoonotic

diseases (WRZD)

Diseases in humans that come from animals and related to

water such as Highly Pathogenic Avian Influenza, some

parasites, or diseases caused by a range of bacteria (e.g.,

pathogenic strains of Escherichia coli). Criteria for determining

WRZD include: 1) the pathogen must spend part of its life

cycle within one or more animal species; 2) it is probable or

conceivable that some life stage of the pathogen will enter

water; and 3) transmission of the pathogen from animals to

humans must be through a water related route (World Health

Organization)





of 340

Chapter One: Introduction

1.1.Overview of the Research

In this dissertation, I present my research developed from a preliminary study in the

provinces1 of Thai Binh (in the Red River Delta) and An Giang (in the Mekong River Delta) in

Vietnam where I completed a field evaluation for controlling and preventing Highly Pathogenic

Avian Influenza (HPAI). In the preliminary study, I observed small-scale integrated (SSI)2

farmers interacting among themselves, their livestock, and the water environment on farms, and I

witnessed how these interactions could potentially transmit water-related zoonotic diseases

(WRZD). In my research, I used the World Health Organization’s (WHO) definition of WRZD

as diseases in humans that come from animals and are related to water such as HPAI viruses,

some parasites, or diseases caused by a range of bacteria (e.g., pathogenic strains of Escherichia

coli such as enterohaemorrhagic E. coli) (WHO, 2004). The criteria for determining WRZD

include: 1) the pathogen must spend part of its life cycle within one or more animal species; 2) it

is probable or conceivable that some life stage of the pathogen will enter water; and 3)

transmission of the pathogen from animals to humans must be through a water-related route

(Moe, 2004).

In Figure 1.1 below, I illustrate potential interactions in the context of SSI farms that may

result in transmission of WRZD.

1 Vietnam is divided into 64 provinces; each province is sub-divided into districts, then communes, and villages. 2 Small-scale integrated (SSI) farming is defined as farming practices in which farmers combine some forms of

livestock and either directly or indirectly depend on agriculture (Hall, Thao, Minh, and Lien, 2006; Irini and

Rapsomanikis, 2005; Pica-Ciarmarra, Tasciotti, L., Otte, and Zezza, 2011).

of 340

Figure 1.1. Water-Related Zoonotic Disease Interactions in the Water Environment

The conceptual model for my research was based on the observation that farmers make

decisions, including management choices related to animal and human health, that are influenced

by their impressions about the water quality on farms. The theoretical foundation for my research

was adapted from the Health Belief Model and the Theory of Planned Behaviour (Becker, 2012;

Janz and Becker, 1984; Rosenstock, Strecher, and Becker, 1988). Both the Health Belief Model

and the Theory of Planned Behaviour explain how individuals perceive a personal threat to

health and/or wellbeing, together with their belief in the benefits and barriers of an action, will

likely predict their health-related behaviours (see Appendix A). Based on the theoretical

foundation, the overall aim of my research was the following: to understand the association

between SSI farmers’ perceptions about the risk factors for transmitting WRZD and their

engagement in mitigating strategies to reduce transmitting WRZD in the provinces of Thai Binh

and An Giang in Vietnam. In my research, I defined a mitigating strategy to reduce transmitting

Adapted from WHO, 2004

On-farm water environment

Microbial pathogens

FarmersLivestock

Water related

zoonotic diseases

of 340

WRZD as a plan (e.g., a plan for managing livestock) that includes one or more activities an SSI

farmer could choose to reduce the transmission of WRZD. An example of a mitigating strategy is

good livestock management, which may include actions, or tactics (e.g., treat animal waste at

source) that SSI farmers could choose to reduce transmitting WRZD (Carr and Bartram, 2004;

WHO, 2004). Moreover, since the Vietnamese Ministry of Agriculture and Rural Development

is proposing to restructure the agriculture system in Vietnam by 2020 (MARD, 2012), studying

the SSI farming model in the context of rural water and public health in Vietnam can contribute

to shaping its agriculture system in the near future.

1.2. Background

1.2.1. Small-Scale Integrated farming as an important agricultural model in Vietnam

Integrated aquaculture-agriculture farming is a common agricultural model in rural areas

in Vietnam (Dang, Milstein, Verdegem, and Verreth, 2006; Vu, Tran, and Dang, 2007). This

model is often referred to by various names such as smallholder, the VAC model3, small-scale

producers, and/or small-scale integrated farms (Bui, Hien, and Dang, 2010). In this model,

farmers combine some types of livestock and directly or indirectly depend on agriculture (Hall et

al., 2006; Irini and Rapsomanikis, 2005; Pica-Ciarmarra et al., 2011). In my research, farming

practices with this combination are referred to as small-scale integrated (SSI) farming and rural

farmers who practice SSI farming are called SSI farmers. The following paragraph provides a

brief review of SSI farmers and SSI farming in Vietnam.

It is reported that 10 million SSI farmers in Vietnam live without adequate sanitation and

9.3 million of them lack safe drinking water (Rechenburg and Herbst, 2006; WHO, 2006).

3 The VAC model: V= Vuon (i.e., Vegetable garden), A= Ao (i.e. Fish pond), C= Chuong (i.e., Livestock housing)

of 340

However, SSI farmers benefit the least from national social economic policies and national

economic growth (VPMO, 2010). More than 80% of SSI farmers in Vietnam own some forms of

livestock (Hoang, 2011; Irini and Rapsomanikis, 2005; Pica-Ciarmarra et al., 2011; WB, 2002;

WB, 2006; WB, 2006), with 87% of them keeping primarily poultry (Irini and Rapsomanikis,

2005).

More than 50% of SSI farmers in Vietnam depend on livestock for food and income

(Hall, Benigno, and Wantanee, 2006; Irini and Rapsomanikis, 2005) and most of their cash

revenue comes from selling live animals (e.g., at home and/or in live poultry markets rather than

selling livestock products) (Irini and Rapsomanikis, 2005). Moreover, although SSI farmers have

been applying disease control methods such as vaccination and biosecurity (Morley, 2010), the

most common way they deal with infected animals is to salvage them (i.e., sell or eat the sick

animals). However, limited support and access to experienced animal health services further

promotes the spread of disease by magnifying the level of panic (Gleeson and Dung, 2010).

Backyard livestock rearing also plays an integral role in recycling nutrients in farming

systems and in generating valuable cash income, however, its development has been hampered

by infectious diseases such as classical swine fever and reproductive failure for sows

(Kamakawa, Thu, and Yamada, 2006). Scavenging ducks are also common and it is believed

they play an important role in maintaining and transmitting the H5N1 (HPAI) virus throughout

the food chain, yet published information is lacking on H5N1 infection among scavenging ducks

(Henning et al., 2011; Morley, 2010). Chapter Two has more detailed information on SSI

farming.

of 340

1.2.2. Rural water and public health issues may challenge the SSI farming model.

Water is the vehicle for transmitting about 80% of all water related-diseases in people

worldwide (Ford and Colwell, 1996). Traditionally, Vietnamese farmers have considered water

as the most important factor for sustainable agriculture (Bach, 2007). However, in the SSI

farming model, a significant amount of untreated excreta from animals and humans is used as

fertilizer, food for fish, irrigation, or is dumped, which contaminates water sources with excreted

pathogens such as E. coli (Dalsgaard, 2001; Dao, 2006; Pham et al., 2011; Porphyre and Nguyen,

2007; Rechenburg and Herbst, 2006; Vu et al., 2007). Thus, although over 70% of SSI farmers

use unsanitary water and have no latrines, their awareness of the importance of a clean water

supply and environmental sanitation is limited (MoC and MARD, 2000).

The transmission of WRZD is considered a key issue in the context of an inadequate

clean water supply and sanitation in the rural areas of Vietnam (Hanington, Hall, and Anyan,

2016; Thang, Tuan, and Huton, 2008). For example, some of the reported important risk factors

for transmission of HPAI viruses were the proximity and increased accessibility of poultry flocks

to water sources, the increased density of ponds and streams, and temporal flooding (Gilbert et

al., 2006; Gilbert et al., 2008; Henning, Pfeiffer, and Vu, 2009; Nguyen, Vuong, Pham, and Vu,

2011; Pfeiffer, Minh, Martin, Epprecht, and Otte, 2007). Water bodies are thought to be

indicators for transmission of HPAI viruses and could provide a temporary reservoir for them at

the village and farm level (Morris and Jackson, 2005). This is important because AI viruses can

persist for extended periods in water, sewage, excreta and animal wastes depending on different

factors such as water temperature (Boere, Galbraith, and Stroud, 2006; Morris and Jackson,

2005; Webster, Peiris, Chen, and Guan, 2006; WHO, 2007). More specifically, the HPAI

(H5N1) virus can persist in water for over a 60-day period, and constitutes a potential source of

of 340

waterborne AI transmission for wild birds and domestic ducks (Brown, Swayne, Cooper, Burns,

and Stallknecht, 2007; Domanska, Minta, Smietanka, Sylvie, and Berg, 2010; Nazir et al., 2010;

Roche et al., 2009). The increasing interaction between domestic and wild birds in a common

water source (Boere et al., 2006; Morris and Jackson, 2005; Zheng et al., 2010) also creates

favourable conditions for co-infection, re-assortment, and evolution of the viruses among host

species (Stallknecht, Shane, Kearney, and Zwank, 1990; Ward, Maftei, Apostu, and Suru, 2009;

Webster, Peiris, Chen, and Guan, 2006). Increasing livestock populations additionally pollute

both soil and surface waters (Bui, 2010; Trinh, 2001), with concerns for the presence of a high

concentration of pathogenic microorganisms in livestock wastes (Vu, 2001; Yajima and

Kurokura, 2008). The direct use of this contaminated water can result in an actual risk to public

health and appropriate risk mitigation measures such as source water protection and/or livestock

management are crucial (Guan and Holley, 2003; Toews, 2009; Trang, Bui, Molbak, Phung, and

Dalsgaard, 2007; Yajima and Kurokura, 2008).

1.2.3. SSI farmers’ perceptions about risk factors for transmitting WRZD may be associated

with their engagement in mitigating strategies to reduce transmitting WRZD

I have not found other studies that explore SSI farmers’ perceptions of risk factors or

their mitigating strategies for reducing WRZD transmission in Vietnam. Some studies have

shown generally that SSI farmers are not aware of, or do not rank highly, the health risks

associated with their use of water contaminated with waste from humans and from livestock

(Keraita et al., 2010). Rather, they view these as unavoidable risk factors related to livestock

production (Knudsen et al., 2008). Moreover, for farmers, health risks from using contaminated

water are not as serious as water that remains on the skin and causes skin problems; however,

health risks from non-composted smelly feces are serious because they enter the body through

of 340

polluted air (Knudsen et al., 2008). Although farmers know proper hygiene, such hygiene

measures are ineffectively practiced (e.g., hand washing without soap). Most likely, this is due to

misconceiving risk factors and/or lacking knowledge about cause-effect relationships, habits,

and/or not ranking the risks high enough while using water (Herbst et al., 2009; Keraita et al.,

2010; Rechenburg and Herbst, 2006).

In some cases, farmers are aware of health risk factors, however, they assess the value of

potential mitigation measures by the overall economic impact on work efficiency rather than by

taking into account the potential health benefits to be gained (Keraita et al., 2010). A number of

studies have shown that some farmers have a basic awareness of health risks relating to

contaminated water use (Herbst et al., 2009; Keraita, Jim'nez, and Drechsel, 2008; Owusu,

Bakang, Abaidoo, and Kinane, 2012; Owusu, Bakang, Abaidoo, and Kinane, 2011;

Weldesilassie, Boelee, Drechel, and Dabbert, 2011). However, these studies only targeted small

samples of peri-urban farmers with reference to crops instead of livestock among rural SSI

farmers. Thus, it is important to understand SSI farmers’ perceptions about the risk factors for

transmitting WRZD, their mitigating strategies to reduce this threat, and the factors associated

with their strategies. This understanding could be useful for the development of mutually

acceptable risk-management strategies and/or interventions (Keraita et al., 2010), especially in a

Vietnamese context.

1.3. Key Research Questions

My research thus addresses the following research questions:

1. What is the socioeconomic profile of SSI farmers in the provinces of Thai Binh and An

Giang in Vietnam?

of 340

2. Is the socioeconomic profile of SSI farmers in the provinces of Thai Binh and An Giang

consistent with an EcoHealth approach in the context of SSI farming?

3. What are the perceptions of SSI farmers concerning the risk factors for transmitting

WRZD in the provinces of Thai Binh and An Giang?

4. What sources of water do SSI farmers use for drinking and domestic purposes4 and what

are the corresponding basic microbial and related indicators5 of water quality for those

sources in the provinces of Thai Binh and An Giang?

5. How does the quality of on-farm water that SSI farmers in the provinces of Thai Binh and

An Giang used for drinking and domestic purposes relate to their perceptions about both

water quality and risk factors for transmitting WRZD?

6. Do SSI farmers in the provinces of Thai Binh and An Giang engage in mitigating

strategies to reduce transmitting WRZD?

7. What factors are associated with SSI farmers’ engagement in mitigating strategies to

reduce the transmission of WRZD in the provinces of Thai Binh and An Giang?

1.4. Research Methods

In my research, I combined qualitative and quantitative methods in a cross-sectional

study with three separate phases: 1) establishing, testing, piloting, and revising study tools (e.g.,

questionnaires and water testing protocols); 2) gathering data from target groups of farmers

through a full field survey that included questionnaires and water quality testing; and 3)

analyzing data and publishing results. Initially, I conducted a pilot study to test and revise the

4 In my research, drinking water is defined as “water used for direct drinking or food processing”. Domestic water is

defined as “water use for domestic purposes but nor for direct drinking or processing food” (MoH, 2009; MoH,

2009) 5 The WHO recommends Escherichia coli (E. coli) colony forming unit (cfu), turbidity, and pH as basic microbial

and related indicators for assessing rural water quality (WHO, 2012)

of 340

study tools before using them in the full field survey where I randomly selected a sample size of

600 eligible households representing SSI farming to participate in my research. Key variables

included demographics, socioeconomic and livestock production characteristics, SSI farmers’

perceptions of risk factors for transmitting WRZD, and the actions SSI farmers take to reduce

transmitting WRZD. Variables surrounding SSI farmers’ engagement in mitigating strategies to

reduce transmitting WRZD were generated using WHO’s recommended mitigating strategies to

cluster the data on farmers’ actions to reduce transmitting WRZD.

Most waves of the HPAI (H5N1) epidemic during the 2003 – 2006 period occurred in the

river delta areas of my two target research areas in Thai Binh and An Giang provinces: these are

two typical provinces of Vietnam in the Red River Delta and Mekong River Delta respectively.

Detailed information of the overall methods used in my research is presented in Chapter

Two. Chapters Three, Four, and Five provide additional details of methods that specifically

relate to each of the chapters.

1.5. A Description of Chapters in the Dissertation

This dissertation consists of six individual chapters that are structured in a manuscript

style format. Beside the introduction and conclusion chapters, I present the research results as a

collection of four manuscript-based chapters that were prepared for submitting to peer-reviewed

journals preferably in the field of agricultural economics, water quality, and public health.

Although each manuscript-based chapter is a standalone manuscript, I arranged the chapters of

this dissertation in a logical and continuous order to present the results according to the order of

my research questions.

In the next section, I present a flow chart outlining how the research questions relate to

each manuscript-based chapter (See Figure 1.2) and a description of chapters in the dissertation.

of 340

Figure 1.2. Dissertation’s Outline

Chapter 2:

Develops a socioeconomic

profile of SSI farmers in

Thai Binh and An Giang

Provinces of Vietnam

Chapter 3:

Presents SSI farmers'

perceptions of risk factors

for WRZD transmission

in Thai Binh and An

Giang provinces of

Vietnam

Chapter 4:

Presents the water quality

among SSI farms in Thai

Binh and An Giang

provinces of Vietnam

Chapter 5:

Presents factors

associated with SSI

farmers' engagement in

mitigating strategies to

reduce transmission of

WRZD in Thai Binh and

An Giang provinces of

Vietnam

Research question 1:

What is the socioeconomic profile of SSI farmers in Thai

Binh and An Giang provinces of Vietnam?


Is the socioeconomic profile of SSI farmers in Thai Binh and

An Giang consistent with an EcoHealth approach in the

context of SSI farming?


What are SSI farmers’ perceptions of risk factors for water

related zoonotic disease (WRZD) transmission in Thai



What are sources and the basic quality of water used by

SSI farmers for drinking and domestic purposes in Thai



How does the quality of water used by SSI farmers relate

to their perceptions of water quality and risk factors for

transmission of WRZD?


Do SSI farmers engage in mitigating strategies to reduce



What are factors associated with SSI farmers’ engagement

in mitigating strategies to reduce transmission of WRZD?

of 340

Chapter One provides an overview of the research, a background of my research, a

summary of each individual chapter in the dissertation, and connections between chapters in the

dissertation. I also acknowledge the contributions of other authors to manuscripts that were

prepared based on this dissertation.

Chapter Two explores the first two research questions by presenting a profile of SSI

farming in the regions of research within the context of the increased documented risk factors

associated with the transmission of WRZD. The first objective of this chapter is to describe the

characteristics of SSI farming within the geographical research regions while considering the risk

factors for transmitting WRZD. The second objective is to use an EcoHealth approach6 to

discuss the potential ecological and health implications of SSI farming in the context of risk

factors for transmitting WRZD. Before conducting the research, I anticipated that the SSI

farming practices used in Thai Binh and An Giang would expose farmers to risk factors for

transmitting WRZD. Furthermore, I expected that not all the characteristics of SSI farming

practices in Thai Binh and An Giang were consistent with an EcoHealth approach. Finally, I

used descriptive and analytical statistics to assist in the analysis of data collected for this chapter.

Based on this chapter, two manuscripts were prepared and submitted for journal publication. My

committee members are acknowledged in the manuscripts for their comments and advice on this

chapter.

6 In my research, EcoHealth (or ecosystem approaches to health) is referred to as an approach that explores the

interface between human health, animal health, and environmental health by using six key pillars/principles:

complexity, trans-disciplinarity, community participation, gender and social equity, sustainability, and knowledge to

action (Hall, 2010; IDRC, 2012). The International Development Research Centre (IDRC) initiated EcoHealth and

developed it based on many years of health and environment research(IDRC, 2012).

of 340

Chapter Three addresses the third research question by presenting a study of SSI farmers’

perceptions about the risk factors for transmitting WRZD in the study areas. Here, I use both the

conceptual and theoretical models for my research (See Appendix A). I also discuss the

importance of these perceptions in researching how SSI farmers engage in mitigating strategies

to reduce transmitting WRZD. My specific objective was to describe four constructs surrounding

SSI famers’ perceptions of risk factors for transmitting WRZD. The first construct was SSI

farmers’ impressions about the quality of general water environment in their villages. The second

construct was SSI farmers’ anticipation of threat to health and/or wellbeing of transmitting

WRZD including the projected susceptibility to and severity of transmission of WRZD. The third

construct was SSI farmers’ perceived expectations including sensed barriers, supposed benefits,

and believed self-efficacy in mitigating the transmission of WRZD. The fourth construct was

beliefs about triggers/cues7 to taking actions that would mitigate transmitting WRZD (See

Appendix A). Descriptive statistics were used to assist in the data analysis for this chapter.

Chapter Four addresses the fourth and fifth research questions by presenting a study of

the types and qualities of water sources that SSI farms in the study area typically use. I also

explore the association between on-farm water quality and the perceptions identified in Chapter

Two and discuss the public health implications of this association. The objectives of my research

in Chapter Four were to 1) describe the sources and assess the basic microbial and related water

quality SSI farms use and 2) to explore the connection between farm water quality and farmers’

perceptions of risk factors for transmitting WRZD. I expected SSI farmers to use multiple water

sources for drinking and domestic purposes. I hypothesized that the frequency of water source

7 A trigger/cue to action is “the stimulus needed to trigger the decision making process to accept a recommended

health action”. Examples of possible cues to actions are chest pains, mass media campaigns, advice from others, a

reminder postcard from a physician, illness of family member or friend, etc.) (BUMC, 2013)

of 340

use for drinking and domestic purposes among SSI farmers in Thai Binh was different from

those in An Giang, and that these water sources in both provinces are of low quality compared to

water quality standards in Vietnam. My other hypothesis in this chapter was that the

socioeconomic status of SSI farmers and their perceptions about risk factors for transmitting

WRZD are associated with microbial content of water on farms. Different methods of data

collection, specifically questionnaires, in-depth interviews, farm visits, and focus group

discussions were used to gather data on the water sources SSI farmers have access to and their

corresponding frequency of use for drinking and domestic purposes.

In Chapter Four, I use two protocols to assess objectively the basic microbial and related

quality indicators (i.e., pH, turbidity, and E. coli colony forming unit count) of on-farm water for

drinking and domestic purposes. These protocols are: 1) a protocol for laboratories to assess on-

farm water quality and 2) a protocol for community health/veterinary workers (CHWS/CVWs) to

quantify E. coli in on-farm water. These protocols follow WHO’s recommendations for assessing

basic indicators of rural water quality (WHO, 2011; WHO, 2011; WHO and OECD, 2003) and

conform with Vietnam’s national technical regulations on drinking and domestic water quality

(MoH, 2009; MoH, 2009). This chapter additionally discusses implications for the research

results for rural water, public health, and for the water regulatory framework for SSI farms in

Vietnam. Descriptive statistics were used to analyze and present the results related to water

sources, frequency of use, and on-farm water quality. Regression analysis was used to model and

test the association between on-farm water quality and SSI farmers’ perceptions of water quality

and risk factors for transmitting WRZD.

Chapter Five addresses the sixth and seventh research questions by presenting SSI

farmers’ engagement in, and the factors associated with mitigating strategies to reduce

of 340

transmitting WRZD (e.g., livestock management). I hypothesized that the socioeconomic status

of SSI farmers and their perceptions of risk factors for transmitting WRZD are associated with

their mitigating strategies to reduce transmitting WRZD. This chapter also discusses implications

for associations between formulating interventions and policies to improve the health of farmers

and animals, and their environment. Descriptive statistics and probit regression were used to

assist in the data analysis, to test the hypothesis, and to present the results.

In Chapter Six, I summarize the results presented in Chapters Two to Five and discuss the

limitations of my research. Then, I provide recommendations for future research and formulating

and revising programs and policies related to WRZD, livestock production, rural water, and

public health in Vietnam. Finally, I conclude my research study.

1.6. Connection Between Chapters

Overall, the results of the research are connected through the chapters in the dissertation.

The results from Chapter Two and Three were used to assist the studies in Chapter Four and

Five. In particular, the results from the research in chapters Two and Three were used as a

foundation for Chapter Four to model factors that are associated with on-farm water quality.

Similarly, the results from chapters Two, Three, Four were used in Chapter Five to explore

factors that are associated with SSI farmers’ engagement in mitigating strategies to reduce

transmitting WRZD. In order to ensure that each manuscript-based chapter is a stand-alone

chapter, there will be some repetition of information in the text of the chapters.

of 340

References

Bach, A. H. (2007). Equivalence in the translation of Vietnamses cultural words in the book

"Wandering through Vietnamese culture" by Huu Ngoc. Vietnam College of Foreign

Languages (VHUN), Hanoi.

Becker, M. H. (2012). The Health Belief Model and Personal Health Behavior. (4 ed.) (vols. 2).

Boere, G. C., Galbraith, C. A., and Stroud, D. A. (2006). Waterbirds Around the World The

Stationery Office, UK.

Brown, J. D., Swayne, D. E., Cooper, R. J., Burns, R. E., and Stallknecht, D. E. (2007).

Persistence of H5 and H7 Avian Influenza viruses in water. Avian Diseases Digest, 2,

e22.

Bui, T. D., Hien, N. T., and Dang, T. T. (2010). Distribution of freshwater snails in family-based

VAC ponds and associated waterbodies with special reference to intermediate hosts of

fish-borne zoonotic trematodes in Nam Dinh Province, Vietnam. Acta Tropica, 116, 15-

23.

Bui, X. M. (2010). Duck farming systems and avian influenza in the Mekong Delta of Vietnam.

In FAO Rome.

BUMC (2013). The Health Belief Model. BU School of Public Health [On-line]. Available:

http://sphweb.bumc.bu.edu/otlt/MPH-Modules/SB/SB721-Models/SB721-Models2.html

Carr, R. and Bartram, J. (2004). The control envelope and risk management. In (pp. 67-92). IWA

Publishing.

Dalsgaard, A. (2001). Health Aspects of the Reuse of Wastewater in Agriculture and

Aquaculture in Vietnam. In Wastewater Reuse in Agriculture in Vietnam: Water


of 340

Management, Environment and Human Health Aspects International Water Management

Institute.

Dang, N. K., Milstein, A., Verdegem, M. C. J., and Verreth, J. A. V. (2006). Food inputs, water

quality and nutrient accumulation in integrated pond systems: A multivariate approach.

Aquaculture, 261, 160-173.

Dao, L. H. (2006). Situation and future direction of environmental protection in livestock

production, Vietnam Department of Animal Husbandary, Ministry of Agriculture and

Rural Development, Vietnam.

Domanska, B. K., Minta, Z., Smietanka, K., Sylvie, M., and Berg, A. T. (2010). H5N1 High

Pathogenicity Avian Influenza Virus Survival in Different Types of Water. Avian

Diseases Digest, 5, e180-e181.

Ford, T. E. and Colwell, R. (1996). A Global Decline in Microbiological Safety of Water: A Call

for Action. American Academy of Microbiolog.

Gilbert, M., Chaitaweesub, P., Parakamawongsa, T., Premashthira, S., Tiensin, T., Kalpravidh,

W. et al. (2006). Free-grazing ducks and highly pathogenic avian influenza, Thailand.

Emerging Infectious Diseases, 12, 227-234.

Gilbert, M., Xiao, X., Pfeiffer, D. U., Epprecht, M., Boles, S., Czarnecki, C. et al. (2008).

Mapping H5N1 highly pathogenic avian influenza risk in Southeast Asia. Proceedings of

the National Academy of Sciences, 105, 4769-4774.

Gleeson, L. J. and Dung, D. H. (2010). Mid-Term Review: Vietnam Integrated National

Operational Program for Avian and Human Influenza (OPI) 2006-2010: HPAI Control

and Eradication in the Agriculture Sector.

of 340

Guan, T. Y. and Holley, R. A. (2003). Pathogen survival in swine manure environments and

transmission of human enteric illness – a review. Journal of Environmental Quality, 32

(2), 392.

Hall, D. C. (2010). The Ecohealth Approach: An introduction to the key concepts. Building

Ecohealth Capacity in Asia, Veterinarians Without Border Canada [On-line]. Available:

https://www.vetswithoutborders.ca/asia/ecohealth

Hall, D. C., Benigno, C., and Wantanee, K. (2006). The impact of avian influenza on small and

medium scale poultry producers in South East Asia (preliminary findings) Food and

Agriculture Organization (FAO).

Hall, D. C., Thao, T. D., Minh, D. V., and Lien, L. V. (2006). Competitiveness of the livestock

sector in Vietnam World Bank paper. WB-EASRD and FAO-TCIP.

Hanington, P., Hall, D. C., and Anyan, W. (2016). The role of environmental management in

snail-borne zoonotic diseases. In S. C. Cork, D. C. Hall, and K. Liljebjelke (Eds.), One

Health Case Studies: Addressing Complex Problems in a Changing World (Expected

2016). 5M Publishing, Sheffield, UK.

Henning, J., Henning, K. A., Morton, J. M., Long, N. T., Ha, N. T., Vu, L. T. et al. (2011).

Highly pathogenic avian influenza (H5N1) in ducks and in-contact chickens in backyard

and smallholder commercial duck farms in Viet Nam. Preventive Veterinary Medicine,

101, 229-240.

Henning, J., Pfeiffer, D., and Vu, L. (2009). Risk factors and characteristics of H5N1 Highly

Pathogenic Avian Influenza (HPAI) post-vaccination outbreaks. Vet.Res., 40.

http://www.vetswithoutborders.ca/asia/ecohealth

of 340

Herbst, S., Benedikter, S., Koester, U., Phan, N., Berger, C., Rechenburg, A. et al. (2009).

Perceptions of water, sanitation and health: a case study from the Mekong Delta,

Vietnam. Water Science and Technology, 60, 699-707.

Hoang, B. T. (2011). Rural employment and life: Challenges to gender roles in Vietnam's

agriculture at present Research Centre for Gender, Family and Environment in

Development, Viet Nam.

IDRC (2012). Ecohealth Research in Practice: Innovative Applications of an Ecosystem

Approaches to Health. International Development Centre (IDRC); Springer.

Irini, M. and Rapsomanikis, G. (2005). Contribution of Livestock to Household income in

Vietnam: A household typology based analysis (Rep. No. PPLPI Working Paper N.21). A

Living from Livestock, FAO.

Janz, N. K. and Becker, M. H. (1984). The Health Belief Model: A Decade Later. Health

Education and Behavior, 11, 1-47.

Kamakawa, A., Thu, H. T. V., and Yamada, S. (2006). Epidemiological survey of viral diseases

of pigs in the Mekong delta of Vietnam between 1999 and 2003. Veterinary

Microbiology, 118, 47-56.

Keraita, B., Drechsel, P., Seidu, R., Amerasinghe, P., Cofie, O., and and Konradsen, F. (2010).

Harnessing Farmers' Knowledge and Perceptions for Health-Risk Reduction in

Wastewater-Irrigated Agriculture. In P.Drechsel, C.A.Scott, L.Raschid-Sally,

M.Redwood, and and A.Bahri (Eds.), Wastewater Irrigation and Health: Assessing and

Mitigating Risk in Low-Income Countries (pp. 337-353). London: Earthscan Publications

Ltd.

of 340

Keraita, B., Drechsel, P., Seidu, R., Amerasinghe, P., Cofie, O., and Konradsen, F. (2010).


Wastewater-Irrigated Agriculture. In P.Drechsel, C. A. Scott, L. Raschid-Sally, M.

Redwood, and A. Bahri (Eds.), Wastewater irrigation and health : assessing and

mitigating risk in low-income countries (pp. 337-353). London: Earthscan Publications

Ltd.

Keraita, B., Jim'nez, B., and Drechsel, P. (2008). Extent and implications of agricultural reuse of

untreated, partly treated and diluted wastewater in developing countries. Agr Vet Sci Nutr

Nat Resour, 3, 15.

Knudsen, L. G., Pham, D. P., Nguyen, H. T., Helle, S., Jensen, P. K., Dalsgaard, A. et al. (2008).

The fear of awful smell: risk perceptions among farmers in Vietnam using wastewater

and human excreta in agriculture. The Southeast Asian journal of tropical medicine and

public health, 39, 341-352.

MARD (2012). Proposal on restructuring the agricultural sector towards greater added value

and sustainable develoment Hanoi: Vietnam Ministry of Agriculture and Rural

Development.

MoC and MARD (2000). National Rural Clean Water Supply and Sanitation Strategy up to the

Year 2020 Hanoi, Vietnam: Ministry of Construction and Ministry of Agriculture and

Rural Development of Vietnam.

Moe, C. L. (2004). What are the criteria for determining whether a disease is zoonotic and water

related. In World Health Organization (WHO). Waterborne Zoonoses: Identification,

Causes and Controls (pp. 27-45). London: IWA Publishing.

of 340

MoH (2009). National technical regulation on domestic water quality (QCVN 02: 2009/BYT).

Hanoi: Department of Preventive Medicine and Environment, Ministry of Health,

Socialist Republic of Vietnam.

MoH (2009). National technical regulation on drinking water quality (QCVN 01: 2009/BYT).



Morley, C. (2010). Bio-Security Stakeholder Analysis, Vietnam (Rep. No. OSRO/RAS/604/USA

B6). The Emergency Centre for Transboundary Animal Diseases (ECTAD), Food and

Agriculture Organization of the United Nations (FAO).

Morris, R. S. and Jackson, R. (2005). Epidemiology of H5N1 Avian Influenza in Asia and

Implications for Regional Control Rome, Italy: Food and Agriculture Organisation of the

UN.

Nazir, J., Haumacher, R., Ike, A., Stumpf, P., Bohm, R., and Marschang, R. E. (2010). Long-

term study on tenacity of avian influenza viruses in water (distilled water, normal saline,

and surface water) at different temperatures. Avian Dis, 54, 720-724.

Nguyen, H. V., Vuong, T. A., Pham, D. P., and Vu, V. T. (2011). Safe use of wastewater in

agriculture and aquaculture (Rep. No. 2). The National Centre of Competence in

Research (NCCR) North-South.

Owusu, V., Bakang, J. E. A., Abaidoo, R. C., and Kinane, M. L. (2012). Perception on untreated

wastewater irrigation for vegetable production in Ghana. Environ Dev Sustain, 14, 135-

150.

of 340

Owusu, V., Bakang, J. E., Abaidoo, R., and Kinane, M. (2011). Perception on untreated

wastewater irrigation for vegetable production in Ghana. Environment, Development and

Sustainability, 1-16.

Pfeiffer, D. U., Minh, P. Q., Martin, V., Epprecht, M., and Otte, M. J. (2007). An analysis of the

spatial and temporal patterns of highly pathogenic avian influenza occurrence in Vietnam

using national surveillance data. Vet J, 174, 302-309.

Pham, D. P., Nguyen, V. H., Hattendorf, J., Zinsstag, J., Phung, D. C., and Odermatt, P. (2011).

Risk factors for Entamoeba histolytica infection in an agricultural community in Hanam

province, Vietnam. Parasites and Vectors, 4, 102.

Pica-Ciarmarra, U., Tasciotti, L., Otte, J., and Zezza, A. (2011). Livestock Assets, Livestock

Income and Rural Households: Cross-Country Evidence from Household Surveys (Rep.

No. ESA Working Paper No. 11-17). Rome: World Bank, FAO, AU-IBAR, ILRI.

Porphyre, V. and Nguyen, Q. C. (2007). Pig production development animal-waste management

and environment protection: a case study in Thai Binh Province, Northern Vietnam.

PRISE Publications, France.

Rechenburg, A. and Herbst, S. (2006). Hygienic Aspects of Sanitation and Water in the Rural

Areas of the Mekong Delta, Vietnam (Rep. No. "Prosperity and Poverty in a Globalised

World Challenges for Agricultural Research"). University of Bonn, Institute for Hygiene

and Public Health, Germany.

Roche, B., Lebarbenchon, C., Gauthier-Clerc, M., Chang, C. M., Thomas, F., Renaud, F. et al.

(2009). Water-borne transmission drives avian influenza dynamics in wild birds: the case

of the 2005-2006 epidemics in the Camargue area. Infect Genet Evol, 9, 800-805.

of 340

Rosenstock, I. M., Strecher, V. J., and Becker, M. H. (1988). Social Learning Theory and the

Health Belief Model. Health Education and Behavior, 15, 175-183.

Stallknecht, D. E., Shane, S. M., Kearney, M. T., and Zwank, P. J. (1990). Persistence of Avian

Influenza Viruses in Water. Avian diseases, 34, 406-411.

Thang, P., Tuan, H., and Huton, G. (2008). Economic impacts of sanitation in Vietnam. World

Bank.

Toews, D. W. (2009). Food, Global Environmental Change and Health: EcoHealth to the

Rescue? Mcgill Journal of Medicine, 12, 85-89.

Trang, D. T., Bui, T. T., Molbak, K., Phung, D. C., and Dalsgaard, A. (2007). Epidemiology and

aetiology of diarrhoeal diseases in adults engaged in wastewater-fed agriculture and

aquaculture in Hanoi, Vietnam. Trop Med Int Health, 12, 23-33.

Trinh, T. L. (2001). Wastewater and its Treatment System in the Southern Part of Vietnam. In

International Water Management Institute: Proceedings of a Workshop held in Hanoi,

Vietnam 14 March 2001 Southern Institute of Water Resources Research.

VPMO (2010). About the organisation of the national poverty survey to prepare for the

implementation of social secuity program and policy in the period of 2011- 2015 (Rep.

No. 1752/ CT-TTg). Vice Prime Minister Office (VPMO), the Government of Vietnam.

Vu, T. K. V., Tran, M. T., and Dang, T. T. S. (2007). A survey of manure management on pig

farms in Northern Vietnam. Livestock Science, 112, 288-297.

Vu, T. T. H. (2001). Treatment of Domestic Wastewater and its Reuse in Farm Irrigation in the

Red River Delta. In Wastewater Reuse in Agriculture in Vietnam: Water Management,

Environment and Human Health Aspects Center for Irrigation and Water Supply,

Vietnam Institute for Water Resources Research.

of 340

Ward, M. P., Maftei, D. N., Apostu, C. L., and Suru, A. R. (2009). Association Between

Outbreaks of Highly Pathogenic Avian Influenza Subtype H5N1 and Migratory

Waterfowl (Family Anatidae) Populations. Zoonoses and Public Health, 56, 1-9.

WB (2002). Memorandum of the President of the International Development Association and the

International Finance Corporation to the Executive Directors on a Country Assistance

Strategy of the World Bank Group for the Socialist Republic of Vietnam (Rep. No. Report

No: 24621-VN). Vietnam Country Management Unit East Asia and the Pacific Region.

The World bank.

WB (2006). Vietnam Food Safety and Agricultural Health Action Plan (Rep. No. Report No.

35231 VN).

WB (2006). Vietnam Plan Action for Food Security and Animals, Plant Protection Hanoi: World

Bank.

Webster, R. G., Peiris, M., Chen, H., and Guan, Y. (2006). H5N1 outbreaks and enzootic

influenza. Emerg Infect Dis [serial on the Internet].


influenza. Emerg Infect Dis [serial on the Internet]. Emerging Infectious Diseases.

Weldesilassie, A. B., Boelee, L. I. N., Drechel, P. A. Y., and Dabbert, S. T. E. P. (2011).

Wastewater use in crop production in peri-urban areas of Addis Ababa: impacts on health

in farm households. Environment and Development Economics, 16, 25-49.

WHO (2004). Waterborne Zoonoses: Identification, Causes, and Control. IWA Publishing.

WHO (2006). Volume II: Wastewater use in agriculture.

WHO (2007). Review of latest available evidence on potential transmission of avian influenza

(H5N1) through water, sewage and ways to reduce the risks to human health (Rep. No.

of 340

WHO/SDE/WSH/06.1). Geneva: World Health Organisation: Water, Sanitation and

Health Public Health and Environment.

WHO (2011). Guidelines for drinking-water quality. (Fourth ed.) World Health Organisation.

WHO (2011). Microbial aspects. In Guidelines for drinking-water quality (pp. 117-153). World

Health Organization.

WHO (2012). Rapid assessment of drinking water quality - A handbook for implementation.

World Health Organization.

WHO and OECD (2003). Assessing Microbial Safety of Drinking Water: Improving Approaches

and Methods. IWA Publishing.

Yajima, A. and Kurokura, H. (2008). Microbial risk assessment of livestock-integrated

aquaculture and fish handling in Vietnam. Fisheries Science, 74, 1062-1068.

Zheng, T., Adlam, B., Rawdon, T. G., Stanislawek, W. L., Cork, S. C., Hope, V. et al. (2010). A

cross-sectional survey of influenza A infection, and management practices in small rural

backyard poultry flocks in two regions of New Zealand. New Zealand Veterinary

Journal, 58, 74-80.

of 340

Chapter Two: Developing a Socioeconomic Profile of Small-Scale Integrated (SSI) Farms

in the Provinces of Thai Binh and An Giang in Vietnam

2.1. Introduction

2.1.1. Overview of Small-Scale Integrated (SSI) farming in Vietnam

In Vietnam, as in other developing countries, economic growth relies primarily on

agricultural production and industrialization associated with rapid urban development. In this

environment in Vietnam, integrated aquaculture agriculture farming is a common model in rural

areas (Dang et al., 2006; Vu et al., 2007). In my research, SSI farming is defined as farming

practices in which farmers combine some types of livestock rearing and directly or indirectly

depend on agriculture (Hall et al., 2006; Irini and Rapsomanikis, 2005; Pica-Ciarmarra et al.,

2011). SSI farming and/or farmers are often referred to by various terms such as smallholder

farmers, the VAC model8, small-scale producers, and/or SSI farms (Bui et al., 2010). However,

most farmers practicing SSI farming benefit the least from national social economic policies and

economic growth (Morley, 2010). For instance, in Vietnam, it is estimated that 10 million rural

farmers still live without adequate sanitation and 9.3 million lack safe drinking water (VPMO,

2010).

The VAC model is an example of SSI farming that encompasses a predominantly small-

scale operation of poultry, pigs, fish, and crops (Bui et al., 2010). The VAC model is a common

integrated aquaculture, agricultural farming model (Bosma et al., 2006; Bui et al., 2010; Dang et

al., 2007; Le et al., 2007). Some farmers also rely on raising free-grazing ducks (Henning et al.,

2011) and dairy cattle (Geurden et al., 2008). Backyard livestock rearing is another example of

SSI farming model, which plays an integral role in recycling nutrients in farming systems and

8 The VAC model: V= Vuon (i.e., Vegetable garden), A= Ao (i.e., Fish pond), C= Chuong (i.e., Livestock housing)

of 340

generating valuable cash income. This model can work effectively but the development of

backyard livestock rearing has been hampered by infectious diseases such as fatal epizootics of

piglets and reproductive failure of sows caused by classical swine fever (CSF) virus (Kamakawa

et al., 2006). Moreover, in this model, a significant amount of untreated excreta of animals and

humans is used as fertilizer, fish food, irrigation, or is dumped, which contaminates water

sources with excreted pathogens (Vu et al., 2007; Pham et al., 2011; Dalsgaard, 2001).

More than 80% of SSI farmers own some forms of livestock (Pica-Ciarmarra et al., 2011)

with 87% of them specifically rearing poultry (Hall et al., 2006; Irini and Rapsomanikis, 2005).

More than 50% of farmers also depend on livestock for nutrition and income (Irini and

Rapsomanikis, 2005) and most of their cash revenue comes from selling live animals rather than

livestock products such as broiler meat (Kamakawa et al., 2006).

In SSI farming, economic factors are one of the most important considerations in driving

access to health care for people in the rural areas. For instance, SSI farmers often consider the

quality of health services and sociocultural and economic factors (e.g., living with extended

family or cost of services), rather than geographical access in utilizing health services (Duong et

al., 2004). Some studies have also reported the health behaviour of SSI farmers as a concern. For

instance, although SSI farmers have been engaged in disease control methods such as

vaccination and bio-security (Gleeson and Dung, 2010), they most commonly deal with infected

animals by salvaging them (i.e., sell or eat sick animals). Gleeson and Dung (2010) further

argued that when SSI farmers confronted a problem (e.g., a zoonotic disease) that no one,

including animal health services, had experience with, they panicked, which further promoted

the spread of the disease. Such situations emphasize the importance of understanding the

farmers’ sense of self-efficacy in dealing with public health or animal health problems.

of 340

2.1.2. Rationale for the research

Water has a profound influence on human health and water-related zoonotic diseases

(WRZD)9 place a heavy burden on the population health services of many countries, especially

developing countries(WHO, 2011). Farmers appear to consider water use an important factor

mainly because of its implications for agriculture and economic development, rather than for

their own health or that of their animals (Bui et al., 2010). Specifically, in Vietnam, there are

increasing reports of WRZDs especially among SSI farmers and their animals in the Red River

and Mekong River Deltas (ProMED, 2014; MoH and MARD, 2011; MARD and MoH, 2006).

However, their awareness level about the health risk factors associated with transmitting WRZD

remains low (Knudsen et al., 2008; Keraita et al., 2010). Thus, it appears that although SSI

farming represents one of the largest agricultural sectors in Vietnam (MARD, 2012), it will be

challenging to reduce transmission of WRZD (see Table 2.1).

In Canada, the International Development Research Centre (IDRC) has initiated

ecosystem (or EcoHealth) approaches to health that have their foundation in many years of

health and environment research (Charron, 2012). Ecosystem approaches to health explore the

interface between humans, animals, and the environment through using six key pillars/principles

consisting of complexity, trans-disciplinarity, community participation, gender and social equity,

sustainability, and knowledge to action (Hall, 2010; IDRC, 2012).

9 In my research, I used WHO’ definition of WRZD as diseases in humans that are spread from animals and are

related to water such as Highly Pathogenic Avian Influenza, some parasites, or diseases caused by a range of

bacteria (e.g., pathogenic strains of Escherichia coli) (WHO, 2004). Criteria for determining WRZD include: 1) the

pathogen must spend part of its life cycle within one or more animal species; 2) it is probable or conceivable that

some life stage of the pathogen will enter water; and 3) transmission of the pathogen from animals to humans must

be through a water related route (Moe, 2004).

of 340

Internationally, case studies and reports have been presented that bring forward the issues

of SSI livestock management, improvement in famers’ incomes, and reduction of WRZD

(Waltner-Toews, 2011; IDRC, 2012). However, to date in Vietnam, I have not found any studies

that provide an EcoHealth insight into SSI farming and that explore farmers’ perceptions of risk

factors for transmitting WRZD and link their mitigation strategies10 for reducing the transmission

of WRZD. Therefore, my research is significant because I use an EcoHealth approach to

examine and describe the socioeconomic profile of SSI farming in the two main river deltas in

Vietnam, while focusing on general socioeconomic status, demographics, and characteristics of

livestock production.

My research highlighted avian influenza (AI) including Highly Pathogenic Avian

Influenza H5N1 (HPAI) because it is one of the WRZD global threats, and it has had a

significant impact on the health and livelihoods of SSI farmers in Vietnam. The major proportion

of the HPAI epidemics happened during 2003-2005 in the rural areas of the two main river deltas

in Vietnam where the most vulnerable SSI farmers reside (Hoang, 2006; Rushton et al., 2006).

Moreover, a number of reported risk factors for transmitting WRZD relate to SSI farming

practices. Water bodies are also believed to be modes of transmission and temporary reservoirs

for AI viruses at the village and farm level (Morris and Jackson, 2005). As well, AI viruses can

persist for extended periods in water, sewage, excreta, and animal wastes varying on different

factors (Boere et al., 2006; Morris and Jackson, 2005; Webster et al., 2006; WHO, 2007).

10 In my research, I defined a mitigating strategy to reduce the transmission of water-related zoonotic diseases

(WRZD) as a plan (e.g., a plan for managing livestock) that includes one or more activities an SSI farmer could

choose to reduce transmitting WRZD. An example of a mitigating strategy is good livestock management, which

may include actions, or tactics (e.g., treat animal waste at source) that SSI farmers could choose to reduce

transmitting WRZD (Carr and Bartram, 2004; WHO, 2004).

of 340

The values of incorporating the principles of an EcoHealth approach are considered in

my research. Together with the research outlined in other chapters of this dissertation, the

research results in this chapter will assist human and animal health authorities, as well as

research and development agencies, to understand better how to influence current SSI farming

management practices. Particularly, the research results will be useful for SSI farmers, health

workers, veterinarians, researchers, and policy makers who have the potential to help revise,

formulate, and implement government policies in SSI farming and in water and health. The

research results will be useful for considering an EcoHealth approach in the context of SSI

farming to reduce risk factors for transmitting WRZD.

2.1.3. SSI farming and risk factors for transmission of HPAI

Highly Pathogenic Avian Influenza (HPAI) during the 2003-10 period in Vietnam caused

119 human cases of HPAI disease - the second highest human case incidence in the world and

case-fatality of 50% (Soares et al., 2010; WHO, 2010). HPAI also had significant impact on

Vietnamese socioeconomic condition with an estimate of 50 million poultry culled (Figuie,

2007; Otte, Hinrichs, Rushton, Roland-Holst, and Zilberman, 2008; Soares et al., 2010) while

68% of poor rural households rely on poultry as a primary source of income (FAO, 2010). The

reported risk factors for transmitting the HPAI virus are related to poultry production and trade,

behaviour and culture, interactions among wild and domestic birds, the water environment,

farming practices, and spatial and temporal variation (See Table 2.1.). The risk factors for

transmitting HPAI are also more likely driven by characteristics operating at the household level

rather than at the village level (Phan, 2010). Ducks pose a significant risk factor in the spread of

low pathogenic avian influenza (LPAI) as the virus can infect ducks without producing clinical

signs. In fact, several studies have concluded that ducks are asymptomatic reservoirs and

of 340

transmitters of HPAI (Bui, 2010; Hulse-Post et al., 2005; Munster et al., 2005; Tracey et al.,

2004). Scavenging ducks are common and are thought to play an important role in maintaining

and transmitting the HPAI virus, but there is a lack of published information about the HPAI

virus infection status among these animals (Henning et al., 2009). Consuming some products

(e.g., eating fresh duck blood) is also considered a high risk factor in the transmission of HPAI to

humans (Care, 2011; Zuckerman, 2011). With respect to seasonality, HPAI outbreaks are now

occurring in the warmer months of the year with ducks being the prominent affected species in

the HPAI (H5N1) outbreaks (Minh et al., 2009).

SSI farming practices are also thought to play an important role in maintaining and

transmitting some infectious diseases (e.g., H5N1) due to the increased potential for interspecies

mixing among livestock, humans, and wild animals (Geurden et al., 2008; Henning et al., 2011).

Table 2.1 provides more information about the main groups of risk factors for transmitting HPAI

as reported by other researchers. Most of these risk factors can be linked with SSI farming

practices.

In summary, SSI farming practices play an important economic role for Vietnam’s rural

population. Nevertheless, there are differences between SSI farming practices in the Red River

Delta and those in the Mekong River Delta. Exploring whether characteristics of these SSI

farming practices are consistent with an EcoHealth approach can help determine potential health

risk factors, and reduce undesirable health consequences for SSI farmers, their animals, and their

environment.

2.1.4. Research questions and objectives

This chapter addresses the first two questions of my research. These questions are 1)

What is the socioeconomic profile of SSI farmers in the provinces of Thai Binh and An Giang in

of 340

Vietnam? and 2) Is the socioeconomic profile of SSI farmers in the provinces of Thai Binh and

An Giang consistent with an EcoHealth approach in the context of SSI farming? More

specifically, my research in this chapter has two objectives. The first objective is to describe

characteristics of SSI farming in the research area in the context of the risk factors for

transmitting WRZD. The second objective is to discuss any potential ecological and health

implications of SSI farming in the context of risk factors for transmitting WRZD through using

an EcoHealth approach (See Table 2.1). I expect that the SSI farming practices used in the

provinces of Thai Binh and An Giang expose farmers to risk factors for transmitting WRZD.

Furthermore, I expect that not all characteristics of SSI farming practice in Thai Binh and An

Giang are consistent with an EcoHealth approach in the context of SSI farming.

of 340

Table 2.1. Risk Factors for Transmission of Highly Pathogenic Avian Influenza (HPAI)

and Their Possible Connection with Small-Scale Integrated (SSI) Farming

Reported risk factors for transmission of HPAI

(order based on subjective priority11)

References

1. Poultry production (e.g., number of chickens,

unevenly distributed poultry supply-and-demand

system, contact with sick or dead poultry)

(Biswas P.K. et al., 2009a; Dinh et al.,

2006; Gilbert et al., 2004; Sheta et al.,

2014; Tiensin et al., 2009; Zhao et al.,

2013; Zhou et al., 2009)

2. Poultry trade (e.g., live poultry trade, poultry

slaughter house, and trade of poultry between farms)

(Desvaux et al., 2014; Fasina et al., 2011;

Henning et al., 2009; Paul et al., 2011;

Sims, 2007; Soares et al., 2010)

3. Duck production (e.g., number of ducks, and

scavenging ducks in rice field)

(Gilbert et al., 2006; Tiensin et al., 2009;

Paul et al., 2014)

4. Behaviours and culture (e.g., eating fresh duck

blood, visiting a stall or market selling live poultry

during the week before the illness)

(Care, 2011; Zuckerman, 2011)

5. Interaction between wild and domestic birds (e.g.,

abundance of domestic ducks and geese, and kept

birds/animals other than poultry)

(Boere et al., 2006; Sheta et al., 2014;

Tiensin et al., 2009; Zheng et al., 2010)

6. Relating to water and environment (e.g., water

bodies at the village and farm level and

environmental characteristics)

(Morris and Jackson, 2005; Paul et al.,

2014)

7. On-farm management practices (e.g., farm

visitors, visiting other farms, and farm biosecurity

management/practices)

(Fasina et al., 2011)

8. Spatial and temporal variation (e.g., river length

within a commune, poultry in close proximity and

increased accessibility to a water body/sources,

temporal flooding, seasonality, and distance to

water sources)

(Biswas P.K. et al., 2009b; Dang et al.,

2006; Fang et al., 2008; Vu et al., 2007;

Ward et al., 2008)

9. Other on-farm production (e.g., vegetation indices

and rice production)

(Gilbert et al., 2004; Pfeiffer et al., 2007)

11 In this table, based on my literature reading and my own judgment of priority, I subjectively grouped the risk

factors of transmission of HPAI (reported by other studies) by categories in order to show their possible connection

to small scale integrated farming systems.

of 340

Figure 2.1. Chapter Two’s Outline

Chapter 2:

Presents a developed socioeconomic profile of SSI farmers in Thai Binh and An Giang Provinces of Vietnam,

including:

Used to assist modeling factors associated with on-farm water quality (Chapter 4) and factors associated

with SSI farmers’ engagement in mitigating strategies to reduce transmission of WRZDs (Chapter 5) in

the provinces of Thai Binh and An Giang in Vietnam


What is the socioeconomic profile of SSI farmers in Thai Binh

and An Giang provinces of Vietnam?

Discussion of SSI

farming and EcoHealth

Farm economics (i.e., the

SSI farmers’ income and

contribution to farm

income of agricultural

production)

Livestock production

and fish production on

the SSI farms

Demographic profile of

the SSI farmers


Is the socioeconomic profile of SSI farmers in Thai Binh and An

Giang consistent with an EcoHealth approach in the context of

SSI farming?

of 340

2.2. Methods

2.2.1. Research design

2.2.1.1.Methods and phases of the research

In my research, quantitative and qualitative methods were used in a cross-sectional

research design. The following three phases were used to conduct the research: 1) establishing,

testing, piloting, and revising the questionnaires and survey protocol; 2) gathering data from the

target groups of farmers through the full field survey; and 3) analyzing the collected data.

Quantitative and qualitative data were collected at the same time. The quantitative data

were collected using questionnaires and objective assessments of microbial and related indicators

of water quality. Quantitative methods were used to: confirm my choice of hypothesis for testing

in my research; clean and check data; and summarize and analyse data12. Key qualitative method

steps included: identifying the target population (i.e., the eligible farmers in the research areas);

designing and testing the survey instruments; and defining and identifying variables that cross-

reference the research hypothesis. The following qualitative method tools were used to explore

culturally and meaningfully farmers’ perceptions and characteristics: in-depth interviews, focus

group discussions (FGD), meetings, field visit notes, observations, desk review, and audio-visual

materials. Qualitative data were used to assist the discussion of research results.

In the first phase of my research, I designed tools for the study based on my knowledge

of the research issues, the conceptual and theoretical models of the research, feedback from local

partners, and a review of the literature. These study tools included questionnaires and qualitative

data collection tools (i.e., guidelines for in-depth interviews, observations, and FGD). The study

tools were tested and used in a pilot study before being revised for use in a full field survey. In

12 STATA version 13 was employed for quantitative analysis.

of 340

the pilot study, I tested the questionnaires and qualitative data collection tools by asking the

enumerators of the field survey to read all the questions and to share their understanding of the

meaning and clarity of the questions. Enumerators were also asked to role-play in pairs

specifically to ask and answer questions contained in the questionnaires, and to review the

guidelines in the qualitative data collection tools and then provide me feedback on the

appropriateness of the questions and procedures for the field survey. Following this, I revised the

questionnaire and the qualitative data collection tools considering the feedback I received from

the enumerators. The enumerators then administered the questionnaire to 30 farmers in Thai

Thuy district of Thai Binh province. The questionnaire was revised once again before I used it in

the second phase (i.e., the full field survey). Details of the pilot study is in section 2.2.2.

In the second phase of my research, I conducted the full survey in the two provinces to

gather data from 600 eligible rural households (i.e., farms), which were randomly selected to

participate in my research. The selection of these rural households is described in the sampling

design section below. Key variables included demographic and socioeconomic characteristics,

livestock production focusing on poultry production, and other key types of on-farm production.

Local partners including commune and village leaders, as well as commune and village

health/veterinary workers played an important role in the field data collection. In the third phase

of my research, the collected data were cleaned, coded, and organized before analyzing.

2.2.1.2.Research areas

The research team’s previous evaluation visit to the Red River and Mekong River Deltas

provided them with a background understanding of the related problems and the situation of SSI

farmers in the two river deltas including the provinces of Thai Binh and An Giang. This

background understanding led to selecting the provinces of Thai Binh and An Giang as the two

of 340

target research areas for my research (See Figures 2.2). More importantly, these two provinces

are located in the centre of the Mekong River Delta and the Red River Delta where the HPAI

epidemic waves were focused (Hoang, 2006; Rushton et al., 2006).

There are 1.6 million and 1.5 million people respectively living in the provinces of Thai

Binh and An Giang. In 2010, the total number of farms/livestock farms were 3,376/2,388 (i.e., in

Thai Binh) and 17,273/218 (i.e., in An Giang) (GSO, 2012), which produced (as of October 1st,

2010) 8,899 (i.e., in Thai Binh) and 4,067 (i.e., in An Giang) heads of poultry (GSO, 2010). The

total area of surface water for aquaculture is 13.4 (i.e., in Thai Binh) and 1.9 (i.e., in An Giang)

thousand hectares (MARD, 2012). However, only 27.7% and 47.5% of the total numbers of

communes in Thai Binh and An Giang respectively have a common sewage system (GSO, 2010;

Nguyen et al., 2010b)

2.1.2.3.Selection criteria

I used four criteria to select participants/rural household residents from the research

areas. The first criterion was rural household residents who produced backyard poultry and

expected to do so for at least the next six months. The second criterion was rural household

residents who had a poultry flock size of about 50 birds (Burgos et al., 2007). The third criterion

was rural household residents who had access to at least one of the main water sources for rural

farmers (e.g., pond, river, well, or canal). The last criterion was rural household residents who

were interested in and willing to consent to participate in the research study. There were no

exclusion criteria.

of 340

Figure 2.2. Research Areas – The Provinces of Thai Binh and An Giang in Vietnam13

13 This figure uses maps produced by FAO and D. Pfeiffer including: (1) Map of poultry density in Vietnam and (2) Map of irrigated and drained agricultural

land (FAO, 2015); and (3) Map of HPAI epidemic waves 2004 (Pfeiffer, 2015). Greater numbers are indicated by darker colours.

of 340

2.2.1.4. Sampling design:

The choice of using both qualitative and quantitative methods together with insufficient

available information relating to effect size and variability challenged my ability to estimate the

sample size for my research. The sample size for the full field survey followed a

recommendation in which a minimum sample size pertaining to the most common qualitative

and quantitative and correlation designs would be 82 participants per group for a two tailed

hypothesis with 0.80 power at the 5% level of significance (Denise, 2010). Moreover, within a

5% to 10% level of significance assuming the large and medium conventional estimated

population correlations of 0.10 and 0.30, a sample size of 785 and 85 participants respectively

would be sufficient to keep the power of 0.80 and risks of statistical errors to standard levels

(Browner et al., 2013). Given this recommendation and the available resources for my research, a

sample size of 600 participants was feasible and acceptable.

I first conducted a pilot study among 30 farmers in which the questionnaires, the

qualitative data collection tools, and the survey procedures were tested, used in the field, and

then revised. The research team was already familiar with the target provinces through being

previously involved in health and agricultural research in these provinces. One district per

province14 was selected. This was based on a deliberate process in which the two districts chosen

had the closest matching population and number of communes to the mean population and

number of communes of all the districts in their respective province. The communes in these

districts reflected: 1) farming villages in Vietnam with a diverse ecosystem (e.g., ponds and

canals) that farmers were using for their poultry production; 2) farming systems engaging in

14 Vietnam is divided into 64 provinces. Each province is sub-divided into districts, then communes, and villages.

of 340

various types of SSI farming and other agricultural activities; and 3) typical risk factors for

WRZD related to livestock production. Within the provinces of Thai Binh and An Giang, the

districts selected were Quynh Phu and Chau Thanh respectively. The desired sample size was

300 households per district.

A two-stage cluster sampling method was used to select target communes as the clusters,

and households in each cluster as the research participants. In the first stage, the existing number

of communes in each district was identified as the existing sampling frame. The number of

clusters selected in each district was defined by dividing the sample size by the estimated mean

number of population elements in each cluster. The clusters were selected based on the sampling

interval (S) obtained by dividing the total population by the number of clusters, and on a random

number (N) using a random number table generated by Microsoft Excel (Microsoft, 2014). In the

second stage, the technique of selecting a random sample from a table of random numbers was

used to select households to participate in the research. Commune leaders and the communal

farmers’ associations actively assisted the research team by providing the list of farmers for the

random selection of the sample.

2.2.1.5.Variables

For the research in this chapter, variables were collected in the following areas of

interest: 1) personal characteristics (e.g., age, education, and other demographic characteristics);

2) household socioeconomic characteristics (e.g., income, poverty status, gender, and family

size); and 3) on-farm and livestock production characteristics (e.g., type of livestock, size of

flock, cropping pattern, and water sources).

Data collection methods included the use of questionnaires, in-depth interviews, focus

group discussion, observations, and local health and veterinary reports. In particular, structured

of 340

questionnaires were designed for interviewing the randomly selected farmers whereas semi-

structured questions and guidelines were designed for conducting the in-depth interviews, the

focus group discussions, and the observations of the participating farmers. Questionnaires

contained both open-ended and closed-ended questions. For the close-ended questions, farmers

had the option either of selecting a scaled response or of refusing to answer the questions.

2.2.2. Pilot study

From July 2012 to June 2013, a pilot study titled “Animals, water, and public health in

Vietnam” was conducted on 30 small-scale farms in the Thai Thuy district of Thai Binh province

in Vietnam. The Canada Excellence Research Chairs (CERC) program funded the pilot study

and the University of Prince Edward Island and the University of Calgary jointly led the one-

year study. The objectives of the pilot study were to examine the types of fish and livestock

raised on SSI farms in Vietnam and to test the study tools (i.e., the questionnaires, the qualitative

data collection tools, and the survey’s procedures). Testing and revising the study tools ensured

that the farmers easily understood the questions, and that the questions were free of ambiguity

while ensuring the validity and reproducibility of the responses. Farmers who participated in the

pilot study were not selected again for the full research study.

2.2.3. Data collection for the full research study

For both the pilot study and the full study, data collection followed the standard

ethical/consent procedures of the University of Calgary and the local research ethics board at the

Hanoi School of Public Health (see Appendix D). The field investigator recruited and trained the

enumerators on how to conduct the survey before they collected data in the field. Commune

collaborators/local guides assisted the field investigator and the enumerators throughout the field

survey. The research team worked with their local partners to identify commune collaborators

of 340

who then informed the selected households of the nature of the survey and requested members of

each household to discuss and nominate a representative to participate in the survey.

Based upon their knowledge, the household representatives were asked to provide their

birthday to determine the ages of farmers. Those who did not remember their birthdays were

asked to check their birth certificate and/or check it with other household members. The lists of

farmers provided by the commune collaborators also noted farmers’ birthdays for cross-

reference. Prior to the interview, consent forms were explained carefully to the selected farmers,

and only farmers who understood and agreed with the consent scripts were interviewed.

Informed consent was also attained for the focus group discussions, in-depth interviews, and

observations. At the end of each interview with a farmer, a bar of soap and a hand towel were

provided to the farmer as tokens of appreciation. No farmers knew about the tokens of

appreciation until after the completion of the interviews. The questionnaires were pretested and

integrated into five mini iPads using licensed Survey Pocket15 and QuestionsPro16 software. The

enumerators administered the questionnaires during in-person interviews with farmers. In each

province, a senior local researcher and I supported the enumerators throughout the data

collection (e.g., accompanying enumerators, observing enumerators in actions, answering

questions if any, having a briefing session with all enumerators before each day of data

collection and a reviewing session at the end of each day of data collection).

In the procedures for collecting qualitative information, the research team purposefully

selected farmers and other key stakeholders (informants) for the focus group discussions and in-

15 SurveyPocket is a survey application to conduct field surveys on iPad, iPhone or Android tablet and smart phone

devices (http://www.surveypocket.com/ ). 16 QuestionPro is a web-based software for creating and distributing surveys

(http://www.questionpro.com/home/howItWorks.html ).

http://www.surveypocket.com/

http://www.questionpro.com/home/howItWorks.html

of 340

depth interviews. Local authorities were asked to share relevant documents for review, if they

were available. Each focus group discussion included from five to ten farmers and took

approximately one hour to complete. The enumerators also facilitated the focus group

discussions by using the predesigned guidelines and guided questions. For the in-depth

interviews, each enumerator asked open-ended questions face-to-face with each farmer who was

representing a household. Heads of households and/or the person with responsibility for

livestock production, water, and related issues in the households were the primary respondents.

These procedures provided opportunities for participants to express qualitative information such

as questions, issues, comments, and inputs, which may not been sufficiently addressed in the

questionnaires.

Both the field investigator and the enumerators were experienced in conducting

interviews and administering questionnaires and they were familiar with the local agriculture

context. The enumerators also not only helped the field investigator in administering the

questionnaires, but in recording the interviews, and probing for answers, comments, and

expressions of the responding farmers. Audio records were then analyzed and summarized.

Farmers were also informed in advance that codes would be used in place of names in the

research to ensure confidentiality and to prevent bias.

Instructions and checklists were used to guide the interviews and the focus group

discussions to ensure consistency and to avoid missing important information. With respect to

the secondary data relating to livestock production, environment, and farmers’ health, the field

investigator asked local authorities to share information. However, the obtained secondary

information was limited as it was not always available, nor comprehensive, and sometimes it was

too general. Information obtained from the in-depth interviews, the focus group discussions, and

of 340

the literature review was used in discussing the results. This information primarily relates to

livestock production, water use, and farmers’ health in the studied areas.

2.2.4. Analysis of the full field survey’s data

The raw data collected from all the farmers who met the selection criteria and fully

participated in the field research were entered into a dataset, cleaned, coded, and edited.

Descriptive statistics were used to inspect frequency distributions (Long and Freese, 2014) and

central tendency (continuous variables) relating to the socioeconomic characteristics of the

participating farmers and their livestock production. Statistical tests were used to test significant

differences of research results between two provinces (e.g., Pearson Chi-squared test, t-test, and

Fisher's exact test).

As the primary researcher, I visually inspected the dataset using graphs and looked for

outliers. Fifteen farmers had young fish production numbers that appeared to deviate markedly

from those of other farmers and might reflect some abnormality or errors in the measurement or

recording. Therefore, I examined the variable for young fish production further and found that 10

farmers in the province of An Giang (i.e., 10 observations in the dataset) had young fish

production (i.e., 1.3 billion young fish per year) that lay outside the normal range of production

compared to the other participating farmers. In addition, the mean years of attending school and

the mean income/person/year were much greater than the rest of the SSI farmers (i.e., 9.4 years

of schooling, and > 20 million Vietnamese Dong (VND)17 respectively compared to 5,769 young

fish per year, 6.9 years of schooling, and < 20 million VND respectively). I also calculated the

interquartile range (IQR) as well as the outer fences (i.e., the lower outer limit and the upper

17 1 CAD = 17,206.42 VND as of Oct. 12, 2015.

of 340

outer limit) of the data on young fish production. The numbers of young fish produced in these

10 farms were more than three times the IQR above the 75th percentile (Q3) of the number of

young fish produced in the farms that produced fish. Furthermore, the numbers of young fish

produced on the 10 farms were much greater than the national mean number of fingerlings

produced per farm, per year, which was 16,472,868 (Ngoc, 2013).

Given the above calculations, I considered the numbers of young fish produced in these

10 farms as extreme outliers, and that including these numbers in analyzing young fish

production may bias the dataset. Therefore, I dropped these young fish production numbers (but

not the 10 farms) in describing young fish production in the participating farms. Furthermore, I

examined the profile of the 10 farms with extreme numbers of young fish production further by

stratifying the data over various income and education levels. This stratification was done to

improve the accuracy (i.e., valid representation) of the strata for the number of young fish

produced among the stratified participating farmers. I also used frameworks for SSI farming

(Dixon et al., 2004; Nguyen, 2014) in further examining the profile of these 10 farms concerning

the numbers of adult fish produced, volume of fish production in kg, level of education, on-farm

income, and number of livestock. I found that two of the 10 farms with extreme outliers in young

fish production were definitely not SSI farms. Therefore, I decided to remove these two farms

entirely from the dataset. Qualitative information gathered from the focus group discussions, in-

depth interviews, and secondary data collections were used to inform with the discussion and

interpretation of the results of the quantitative analysis. The focus of this dissertation was not on

reporting qualitative results analyzed using Nvivo.

of 340

2.3. Results

2.3.1. Demographics

2.3.1.1.Age, years of farming, years of attending school, and household size

The mean ages of participating farmers in Thai Binh and An Giang were 47.5 years and

44.4 years respectively (range 17-85) (Table 2.2). Participating farmers had a wide range of

years of integrated farming experience (i.e., a range of 1-40 years and a range 1-50 years in An

Giang and Thai Binh respectively). At the time of the study, farmers in An Giang appeared to

spend less years working as SSI farmers than those in Thai Binh did (i.e., 8 years and 12 years

respectively). Participating farmers spent nearly six years and eight years in school in An Giang

and Thai Binh respectively. The mean family size was four people in both An Giang and Thai

Binh. The mean number of children (under 18 years old) was about one.

Table 2.2. Demographics of Small-Scale Integrated (SSI) Farm Study Participants in the

Provinces of Thai Binh and An Giang, Vietnam

Variable Obs Mean Std. Dev. Min Max

Thai Binh

Age in years 299 47.5*** 11.9 17 85

Years of schooling 300 8.2*** 2.7 2 18

Household size 300 4.2*** 1.3 2 9

No. of people <18 yrs

old 300 1.0*** 0.9 0 3

Years of farming 300 11.7*** 8.7 1 50

An Giang

Age in years 298 44.4 10.5 19 76

Years of schooling 298 5.6 3.2 0 16

Household size 298 4.7 1.3 2 10

No. of people <18 yrs

old 298 1.4 1.0 0 5

Years of farming 298 7.6 7.2 1 40

(***) indicates a statistical significant difference between Thai Binh and An Giang p < 0.01

of 340

Table 2.3 shows that most farmers in both provinces attended some years in school up to

high school grades18. However, only 54/300 and 37/298 of farmers in Thai Binh and An Giang

respectively attended 12 years of school (high school level). Most of the participating farmers in

Thai Binh spent some years attending secondary school while most farmers in An Giang

attended primary school only from grade one to grade six. Participating farmers in An Giang had

mainly attended primary school while those in Thai Binh had mostly attended up to the

secondary school level. Sixteen farmers in Thai Binh (5% of the participating farmers in Thai

Binh) had also attended post-secondary school. This number was much greater than the number

of farmers who attended post-secondary school in An Giang (i.e., one or 0.3% of the

participating farmers in An Giang).

Table 2.3. Types of School Attended by Small-Scale Integrated (SSI) Farmers in the


Types of school

attended*

Provinces

Thai Binh An Giang Total

Number Percent Number Percent Number Percent

No school 0 0.0 7 2.3 7 1.2

Primary school 56 18.7 200 67.1 256 42.8

Secondary

school 174 58.0 53 17.8 227 38.0

High school 54 18.0 37 12.4 91 15.2

Post-high school 16 5.3 1 0.3 17 2.8

Total 300 100.0 298 100.0 598 100.0

(*) significant difference between Thai Binh and An Giang with p = 0.000, two-tailed Fisher's exact test

2.3.1.2.Gender and gender role

Most farmers who were nominated by their family members to participate in the survey

were men (206/300 or 68.7% and 222/298 or 74.5% in Thai Binh and An Giang respectively)

18 In Vietnam, a primary school includes grade one to grade five classes, a secondary school includes grade six to

grade nine classes, and a high school includes grade ten to grade twelve classes

of 340

and most of them were heads of households (195/300 or 65.0% and 208/298 or 69.1% in Thai

Binh and An Giang respectively (Table 2.4). However, it was common for both men and women

to share responsibility for family health (159/299 or 53.2% and 197/292 or 67.5% of households

in Thai Binh and An Giang respectively) (Table 2.5). In Thai Binh, male farmers were mainly

responsible for livestock production, fish production, and family health in 144/295 (38.6%),

134/287 (46.7%), and 67/299 (22.4%) of households respectively. Meanwhile, female farmers

were mainly responsible for livestock production, fish production, and family health in 54/295

(18.3%), 29/287 (10.1%), and 73/299 (24.4%) respectively. In An Giang, male farmers were

mainly responsible for livestock production, fish production, and family health in 130/294

(44.2%), 53/186 (28.5%), and 64/292 (21.9%) of households respectively. Meanwhile, female

farmers were mainly responsible for livestock production, fish production, and family health in

37/294 (12.6%), 12/186 (6.5%), and 31/292 (10.6%) respectively (Table 2.5).

of 340

Table 2.4. Gender of the Participating Farmers in the Provinces of Thai Binh and An

Giang in Vietnam and Their Relationship with Other Household Members

Province/

Gender Relationship with other household members

Head of household a Other roles b Total

Thai Binh Number Percent Number Percent Number Percent

Male 195 65.0 11 3.7 206 68.7

Female 24 8.0 70 23.3 94 31.3

An Giang

Male 206 69.1 16 5.4 222 74.5

Female 16 5.4 60 20.1 76 25.5

Total

Male 401 67.0 27 5.0 428 72.0

Female 40 6.0 130 22.0 170 28.0 (a) no significant difference between Thai Binh and An Giang with p = 0.677, Pearson Chi-squared test

(b) no significant difference between Thai Binh and An Giang with p = 0.690, Pearson Chi-squared test

Table 2.5. Gender Roles in Small-Scale Integrated (SSI) Farm Households in the Provinces

of Thai Binh and An Giang in Vietnam

Gender roles

Provinces



Mainly responsible for livestock production a

Male 114 38.6 130 44.2 244 41.4

Female 54 18.3 37 12.6 91 15.4

Both 127 43.1 127 43.2 254 43.1

Total 295 100.0 294 100.0 589 100.0

Mainly responsible for fish production

Male 134** 46.7 53 28.5 187 39.5

Female 29** 10.1 12 6.5 41 8.7

Both 124** 43.2 74 39.8 198 41.9

Not sure 0** 0.0 47 25.3 47 9.9

Total 287 100.0 186 100.0 473 100.0

Mainly responsible for family health

Male 67** 22.4 64 21.9 131 22.2

Female 73** 24.4 31 10.6 104 17.6

Both 159** 53.2 197 67.5 356 60.2

Total 299 100.0 292 100.0 591 100.0 (a) no significant difference between Thai Binh and An Giang with p = 0.121, Pearson Chi-squared test

(**) significant difference between Thai Binh and An Giang with p = 0.000, Pearson Chi-squared test

of 340

2.3.2. On-farm production

2.3.2.1.Fish production

In Thai Binh, SSI farmers with either secondary school or post-high school or incomes of

10 to 40 million VND per person per year produced the greatest mean number of young fish per

year. In An Giang, SSI farmers with either primary school or secondary school and incomes of

10 to 40 million VND per person per year produced the greatest mean number of young fish per

year (Table 2.6).

Table 2.6. Mean Quantity of Young Fish Produced Per Year on Small-Scale Integrated

(SSI) Farms in the Provinces of Thai Binh and An Giang, Vietnam, Stratified for

Education and Income

Provinces/

Income b

Highest level of educational attainment a

1 2 3 4 5

Thai Binh

< 4.8 - 3,800.0 2,500.0 - -

- (5,377.7) (4,203.6) - -

4.8-10 - 2,469.2 6,103.6 933.6 520.0

- (2,593.1) (17,732.2) (896.5) (678.8)

10-20 - 9,555.6 5,146.1 3,472.7 14,190.0

(12,875.5) (17,010.5) (3,733.4) (21,759.3)

20-40 - 1,000.0 10,000.0 1,000.0 4,000.0

- n.ac (11,742.0) n.a n.a

>40 - 10,000.0 1,508.3 800.0 200.0

- n.a (985.1) (244.9) n.a

An Giang

< 4.8 3 1,307.4 75.0 129.5 -

- (2,001.0) (35.4) (247.0) -

4.8-10 - 1,357.6 3,100.0 1,850.0 -

- (1,544.8) (3,489.3) (2,333.4)

10-20 - 18,006.1 6,000.0 7,006.7 -

- (27,239.7) n.a (11,263.2) -

20-40 - 14,000.0 100,000.0 100,000.0 -

- (8,485.3) n.a n.a -

>40 - - - - - (.) Numbers in brackets are standard deviations; a 1 = no school; 2 = primary school; 3 = secondary school; 4 = high

school; and 5 = post-high school; b Income was reported in millions of Vietnamese dong (VND) c No s.d. is reported where only one observation exists

of 340

2.3.2.2.Livestock production

Table 2.7 shows that SSI farms in Thai Binh raised more chickens, ducks, and pigs than

SSI farms in An Giang. However, the mean number of cattle raised on a SSI farm in An Giang

was about three times greater than in Thai Binh.

2.3.2.3.On-farm production other than fish and livestock production

In addition to livestock production and fish production, participating farmers in An Giang

and Thai Binh also had other on-farm production of which rice and vegetable production were

the two most common types (Table 2.8). However, very few farmers were producing cassava,

bamboo, or peanuts.

Table 2.7. Annual Livestock Production on Small-Scale Integrated Farms (SSI) in the Thai

Binh and An Giang Provinces in Vietnam

Province/Livestock Annual on-farm livestock production by province

No. of farms Percent of farms Mean sd Max Min

Thai Binh

Number of pigs 193 65 24.8*** 18.9 70 1

Number of ducks 126 42 57.1*** 56.9 200 1

Number of chickens 272 91 33.5*** 22.4 100 1

Number of cattle 65 22 1.4*** 1.0 5 1

An Giang

Number of pigs 101 34 7.7 9.7 50 1

Number of ducks 119 40 14.1 18.8 100 1

Number of chickens 201 67 11.9 10.7 52 1

Number of cattle 128 43 3.9 2.1 10 1 (***) indicates a statistical significant difference between Thai Binh and An Giang p < 0.01, t-test

2.3.2.4.Occupation other than farming among participating farmers

Participating farmers were asked to indicate any other work or employment they had in

addition to working as integrated small-scale livestock farmers. Table 2.9 shows that in both

provinces, almost half of the participating farmers had no additional work. Among those who

had additional jobs, the most common in Thai Binh and An Giang were owning a small

of 340

business/being a vendor, working communal committees19, or doing seasonal work respectively.

The most commonly reported seasonal work included temporary work relating to construction,

brick production, mechanics, motorbike taxi, rice processing, and on-call locally hired workers.

Table 2.8. On-farm Production Other Than Fish and Livestock on Small-Scale Integrated

(SSI) Farms in the Thai Binh and An Giang Provinces in Vietnam

Province Numbers and percent of farms by commodity and province

Rice Vegetables Fruits Corn Peanuts Cassava Bamboo

Thai Binh a

Number 234.0 216.0 195.0 26.0 22.0 5.0 7.0

Percent 78.0 72.0 65.0 8.7 7.3 1.7 2.3

An Giang b

Number 207.0 77.0 46.0 29.0 6.0 23.0 15.0

Percent 69.5 26.8 15.4 9.7 2.0 7.7 5.0

Both provinces c

Number 441.0 293.0 241.0 55.0 28.0 28.0 22.0

Percent 73.7 49.0 40.3 9.2 4.7 4.7 3.7 a n = 300 (in the province of Thai Binh) b n = 298 (in the province of An Giang) c n = 598 (in the provinces of Thai Binh and An Giang)

19 This may include a position at a commune health centre, commune veterinary station, commune farmers’ union,

or commune women’s union

of 340

Table 2.9. Occupations Other Than Farming Among Small-Scale Integrated (SSI) Farmers

in the Provinces of Thai Binh and An Giang, Vietnam

Occupation other

than farming

Numbers of farmers by occupation and province



No additional work 163 55.1 137 46.1 300 50.6

Communal staff 20 6.8 7 2.4 27 4.6

Small

business/vendor 60 20.3 39 13.1 99

16.7

Small production 29 9.8 7 2.4 36 6.1

Seasonal work 19 6.4 107 36.0 126 21.2

Retired 5 1.7 0 0.0 5 0.8

Total 296 100.0 297 100.0 593 100.0

2.3.3. Farm economics

2.3.3.1.Income of SSI farmers

Most SSI farmers reported their income from farming (expressed in VND per person per

year) as being between 4.8 million VND (equivalent to $244 CAD)20 to 20 million VND

(equivalent to $1,015 CAD) (Table 2.10). A limited number of farmers had income per year that

was higher than 40 million VND (equivalent to $2,030 CAD) – specifically 18 farmers (6.1%)

and six farmers (2.0%) in Thai Binh and An Giang respectively. Eleven (4%) and 118 (36%) of

participating farmers in Thai Binh and An Giang respectively had annual incomes that were

lower than the cut-off income level21 for poverty set by the Vietnamese government.

20 At the exchange rate of $1CAD = 19,700 VND on 5th December 2013 (Source: Vietinbank) 21 Less than 4.8 million VND per person per year – equivalent to $244 CAD per person per year

of 340

Table 2.10. Income of Small-Scale Integrated (SSI) Farmers in the Provinces of Thai Binh

and An Giang, Vietnam

On-farm income

per person per

year (VND

million)

Number and% of farmers by income and province



< 4.8 11*** 3.7 107 36.1 118 20.0

4.8-10 106*** 36.1 104 35.1 210 35.6

10-20 127*** 43.2 61 20.6 188 31.9

20-40 32*** 10.9 18 6.1 50 8.5

> 40 18*** 6.1 6 2.0 24 4.1

Total 294 100.0 296 100.0 590 100.0 (***) significant difference between Thai Binh and An Giang with p = 0.000, Pearson Chi-squared test

2.3.3.2.Contribution to total farm income from various kinds of agriculture production

Participating farmers were also asked to rate the level of income that they obtain from

their various on-farm production activities. Table 2.11a-b shows that although poultry, fish,

duck, pig, and rice production brought more income to SSI farmers than other vegetable

production, about 28% to 53% of farmers in An Giang and Thai Binh respectively indicated that

they receive either moderate or moderate to high farm income from poultry production. Yet,

approximately 5% of farmers in the province of Thai Binh considered fish, duck, and pig

production as the top three farm activities for providing “extremely high” contributions to on-

farm income compared to other forms of agricultural production. In the province of An Giang,

less than 2% of farmers considered fruit, fish, duck, and pig production as farm activities for

providing “extremely high” contributions to on-farm income compared to other forms of

agricultural production.

of 340

Table 2.11a. Farmers’ Perceptions of the Importance of Contribution to Total Farm

Income from Various Types of Agricultural Production on Small-Scale Integrated (SSI)

Farms in the Province of Thai Binh, Vietnam

Levela Contribution of on-farm production to farm income (%) by agricultural activity

Poultr

y

Fish Duck Pig Rice Cassav

a

Bamboo Peanut Cor

n

Vegetab

le

Fruit

1 2.2 c 5.4 4.6 4.2 - - 10.0 - - - -

2 22.3 57.9 26.2 60.6 6.7 - 10.0 - 21.4 9.1 5.8

3 53.1 23.0 57.9 28.6 53.1 - 20.0 40.0 50.0 50.0 54.7

4 15.0 10.8 6.7 6.1 34.3 - 50.0 50.0 28.6 16.8 21.6

5 7.3 2.9 4.6 0.5 5.9 - 10.0 10.0 - 24.0 17.9

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Nb 273 278 195 213 239 0 10 20 28 208 190

a Levels of contribution to income: 1 = extremely high contribution; 2 = high contribution; 3 = moderately high

contribution; 4 = low contribution; and 5 = no contribution; b N = number of respondents; c For example, of 273

respondents who raised poultry, 2.2% indicated that poultry production made an “extremely high” contribution to

farm income

Table 2.11b. Farmers’ Perceptions of the Importance of Contribution to Total Farm

Income from Various Types of Agricultural Production on Small-Scale Integrated (SSI)

Farms in the Province of An Giang, Vietnam

Levela

Contribution of on-farm production to farm income (%) by agricultural activity

Poultr

y

Fish Duck Pig Rice Cassav

a

Bamboo Peanut Cor

n

Vegetabl

e

Fruit

1 - - 0.4 0.4 1.9 - - - - - 0.5

2 3.9 c 12.5 2.4 10.3 14.1 6.6 - 2.0 8.1 4.1 -

3 28.5 18.1 17.9 17.3 38.9 5.2 0.5 3.9 4.3 4.6 1.4

4 28.5 16.0 16.3 10.7 15.9 - 6.2 1.0 1.4 16.8 16.0

5 39.1 53.4 63.1 61.3 29.3 88.2 93.3 93.3 86.1 74.6 82.2

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Nb 256 232 246 243 270 211 209 208 209 220 219

a Levels of contribution to income: 1 = extremely high contribution; 2 = high contribution; 3 = moderately high

contribution; 4 = low contribution; and 5 = no contribution; b N = number of respondents; c For example, of 256

respondents who raised poultry, 3.9% indicated that poultry production made an “high” contribution to farm income

of 340

2.4. Discussion

2.4.1. SSI farming practice and its profile in the provinces of Thai Binh and An Giang

Overall, SSI farming practice including the VAC was promoted as a model in rural areas

in Vietnam for a couple of decades after the introduction of the “Doi moi” reform policy in

198622. This is an important policy milestone with a shift from a centrally planned and

subsidized system to a market oriented system. This policy shift has led to the allocation of

production materials such as land to individual farmers instead of to collective ownership. The

VAC model was introduced as an integrated farm management system focusing on three core

components – the garden, the fishpond, and animal husbandry – with the aim of improving

farming practices and achieving high economic efficiency through optimizing the use of land and

water with low capital investment and inputs (Le, 2001; Pham, 2010). In some other Asian

countries, several integrated farming practices that are similar to the VAC model also exist such

as integrated fish-horticulture farming (India), a culture of short-cycle species in seasonal ponds

and ditches (Bangladesh), fodder-fish integration practices (Malaysia), and embankment fish

culture and integrated grass-fish farming (China) (Pullin, 2001).

My research in this chapter shows that there were differences between ages of the farmers

from the provinces of Thai Binh and An Giang. The participating farmers from An Giang were

younger and had fewer years of schooling compared to the participating farmers in Thai Binh.

Rural farmers are still practicing the small-scale integration of various on-farm production

activities with livestock, fish, crop, and so on, which is similar to the reported structure of SSI

farming in Vietnam (Devendra and Thomas, 2002). However, the participating farmers in this

22 In 1896, the Vietnamese government introduced “Doi moi” policy, which indicates a shift the Vietnamese

economy from a centrally planned economy to a market oriented economy (Dang, 2015; Le, 2013).

of 340

study preferred to refer to their farming practice as a normal/traditional way that most farmers

practice farming instead of as an SSI model. This may be explained by the fact that SSI farming

has been introduced in rural areas of Vietnam over the last couple of decades (Hop, 2003).

Farmers with secondary school education and greater income appeared to produce the greatest

mean number of young fish per year. Concerning fish production in An Giang, a number of

farmers had recently stopped producing fish on farms and the qualitative study revealed three

main reasons for this. First, producing fish required greater investment yet yielded low profits.

Second, the restructuring of the irrigation system limited the volume and quality of water input

for ponds. Third, farmers felt it was more convenient to catch fish from canals or rivers or to buy

fish for food than to produce fish on the farm. The insight from this is that programs and policies

targeting livestock production among SSI farms should address different kinds of livestock and

place an emphasis on poultry production.

A majority of reported risk factors for HPAI are related to SSI farming (See Table 2.1)

yet future food availability is projected to depend on improving the productivity for livestock

production in SSI farming systems (CGIAR, 2011; Thomas et al., 2002), with Vietnam being

among the top five countries in Asia to have a large number of small farms (Oksana, 2005).

Moreover, small-scale livestock production is the fastest growing sub-sector of agriculture in

Vietnam and will likely be a significant sub-sector for the next 20 years. Therefore, it is

important to review SSI farming using an EcoHealth approach to determine how to improve SSI

farming in the area of managing risk factors for emerging infectious diseases.

2.4.2. SSI farming and EcoHealth

With respect to complexity and trans-disciplinarity, SSI farming practice is truly a

complex environment where there is an interface between humans, animals, and the

of 340

environment. The integration of gardens, fishponds, and animal husbandry is an example of

complexity where animal manure is used as fertilizer and food for plants and fish. Livestock

production can also bring income and food for farmers, food for animals (e.g., using throwaway

parts of fish as food for animals), and wastewater for gardens. Products from gardens can

likewise be used as food for fish and animals, and bring income for farmers. Such examples of

integration can help optimize on-farm resources. However, this integration correspondingly has

potential linkages to increasing risk factors for transmitting infectious diseases from livestock

and fish to humans (Dang and Dalsgaard, 2012; Huong et al., 2014; Markwell and Shortridge,

1982; Nguyen et al., 2010a; Nguyen et al., 2010b). Moreover, complexity is also reflected in

geographical variation in SSI farming. For instance, the mean volume of fish production for

participating farmers in An Giang is significantly greater than in Thai Binh, and the mean

number of cattle raised in a household in An Giang was significantly greater than in Thai Binh.

Therefore, this complexity requires the knowledge and involvement of various disciplines such

as veterinary health, public health, water security, the private sector, and so on to not only

improve farming practices and generate income but also to enhance the health of farmers and

their animals and to help them avoid polluting their environment.

Regarding community participation, although farmers are the owners and implementers

of this SSI farming practice, their involvement in designing, implementing, and monitoring

policies related to livestock production, rural water supplies, and sanitation is limited (Le et al.,

2014). Yet, to optimize the limited water resources available for farmers and reduce the negative

socioeconomic and health impacts on the rural population, it is important to more effectively

enhance community participation in the policy making process not only with policy makers but

of 340

also with farmers themselves. Such actions could more explicitly help address the issues related

to SSI farming such as water, health, and small-scale livestock production.

In connection with gender equity, my research results show that male farmers played a

greater role in the SSI farms than female farmers did. More specifically, male farmers headed

most of the SSI farms. The percentages of farms with male farmers mainly responsible for

livestock production, fish production, and family health were greater than those percentages for

female farmers except for family health in Thai Binh. I think these results are interesting. I have

not found any other studies that explore the same gender equity issues among SSI farms (i.e.,

gender roles in livestock production, fish production, and family health) to compare with the

results of my research. However, some case studies of roles of women in agriculture reported

that women heavily participate in livestock production and the aquaculture sector (FAO, 2011;

Ranney et al., 2011). My research results focused on gender roles and main responsibilities for

livestock production, fish production, and family health. However, cautions may be needed in

generalizing the results as gender roles may change depending on types/scales of livestock

production on farms. Male farmers may have reported differently as they may feel uncomfortable

with female farmers being mainly responsible for livestock production, fish production, and

family health. It may also be cultural issues where men claim responsibility for women’s roles.

Furthermore, there may be differences between, for example, being responsible for livestock

production (e.g., making decisions) and doing tasks relating to livestock production (e.g., feeding

livestock). Therefore, future research may need to consider these differences/possible changes in

order to understand more fully gender roles among small-scale farming.

Concerning social equity among SSI farms, the application of the VAC model was

reported to generate a mean income from about 50 million VND to 70 million VND per year, per

of 340

household – equivalent to $2,530 CAD to $3,335 CAD per household, per year (Pham, 2010).

Among the study’s participating farmers, the mean number of people in a household was 4.4.

Most of the participating farmers earned 4.8 million VND to 10 million VND per person, per

year, which is equivalent to 21.12 million VND to 44 million VND per household, per year. This

level of combined household income – equivalent to $1,072 CAD to $2,233 CAD respectively –

is lower than the income level Pham (2010) reported from the VAC model. Further, about one in

five of the SSI farmers had income that fell below the existing poverty line set by the

government of Vietnam. Approximately, ten times more SSI farms were below the poverty

income level in An Giang than in Thai Binh. The mean family sizes of the participating

households in Thai Binh and An Giang were slightly larger than the 2009 mean family size in

Vietnam which was 3.8 (GSO, 2011).

Given my research results shows a decrease in mean income with a slightly larger mean

family size, it is understandable that farmers are looking for off-farm seasonal work to earn

additional income. However, with limited education and training, it is a challenge for farmers to

look for off-farm seasonal work. Moreover, the percentage of farmers who attended university in

the provinces of Thai Binh and An Giang (i.e., 5.3% and 0.3% respectively) was lower than the

percentage of those in the general population in the Red River Delta and the Mekong River Delta

who completed university – 6.3% and 2% respectively (GSO, 2011). In-depth interviews with

farmers showed that farmers who took on off-farm seasonal work were from the younger

generation and had to work far from home in poor conditions with limited social benefits.

Moreover, although their income from off-farm seasonal work was greater than that from

livestock production, their away-from-home living expenses were also high.

of 340

Pertaining to the idea of introducing the SSI farming model was the goal of enhancing the

sustainability of farmers’ livelihoods. Sustainability was introduced at the beginning of the

model in which the nature of integrating various on-farm production activities – from fish and

livestock to crop production – were for optimizing on-farm resources to produce greater

sustainability for farmers’ livelihoods (Hop, 2003).

However, from the standpoint of an EcoHealth approach in the context of SSI farming,

several key issues may influence the sustainability of both the SSI farming model itself and the

health of farmers and their animals. These issues include the following: the level of income from

integrated farming and the trend among farmers of doing off-farm work; the attention paid to the

health of farmers and their livestock; guidelines for managing health risk in the model; private

sector connections and markets; and protection of the environment, particularly the quality of on-

farm water while practicing integrated farming. In addition, SSI farmers are not equipped to face

challenges and competitiveness when Vietnam joins the World Trade Organization (Nguyen,

2015; Vu, 2016; Pham, 2008; Vo, 2005). Meanwhile, existing public policies have not been

updated to support farmers, which may worsen the undesirable consequences of SSI farming,

instead of improving them (Hall and Le, 2009; Le et al., 2014). Among the above issues, the

income of farmers was explored in the SSI model, as well as for the participating farms in my

research. However, other issues, such as those mentioned above, need to be studied further to

better understand potential implications for an EcoHealth approach in the context of SSI farming

2.5. Conclusions

The SSI farming model is a common agricultural practice in Vietnam and will continue to

be an important part of agriculture in Vietnam for the future. This model has proven to be

successful in helping farmers optimize their on-farm resources and generate food and income.

of 340

Reported risk factors for transmission of WRZD with respect to SSI farmers could lead to

unintended health consequences for not only farmers and their animals, but also for their

environment. The characteristics of SSI farmers described in this study can also be used as a

basis for future studies of the SSI farming model. The EcoHealth approach has been considered a

good model for community development (Finkelman et al., 2008). In Vietnam, there are some

valuable examples of applying the EcoHealth approach in academic research, policy advocacy,

and community development (Hall et al., 2012; Nguyen-Viet et al., 2013; Nguyen-Viet et al.,

2015). Nevertheless, when using an EcoHealth perspective and taking into consideration the

increase of WRZD, this model must be reviewed to ensure that it will not be a simple agricultural

economic model in which food and income are outputs. Rather, it should be a more holistic

EcoHealth-based model in which the health of farmers and the health of animals are interwoven

with on-farm production and the protection of the environment, and these are considered an

equally important output of the model.

This chapter presented results relating to the profile of farmers who practice the SSI

farming model in the provinces of Thai Binh and An Giang in Vietnam. However, further studies

are needed to understand the SSI model more comprehensively with respect to an EcoHealth

approach. Researchable issues relating to an SSI farming model may include risk factors for

transmission of WRZD among humans and animals; farmers’ perceptions of health-related

factors; the contributing factor of farmers’ mitigating strategies for WRZD transmission; on-farm

environmental health, especially water public health; and the contribution of improved SSI

farming methods to reduce the exposure to risk factors in WRZD transmission.

of 340

References

Biswas P.K., Christensen J.P., Ahmed S.S.U., Das A., Rahman M.H., Barua H. et al. (2009a).

Risk for infection with highly pathogenic avian influenza virus (H5N1) in backyard

chickens, Bangladesh. Emerg Infect Dis [serial on the Internet].

Biswas P.K., Christensen J.P., Ahmed S.S.U., Das A., Rahman M.H., Barua H. et al. (2009b).

Risk for infection with highly pathogenic avian influenza virus (H5N1) in backyard

chickens, Bangladesh. Emerg Infect Dis [serial on the Internet].



Bosma, R. H., Phong, L. T., Kaymak, U., Van Den Berg, J., Udo, H. M. J., van Mensvoort, M. E.

F. et al. (2006). Assessing and modelling farmers' decision-making on integrating

aquaculture into agriculture in the Mekong Delta. NJAS - Wageningen Journal of Life

Sciences, 53, 281-300.

Browner, W. S., Newman, T. B., and Hulley, S. B. (2013). Estimating Sample Size and Power:

Applications and Examples. In S.B.Hulley, S. R. Cumming, W. S. Browner, D. G. Grady,

and T. B. Newman (Eds.), Designing Clinical Research (Fourth Edition ed.,

Philadelphia: Wolters Kluwer Health and Lippincott Williams and Wilkins.

Bui, T. D., Hien, N. T., and Dang, T. T. (2010). Distribution of freshwater snails in family-based

VAC ponds and associated waterbodies with special reference to intermediate hosts of

fish-borne zoonotic trematodes in Nam Dinh Province, Vietnam. Acta Tropica, 116, 15-

23.


In Rome: FAO.

of 340

Burgos, S., Hong Hanh, P. T., Roland-Holst, D., and Burgos, S. A. (2007). Characterization of

Poultry Production Systems in Vietnam. International Journal of Poultry Science, 6, 709-

712.

Care (2011). Community - based Surveillance for Avian Influenza in Vietnam: Experiences,

Results, Discussions, and Recommendations Care International.


Publishing.

CGIAR (2011). Pigmeat value chain in Vietnam: Background proposals for the CGIAR

Research Program on Livestock and Fish International Livestock Research Institute

(ILRI).

Charron, D. F. (2012). Ecohealth Research in Practice: Innovative Applications of an Ecosystem





Institute.

Dang, H. V. (2015). A new approach to explain policy reforms in Vietnam during doi moi by

developing and validating a major policy change model for Vietnam (Order No.

3554150).

Dang, N. K., Milstein, A., Verdegem, M. C. J., and Verreth, J. A. V. (2006). Food inputs, water

quality and nutrient accumulation in integrated pond systems: A multivariate approach.

Aquaculture, 261, 160-173.

of 340

Dang, N. K., Phong, L. T., Verdegem, M. J. C., Duong, L. T., Bosma, R. H., and Little, D. C.

(2007). Integrated freshwater aquaculture, crop and livestock production in the Mekong

delta, Vietnam: Determinants and the role of the pond. Agricultural Systems, 94, 445-

458.

Dang, S. T. T. and Dalsgaard, A. (2012). Escherichia coli Contamination of Fish Raised in

Integrated Pig-Fish Aquaculture Systems in Vietnam. Journal of Food Protection, 75,

1317-1319.

Denise, F. P. (2010). Correlation and Simple Regression. In Maura Connor, Pamela Fuller, and

Patrick Walsh (Eds.), Statistics and Data Analysis for Nursing Research (Second Edition

ed., pp. 197-222). Julie Levin Alexander.

Desvaux, S., Nguyen, C. O., Vu, D. T., Henriquez, C., Ky, V. D., Roger, F. et al. (2014). Risk of

Introduction in Northern Vietnam of HPAI Viruses from China: Description, Patterns and

Drivers of Illegal Poultry Trade. Transboundary and Emerging Diseases.

Devendra, C. and Thomas, D. (2002). Cropanimal systems in Asia: importance of livestock and

characterisation of agro-ecological zones. Agricultural Systems, 71, 5-15.

Dinh, P. N., Long, H. T., Tien, N. T. K., Hien, N. T., Mai, L. T. Q., Phong, L. H. et al. (2006).

Risk factors for human infection with avian influenza A H5N1, Vietnam, 2004. Emerg

Infect Dis.

Dixon, J., Tanyeri-Abur, A., and Wattenbach, H. (2004). Framework for analyzing impacts of

globalization on smallholders. Food and Agriculture Organization of the United Nations

(FAO) [On-line]. Available: http://www.fao.org/docrep/007/y5784e/y5784e02.htm

Duong, D. V., Binns, C. W., and Lee, A. H. (2004). Utilization of delivery services at the

primary health care level in rural Vietnam. Social Science and Medicine, 59, 2585-2595.

http://www.fao.org/docrep/007/y5784e/y5784e02.htm

of 340

Fang, L. Q., de Vlas, S. J., Gong, P., Gong P, and et al. (2008). Environmental Factors

Contributing to the Spread of H5N1 Avian Influenza in Mainland China. PLoS ONE, 3.

FAO (2011). The State of Food and Agriculture: Women in Agriculture - closing the gender gap

for development. Food and Agriculture Organization (FAO) [On-line]. Available:

http://www.fao.org/docrep/013/i2050e/i2050e.pdf

FAO (2010). The fight against Highly Pathogenic Avian Influenza and other Emerging

Infectious Diseases. FAO Vietnam [On-line]. Available:

http://www.fao.org/docrep/015/an164e/an164e00.pdf

FAO (2015). Maps: Poultry Production and Avian Influenza in Vietnam. Food and Agriculture

Organization of the United Nations (FAO) [On-line]. Available:

http://www.fao.org/ag/againfo/programmes/en/pplpi/map_hpaivietnam.html

Fasina, F. O., Rivas, A. L., Bisschop, S. P. R., Stegeman, A. J., and Hernandez, J. A. (2011).

Identification of risk factors associated with highly pathogenic avian influenza H5N1

virus infection in poultry farms, in Nigeria during the epidemic of 20062007. Preventive

Veterinary Medicine, 98, 204-208.

Figuie, M. (2007). Future of Poultry Farmers in Vietnam after HPAI. FAO and MARD Workhop

[On-line]. Available: https://agritrop.cirad.fr/539844/1/document_539844.pdf

Finkelman, J., MacPherson, N., Silbergeld, E., and Zinsstag, J. (2008). External Review of the

IDRC Ecohealth Program Initiative: Final Report. International Development Research

Centre (IDRC) [On-line]. Available: https://idl-bnc.idrc.ca/dspace/handle/10625/37572

Geurden, T., Somers, R., Thanh, N. T. G., Vien, L. V., Nga, V. T., Giang, H. H. et al. (2008).

Parasitic infections in dairy cattle around Hanoi, northern Vietnam. Veterinary

Parasitology, 153, 384-388.

http://www.fao.org/docrep/013/i2050e/i2050e.pdf

http://www.fao.org/docrep/015/an164e/an164e00.pdf

http://www.fao.org/ag/againfo/programmes/en/pplpi/map_hpaivietnam.html

of 340

Gilbert, M., Chaitaweesub, P., Parakamawongsa, T., Premashthira, S., Tiensin, T., Kalpravidh,

W. et al. (2006). Free-grazing ducks and highly pathogenic avian influenza, Thailand.

Emerging Infectious Diseases, 12, 227-234.

Gilbert, M., Wint, W., and Slingenbergh, J. I. (2004). The ecology of Highly Pathogenic Avian

Influenza in East and Southeast Asia: outbreak distribution, risk factors and policy

implications. (Rep. No. 43). Food and Agriculture Organization.

Gleeson, L. J. and Dung, D. H. (2010). Mid-Term Review: Vietnam Integrated National

Operational Program for Avian and Human Influenza (OPI) 2006-2010: HPAI Control

and Eradication in the Agriculture Sector.

GSO (2010). Summary results of the Vietnam household living standard survey 2010 Vietnam

General Statistics Office (GSO).

GSO (2011). The 2009 Vietnam Population and Housing Census General Statistics Office,

Ministry of Planning and Investment Vietnam.

GSO (2012). Results of the 2011 Rural, Agricultrure and Fishery Census: Part 4 - Households

using aquacultural land by the size of land used and province Hanoi: Statistical

Publishing House - General Statistics Office.

Hall, C. D., Dinh, X. T., Vu, H. H., Tran, P., and Le, B. Q. (2012). Eco-health message on

dengue in Vietnam.

Hall, D. C. and Le, Q. B. (2009). The midterm evaluation of the Government of Vietnam-UN

Joint Programme to Fight Highly Pathogenic Avian Influenza 2007-2010: Strengthening

the Management of Public Health Emergencies in Vietnam – with a focus on the

Prevention and Control of Diseases of Epidemic Potential including Highly Pathogenic

of 340

Avian Influenza (HPAI). The Government of Vietnam-UN Joint Programme to Fight

Highly Pathogenic Avian Influenza.

Hall, D. C. (2010). The Ecohealth Approach: An introduction to the key concepts Building

Ecohealth Capacity in Asia (BECA).



Henning, J., Henning, K. A., Morton, J. M., Long, N. T., Ha, N. T., Vu, L. T. et al. (2011).

Highly pathogenic avian influenza (H5N1) in ducks and in-contact chickens in backyard

and smallholder commercial duck farms in Viet Nam. Preventive Veterinary Medicine,

101, 229-240.

Henning, J., Pfeiffer, D., and Vu, L. (2009). Risk factors and characteristics of H5N1 Highly

Pathogenic Avian Influenza (HPAI) post-vaccination outbreaks. Vet.Res., 40.

Hoang, T. D. (2006). HPAI Epidemic and lessons learned in Viet Nam. In APEC.

Hop, L. T. (2003). Programs to Improve Production and Consumption of Animal Source Foods

and Malnutrition in Vietnam. The Journal of Nutrition, 133, 4006S-4009S.

Hulse-Post, D. J., Sturm-Ramirez, K. M., Humberd, J., Seiler, P., Govorkova, E. A., Krauss, S. et

al. (2005). Role of domestic ducks in the propagation and biological evolution of highly

pathogenic H5N1 influenza viruses in Asia. Proc Natl Acad Sci USA, 102, 10682-10687.

Huong, L. Q., Madsen, H., Anh, L. X., Ngoc, P. T., and Dalsgaard, A. (2014). Hygienic aspects

of livestock manure management and biogas systems operated by small-scale pig farmers

in Vietnam. Science of the Total Environment, 470471, 53-57.



of 340




Kamakawa, A., Thu, H. T. V., and Yamada, S. (2006). Epidemiological survey of viral diseases

of pigs in the Mekong delta of Vietnam between 1999 and 2003. Veterinary







Ltd.

Knudsen, L. G., Phuc, P. D., Hiep, N. T., Samuelsen, H., Jensen, P. K., Dalsgaard, A. et al.

(2008). The fear of awful smell: Risk perceptions among farmers in Vietnam using

wastewate and human excreta in agriculture. S E Asia J Trop Med Pub Health, 39, 341-

352.

Le, B. Q., Hall, C. D., and Cork, S. C. (2014). Water related policies (WRPs) and implications

for reductions of risk factors of waterborne diseases among livestock farmers in Vietnam.

In Ecohealth 2014 Conference Montreal, Canada: International Association for Ecology

and Health.

Le, H. H. (2013). Vietnam's domestic-foreign policy nexus: Doi Moi, foreign policy reform, and

Sino-Vietnamese normalization. Asian Politics and Policy, 5, 387.

of 340

Le, L. T. (2001). Integrated Farming Systems: The VAC system in Northern Vietnam. Food and

Agriculture Organization of the United Nations (FAO) [On-line]. Available:

http://www.fao.org/docrep/005/Y1187E/y1187e00.htm#TopOfPage

Le, T. P., Udo, H. M. J., Mensvoort, M. E. F. v., Bosma, R. H., Le, Q., Dang, K. N. et al. (2007).

Integrated agriculture-aquaculture systems in the Mekong Delta, Vietnam: an analysis of

recent trends. Asian Journal of Agriculture and Development, 4, 51-66.

Long, J. S. and Freese, J. (2014). Regression Models for Categorical Dependent Variables Using

Stata. (Third ed.) Texas: Stata Press College Station.

MARD (2012). Provincial rural and agriculture database Vietnam: Planning Department,

Ministry of Agriculture and Rural Development.

MARD and MoH (2006). Vietnam Integrated National Plan for Avian Influenza Control and

Human Pandemic Influenza Preparedness and Response 2006-2008 Hanoi: Vietnam

Ministry of Health and Ministry of Agriculture and Rural Development.

Markwell, D. D. and Shortridge, K. F. (1982). Possible waterborne transmission and

maintenance of influenza viruses in domestic ducks. Applied and Environmental


Microsoft (2014). Random Number Generation. Microsoft [On-line]. Available:

http://support.microsoft.com/kb/86523

Minh, P. Q., Morris, R. S., Schauer, B., Stevenson, M., Benschop, J., Nam, H. V. et al. (2009).

Spatio-temporal epidemiology of highly pathogenic avian influenza outbreaks in the two

deltas of Vietnam during 2003-2007. Preventive Veterinary Medicine, 89, 16-24.


http://support.microsoft.com/kb/86523

of 340




MoH and MARD (2011). The Vietnam Integrated National Operational Program on Avian

Inlfuenza, Pandemic Preparedness and Emerging Infectious Diseases (APEID) 2010-

2015: Strengthening responses and improving prevention through a One Health

approach. Hanoi: Vietnam Ministry of Agriculture and Rural Development and Ministry

of Health.

Morley, C. (2010). Bio-Security Stakeholder Analysis, Vietnam (Rep. No. OSRO/RAS/604/USA

B6). The Emergency Centre for Transboundary Animal Diseases (ECTAD), Food and

Agriculture Organization of the United Nations (FAO).



UN.

Munster, V. J., Wallensten, A., Baas, C., Rimmelzwaan, G. F., Schutten, M., Olsen, B. r. et al.

(2005). Mallards and Highly Pathogenic Avian Influenza Ancestral Viruses, Northern

Europe. Emerging Infectious Diseases, 11, 1545-1551.

Ngoc, T. (2013). Results of the national fingerlings production during the first 6 months of 2013.

Vietnam Directorate of Fisheries [On-line]. Available: http://www.fistenet.gov.vn/e-nuoi-

trong-thuy-san/a-san-xuat-giong/ket-qua-san-xuat-giong-thuy-san-6-thang-111au-nam-

2013

http://www.fistenet.gov.vn/e-nuoi-trong-thuy-san/a-san-xuat-giong/ket-qua-san-xuat-giong-thuy-san-6-thang-111au-nam-2013



of 340

Nguyen, A. T. L., Madsen, H., Dalsgaard, A., Nguyen, P. T., Dao, T. T. H., and Murrell, K. D.

(2010a). Poultry as reservoir hosts for fishborne zoonotic trematodes in Vietnamese fish

farms. Veterinary Parasitology, 169, 391-394.

Nguyen, C. K., Pham, D. P., Tran, H. B., and Nguyen, H. V. (2010b). Health Risks from Excreta

and Wastewater to Vietnamese Farmers.

Nguyen, V. B. (2015). Agriculture restructuring: Towards higher global competitiveness and

food security. Vietnamese Academy of Agricultural Sciences [On-line]. Available:

http://ageconsearch.umn.edu/bitstream/225580/2/VanBo2014.pdf

Nguyen, V. D. (2014). Results of the Control and Prevention of Animal Diseases in early 2014

Thai Binh: Thai Binh Department of Veterinary.

Nguyen-Viet, H., Pham, G., and Tung, D. X. (2013). Field building leadership initiative:

advancing ecohealth in Southeast Asia - the first year journey. 10th Year Anniversary of

Veterinary Public Health Centre for Asia Pacific. Veterinary Public Health Centre for

Asia Pacific (VPHCAP), Chiang Mai University: 192-197. Conferecne proceedings.

Nguyen-Viet, H., Doria, S., Tung, D., Mallee, H., Wilcox, B., and Grace, D. (2015). Ecohealth

research in Southeast Asia: past, present and the way forward. Infectious Diseases of

Poverty, 4, 5.

Oksana, N. (2005). Small farms: current status and key trends. Future of Small Farms Research

Workshop, Wye College [On-line]. Available:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.146.4632andrep=rep1andtype=

pdf

http://ageconsearch.umn.edu/bitstream/225580/2/VanBo2014.pdf

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.146.4632&rep=rep1&type=pdf

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.146.4632&rep=rep1&type=pdf

of 340

Otte, J., Hinrichs, J., Rushton, J., Roland-Holst, D., and Zilberman, D. (2008). Impacts of avian

influenza virus on animal production in developing countries. CAB Reviews: Perspectives

in Agriculture, Veterinary Science, Nutrition and Natural Resources, 3.

Paul, M., Wongnarkpet, S., Gasqui, P., Poolkhet, C., Thongratsakul, S., Ducrot, C. et al. (2011).

Risk factors for highly pathogenic avian influenza (HPAI) H5N1 infection in backyard

chicken farms, Thailand. Acta Tropica, 118, 209-216.

Paul, M. C., Gilbert, M., Desvaux, S. p., Rasamoelina Andriamanivo, H., Peyre, M., Khong, N.

V. et al. (2014). Agro-Environmental Determinants of Avian Influenza Circulation: A

Multisite Study in Thailand, Vietnam and Madagascar. PLoS ONE, 9, e101958.

Pfeiffer, D. (2015). Epidemic Waves 2004-2005. Food and Agriculture Organization of the

United Nations (FAO) [On-line]. Available:

http://www.fao.org/ag/againfo/programmes/en/pplpi/docarc/hpai_05epidemicwaves.html

Pfeiffer, D. U., Minh, P. Q., Martin, V., Epprecht, M., and Otte, M. J. (2007). An analysis of the

spatial and temporal patterns of highly pathogenic avian influenza occurrence in Vietnam

using national surveillance data. Vet J, 174, 302-309.

Pham, D. P., Nguyen, V. H., Hattendorf, J., Zinsstag, J., Phung, D. C., and Odermatt, P. (2011).

Risk factors for Entamoeba histolytica infection in an agricultural community in Hanam

province, Vietnam. Parasites and Vectors, 4, 102.

Pham, V. K. (2008). Vietnamese agriculture: One year entering the WTO. Research Group

Discussion, Paper No.15, June 2008, School of Economics, The University

ofQueensland, Brisbane, Australia [On-line]. Available:

http://www.uq.edu.au/economics/eaerg/dp/EAERG_DP15.pdf

http://www.fao.org/ag/againfo/programmes/en/pplpi/docarc/hpai_05epidemicwaves.html

http://www.uq.edu.au/economics/eaerg/dp/EAERG_DP15.pdf

of 340

Pham, V. T. (2010). VAC integrated system with entire energy chain in Vietnam. Food and

Agriculture Organization of the United Nations (FAO) [On-line]. Available:

http://www.fao.org/bioenergy/26371-0b904ea1d132260451c3231064e3f440c.pdf

Phan, Q. M. (2010). Epidemiological Studies of Highly Pathogenic Avian Influenza in Vietnam.

Massey University.




ProMED (2014). Number of alerts in Vietnam in the past 12 months for influenza. ProMED [On-

line]. Available: http://healthmap.org/promed/?p=152

Pullin, R. (2001). Integrated agriculture-aquaculture and the environment. Food and Agriculture

Organization of the United Nations (FAO) [On-line]. Available:


Ranney, T., Enriquez, G., Cropenstedt, A., Gerosa, S., Lowder, S., Matuscke, I. et al. (2011). The

Role of Women in Agriculture. Food and Agriculture Organization (FAO) [On-line].

Available: http://www.fao.org/docrep/013/am307e/am307e00.pdf

Rushton, J., McLeod, A., and Lubroth, J. (2006). Managing transboundary animal disease.

Agriculture and Consumer Protection, FAO [On-line]. Available:

http://www.fao.org/docrep/009/a0255e/a0255e06.htm

Sheta, B. M., Fuller, T. L., Larison, B., Njabo, K. Y., Ahmed, A. S., Harrigan, R. et al. (2014).

Putative human and avian risk factors for avian influenza virus infections in backyard

poultry in Egypt. Veterinary Microbiology, 168, 208-213.

http://www.fao.org/bioenergy/26371-0b904ea1d132260451c3231064e3f440c.pdf

http://healthmap.org/promed/?p=152


http://www.fao.org/docrep/013/am307e/am307e00.pdf

http://www.fao.org/docrep/009/a0255e/a0255e06.htm

of 340

Sims, L. D. (2007). Lessons learned from Asian H5N1 outbreak control. Avian Diseases Digest,

51, 174-181.

Soares, R. J., Ortiz-Pelaez, A., Lai, T. K. L., Dinh, Q., Otte, J., and Pfeiffer, D. (2010).

Associations between attributes of live poultry trade and HPAI H5N1 outbreaks: a

descriptive and network analysis study in northern Vietnam. BMC Veterinary Research,

6, 10.

Thomas, D., Zerbini, E., Parthasarathy Rao, P., and Vaidyanathan, A. (2002). Increasing animal

productivity on small mixed farms in South Asia: a systems perspective. Agricultural

Systems, 71, 41-57.

Tiensin, T., Ahmed, S. S. U., Rojanasthien, S., Songserm, T., Parntep Ratanakorn, P., and et al.

(2009). Ecologic Risk Factor Investigation of Clusters of Avian In?uenza A (H5N1)

Virus Infection in Thailand. Journal of Infectious Diseases.

Tracey, J. P., Woods, R., Roshier, D., West, P., and Glen, R. (2004). The role of wild birds in the

transmission of avian influenza for Australia: an ecological perspective. EMU Austral

Ornithology, 104, 109-124.

Trinh, T. L. and Nguyen, P. L. (2009). Assessment of surface water quality by water quality

index (WQI) at the Cai Sao canal, An Giang province, Vietnam. Livestock Research for

Rural Development [On-line]. Available: http://www.lrrd.org/lrrd23/7/lan23151.htm

Vo, T. X. (2005). Changing the Life of Vietnamese farmers. Angiang University, Vietnam [On-

line]. Available: http://www.crawfordfund.org/wp-content/uploads/2014/03/xuan-

Changing-the-Life-Vietnamese-Farmers.pdf

http://www.lrrd.org/lrrd23/7/lan23151.htm

http://www.crawfordfund.org/wp-content/uploads/2014/03/xuan-Changing-the-Life-Vietnamese-Farmers.pdf

http://www.crawfordfund.org/wp-content/uploads/2014/03/xuan-Changing-the-Life-Vietnamese-Farmers.pdf

of 340

VPMO (2010). About the organisation of the national poverty survey to prepare for the

implementation of social secuity program and policy in the period of 2011- 2015 (Rep.

No. 1752/ CT-TTg). Vice Prime Minister Office (VPMO), the Government of Vietnam.

Vu, H. Q. (2016). Vietnamese Farmers Face the WTO: Implications for Sociology. Vietnam

Institute of Public Opinion Research [On-line]. Available:

http://www.ios.sinica.edu.tw/cna/download/6b_Quang.pdf

Vu, T. K. V., Tran, M. T., and Dang, T. T. S. (2007). A survey of manure management on pig

farms in Northern Vietnam. Livestock Science, 112, 288-297.

Waltner-Toews, D. (2011). Ecohealth: A Primer. Veterinarians without Borders/ Vétérinaires

sans Frontières - Canada.

Ward, M., Maftei, D., Apostu, C., and Suru, A. (2008). Environmental and anthropogenic risk

factors for highly pathogenic avian influenza subtype H5N1 outbreaks in Romania,

20052006. Veterinary Research Communications, 32, 627-634.




WHO (2007). Review of latest available evidence on potential transmission of avian influenza




WHO (2011). Water and Public Health. World Health Organization [On-line]. Available:

http://www.who.int/water_sanitation_health/dwq/S01.pdf

http://www.ios.sinica.edu.tw/cna/download/6b_Quang.pdf


of 340

WHO (2010). Cumulative number of confirmed human cases of avian influenza A(H5N1)

reported to WHO. World Health Organization [On-line]. Available:

http://www.who.int/influenza/human_animal_interface/H5N1_cumulative_table_archives

/en/

Yap, K. L., Yasmin, A. M., Wong, Y. H., Ooi, Y. E., Tan, S. C., Jegathesan, M. et al. (1992). A

one year community-based study on the incidence of diarrhoea and rotavirus infection in

urban and suburban Malaysian children. Med J Malaysia, 47, 303-308.

Zhao, K., Gu, M., Zhong, L., Duan, Z., Zhang, Y., Zhu, Y. et al. (2013). Characterization of

three H5N5 and one H5N8 highly pathogenic avian influenza viruses in China.

Veterinary Microbiology, 163, 351-357.

Zheng, T., Adlam, B., Rawdon, T. G., Stanislawek, W. L., Cork, S. C., Hope, V. et al. (2010). A

cross-sectional survey of influenza A infection, and management practices in small rural

backyard poultry flocks in two regions of New Zealand. New Zealand Veterinary

Journal, 58, 74-80.

Zhou, L., Liao, Q., Dong, L., Huai, Y., Bai, T., Xiang, N. et al. (2009). Risk Factors for Human

Illness with Avian Influenza A (H5N1) Virus Infection in China. Journal of Infectious

Diseases, 199, 1726-1734.

Zuckerman, J. N. (2011). Information on avian influenza ("bird flu") Academic Centre for Travel

Medicine and Vaccines and Royal Free Travel Health Centre, WHO Collaborating Centre

for Travel Medicine.

http://www.who.int/influenza/human_animal_interface/H5N1_cumulative_table_archives/en/

http://www.who.int/influenza/human_animal_interface/H5N1_cumulative_table_archives/en/

of 340

Chapter Three: Small-Scale Integrated Farmers’ Perceptions about Risk Factors for

Transmitting Water-Related Zoonotic Disease (WRZD) in the Provinces of Thai Binh and

An Giang in Vietnam

3.1. Introduction

The foundations for my research in this chapter were both a conceptual model and a

theoretical model to determine the perceptions of small-scale integrated (SSI) farmers about risk

factors concerning the transmission of water related zoonotic disease (WRZD)23 in the provinces

of Thai Binh and An Giang in Vietnam. The conceptual model is that farmers make decisions,

including management decisions, related to animal and human health that are influenced by their

impressions about the quality of water used on farms. The theoretical model of my research is

adapted from both the Health Belief Model and the Theory of Planned Behaviour (Becker, 2012;

Janz and Becker, 1984b; Rosenstock et al., 1988). The theoretical model of my research indicates

that people’s perceptions of a personal threat to their health and or wellbeing, together with their

belief in the benefits and barriers of an action, will predict the likelihood of their health-related

behaviours (See section 3.2. and Appendix A). My research in this chapter explored the first part

of the theoretical model, which includes key constructs of perceptions. These constructs were: 1)

SSI farmers’ perceived quality of general water environment in their villages; 2) SSI farmers’

perceived threat of WRZD (e.g., perceived susceptibility to WRZD); 3) SSI farmers’

expectations (e.g., perceived barriers of taking actions to reduce the transmission of WRZD); and

23 In this research, I used the World Health Organization’s (WHO) definition of WRZD as diseases in humans that

are spread from animals and related to water such as Highly Pathogenic Avian Influenza, some parasites, or diseases

caused by a range of bacteria (e.g., pathogenic strains of Escherichia coli) (WHO, 2004). The criteria for

determining WRZD include: 1) the pathogen must spend part of its life cycle within one or more animal species; 2)

it is probable or conceivable that some life stage of the pathogen will enter water; and 3) transmission of the

pathogen from animals to humans must be through a water related route (Moe, 2004).

of 340

4) SSI farmers’ perceived triggers/cues24 to taking actions that would mitigate25 the transmission

of WRZD.

3.3.1. Rationale for studying farmers’ perceptions of risk factors for transmission of WRZD

Perceptions in general are influenced by and reflected by individuals’ experiences and

actions, and can influence people’s health care seeking behaviors (Aburto et al., 2010; Ajzen,

1985; Levy and Myers, 2004; OECD, 2012). Farmers are not aware of, nor do they highly rank

health risk factors associated with their use of water contaminated by excreta from humans and

livestock. Rather, they view these risk factors related to production as unavoidable (Knudsen et

al., 2008b). Farmers did not consider health risks from water use to be significant because water

remains on the skin and can only cause skin problems; however, they thought that health risks

from non-composted smelly feces are serious because their odor enters the body through polluted

air (Knudsen et al., 2008b). Gender also plays a role because farmers mainly consider hygiene

and health to be women’s issues (Knudsen et al., 2008a).

Moreover, although farmers have a basic knowledge of proper hygiene, they carry out

hygienic practices (e.g., hand washing with soap) in an unsuitable manner (e.g., hand washing

without soap). Most likely, this comes from a misconception about the risk factors for the

transmission of WRZD, a lack of knowledge about cause-effect relationships, habits, and/or not

ranking the risk factors for transmission of WRZD highly while using water (Keraita et al.,

2010). In cases where farmers are aware of the health risk factors for transmitting WRZD, they

24 A trigger/cue to action is “the stimulus needed to trigger the decision making process to accept a recommended

health action”. Examples of possible cues to actions are chest pains, mass media campaigns, advice from others, a

reminder postcard from a physician, illness of family member or friend, etc.) (BUMC, 2013) 25 In this research, I defined “to mitigate transmission of WRZD” as to reduce the transmission of WRZD. An

example of an action to mitigate transmission of WRZD is to treat animal waste at source (Carr and Bartram, 2004;

WHO, 2004).

of 340

assess measures to mitigate transmission of WRZD by considering its overall economic impact

on work efficiency rather than taking into account the potential health benefits of such efforts

(Herbst et al., 2009; Keraita et al., 2008a; Owusu et al., 2011; Weldesilassie et al., 2011).

A number of studies in Vietnam have shown that some farmers do have a basic

awareness of health risks related to water use (Herbst et al., 2009; Keraita et al., 2008b; Keraita

et al., 2010; MoC and MARD, 2000a; Owusu et al., 2011). However, these studies had small

sample sizes of peri-urban farmers, did not include SSI farmers, and did not consider the risk

factors for transmission of WRZD. In Canada, a research study investigated an in-depth

understanding of participants’ perceptions and reasons for using water sources other than

municipal water (e.g., bottled water) and using in-home water treatment devices (Quinlan, 2005).

However, participants in this research were selected from the public, and not from SSI farmers.

A more comprehensive understanding of SSI farmers’ perceptions of risk factors for

transmission of WRZD (i.e., this chapter) may assist future studies about on-farm water quality

and public health and factors that contribute to farmers’ mitigating strategies against transmitting

WRZD (i.e., Chapter Four and Chapter Five respectively). It may also inform future research

relating to livestock production, water quality, and WRZD. Furthermore, this understanding can

contribute to informing related policies, formulating WRZD risk-reduction measures, and

developing mutually acceptable risk-management strategies and/or interventions.

3.1.2. Research question and objective

In this chapter, I present an exploratory research study addressing the following research

question: What are SSI farmers’ perceptions about risk factors for transmission of WRZD in the

provinces of Thai Binh and An Giang in Vietnam? The objective of this chapter is to describe the

perceptions of SSI farmers in these provinces about the risk factors for transmitting WRZD.

of 340

More specifically, in this chapter I describe the key constructs of SSI famers’ perceptions about

risk factors for the transmission of WRZD. I expect that SSI farmers have some impressions of

the general quality of water on farms and moderate threats of untreated water and WRZD to their

health and wellbeing. SSI farmers may indicate that economic related factors (e.g., values to

water for livestock production, and costs and economic gains of taking mitigating actions) as

their common barriers and benefits of concerns. SSI farmers may think that they have some

capacity in taking actions to reduce transmission of WRZD. I also hypothesize that there are

differences in perceptions of risk factors for transmission of WRZD between SSI farmers of the

two provinces.

This chapter’s results were also used in testing the hypotheses in Chapters Four and Five

of this manuscript. In the next section (i.e., Figure 3.1), I present a pictorial representation of the

structure of Chapter Three and its connection to Chapter Four and Five.

3.2. Methods

General information outlining the methods used in this research was presented in Chapter

Two (e.g., the research areas, the sample size, the pilot study, and the survey tools). In this

section, I will focus on aspects of the methods that I used to explore SSI farmers’ perceptions of

risk factors for the transmission of WRZD in the Vietnamese provinces of Thai Binh and An

Giang26. The questionnaires developed to collect information on farmers’ perceptions of risk

factors for the transmission of WRZD were designed based on a theoretical model adapted from

the Health Belief Model (Becker, 2012; Rosenstock et al., 1988) and the Theory of Planned

Behaviours (Payne et al., 2002) (See Appendix A).

26 Vietnam is divided into 64 provinces. Each province is sub-divided into districts, then communes, and villages.

of 340

Figure 3.1. Chapter Three’s Outline

Used to assist modeling factors associated with on-farm water quality (Chapter 4) and factors associated with SSI farmers’ engagement in

mitigating strategies to reduce transmission of WRZDs (Chapter 5) in the provinces of Thai Binh and An Giang in Vietnam


What are SSI farmers’ perceptions about risk factors for water related

zoonotic disease (WRZD) transmission in the provinces of Thai Binh

and An Giang in Vietnam?

Perception of trigger /cues

to taking mitigating

actions to reduce

transmission of WRZD

Perception of expectations

(e.g., perceived barriers of

and self-efficacy in taking

mitigating actions to reduce


Perceptions of threat to

health and/or wellbeing

(i.e., perceived

susceptibility to and

severity of WRZD)

Perceptions of general

water quality in the

villages

Chapter 3:

Presents SSI farmers' perceptions about risk factors for WRZD transmission in the provinces of Thai Binh and An Giang in Vietnam

of 340

More specifically, the theoretical model adapted from the Health Belief Model and the

Theory of Planned Behaviours states that people’s belief in a personal threat combined with their

belief in the benefits, barriers, and self-efficacy of an action will likely predict their behaviours.

Key constructs relating to people’s perceptions have been shown to have possible linkages to

their actions (See Appendix A). These constructs include the following: socioeconomic and

demographic factors; perceptions of threat (i.e., perceived susceptibility to and severity of

diseases); expectations (i.e., perceived benefits, perceived barriers, and self-efficacy); and cues to

action (e.g., media, personal influence, reminder) (Becker, 1974; Janz and Becker, 1984a;

Rosenstock et al., 1988). These constructs were used to guide the design of the questions used to

obtain the information relating to farmers’ perceptions of risk factors for the transmission of

WRZD, and to inform the hypothesis of the research. Various statistical tests (e.g., Pearson Chi-

squared test, t-test, and Fisher's exact test) were used to analyze the differences between farmers

of the two provinces regarding their perceptions of risk factors for WRZD transmission.

I used my knowledge of public health in Vietnam, especially of avian influenza

prevention and controls in Vietnam (Hall and Le, 2009; Le et al., 2014b; Le and Hall, 2011b; Le

and Hall, 2011a), to begin the design of my research study tools (i.e., the questionnaires, the

guidelines for in-depth interviews, and focus-group discussions). The study tools were then

revised based on my review of the literature, reports, and guidelines from the World Health

Organization (WHO, 2011; WHO, 2014a; WHO, 2014b), the Global Burden of Disease (GBD,

2014), and the Vietnamese government (MARD, 2012; MoC and MARD, 2000b; MoH, 2009b;

MoH, 2009a). A pilot study was conducted to test and further revise the study tools before using

them in the full field survey (see Chapter Two for more detailed information about the pilot

study).

of 340

SSI farmers’ perceptions of risk factors for the transmission of WRZD were explored in

the study; these included their perceptions about the following: the general condition of water

sources in their villages; susceptibility to and severity of WRZD transmission through contact

with livestock and water; expectations with respect to WRZD transmission (i.e., perceived

barriers, perceived benefits, and perceived self-ability to take actions that would reduce WRZD

transmission); and triggers for taking actions to reduce WRZD transmission (see Appendix A).

Both qualitative survey tools (i.e., in-depth interviews, focus group discussions, and

observations) and quantitative survey tools (e.g., Likert scale questionnaires and a close-ended

question checklist) were used to collect the information about SSI farmers’ perceptions of risk

factors for transmitting WRZD, which forms the hypothesis of this research. The following

paragraphs briefly explain the questions asked during the survey (For more detailed information

about these questions, see Appendix A).

Concerning SSI farmers’ perceptions about the overall water environment in their

villages, I asked the SSI farmers who participated in the study to subjectively assess the general

condition of the water environment in their villages, and the levels of harm to human and animal

health from the different untreated sources of water on their farms. I also asked them to indicate

factors that may influence their thoughts on the quality of water on their farms. First, using a

Likert scale questions (1 = very poor to 5 = excellent), the SSI farmers were asked to rank their

thoughts about the general condition of the water environment in their villages. Second, farmers

were also asked to indicate their perceived levels of harm to their health if they used untreated

sources of water (i.e., not boiled or filtered) such as drinking untreated water27, handling

27 Drinking water is defined as “water used for direct drinking or processing food” (MoH, 2009b)

of 340

domestic water28, handling pond water, and drinking bottled water29. Levels of harm were scaled

from 1 = no harm to 5 = extreme harm. Finally, SSI farmers were asked to indicate factors that

may influence their thoughts about water quality. These factors came from the literature as well

as the lessons learned from the pilot study. The SSI farmers had options to choose one or more of

these factors or to specify more details when answering this question.

Together, the perceived susceptibility towards diseases and the perceived severity of

diseases inform levels of perceptions of threat. Regarding perceptions of susceptibility to WRZD

as related to small-scale livestock production and water use, I asked the SSI farmers for their

thoughts about their levels of susceptibility to HPAI viruses, diarrhoea, coliform bacteria, and

parasites from using untreated water sources (i.e., not boiled or filtered). By using a scale from 1

to 5 (i.e., from 1 = no susceptibility to 5 = extremely high susceptibility), SSI farmers had the

option to indicate their thoughts about their levels of susceptibility to these diseases. This

question measured to what degree farmers perceived their susceptibility to a disease when using

untreated water; it was not about what causes them to be more or less susceptible to a disease. In

terms of farmers’ perceptions about the severity of diseases, I asked the SSI farmers to indicate

their thoughts about how severe diseases from HPAI viruses, diarrhoea, coliform bacteria, and

parasites are by using a scale of severity from 1 to 5 (i.e., 1= trivial to 5 = very severe).

Expectations are also a construct of perceptions. Expectations may contribute to

influencing farmers’ health behaviours. Expectations in this research have two sub-constructs

including 1) any perceived barriers that may prevent farmers from taking mitigating actions

28 Domestic water is defined as “water used for domestic use but not for direct drinking or processing food” (MoH,

2009a) 29 In my survey, bottled water was referred to as water packed in bottles by approved suppliers and available as a

commercial product. Farmers bought bottled water for drinking purpose without any further treatment by

themselves.

of 340

against the transmission of WRZD and 2) perceived benefits that encourage farmers to take or

not take mitigating actions against transmitting WRZD. Regarding perceived barriers to actions,

I asked SSI farmers a question about what barriers might prevent them from taking mitigating

actions against the transmission of WRZD. This question had several options for the SSI farmers

to choose from, as well as an option to specify detailed answers. In terms of the perceived

benefits of taking or not taking actions, I asked the SSI farmers to specify what benefits

encouraged or did not encourage them to take mitigating actions against the transmission of

WRZD. Farmers were asked to choose one of the benefits listed in the questionnaire and/or to

indicate benefits that were not listed in the question.

Perceived self-efficacy is indicated in the theoretical model of this research as one

construct of perceptions that may influence people’s health behaviours. I asked the SSI farmers

to indicate their faith in their ability to perform four key clusters of mitigating actions against

transmitting WRZD. The first cluster of mitigating actions was livestock health management

(e.g., manage animal waste). The second cluster was taking mitigating actions at the water

storage, treatment, and distribution level (e.g., storing water in a confined tank). The third cluster

was taking mitigating actions relating to the protection of water sources (e.g., using of buffer

zones). The forth cluster was taking mitigating actions at points of use/household levels (e.g.,

hand washing with soap). These clusters were selected based on the WHO’s guidelines about the

control envelope and risk management for WRZDs (WHO, 2004).

Cues to actions such as media, personal influence, and reminders together with

perceptions of threats and expectations are indicated in the theoretical model of this research as

one of the key constructs that may influence people’s health behaviours. In the cue to actions

question, there were options for triggers or cues to actions as well as an option for specifying

of 340

more detail. I asked the SSI farmers to select the triggers/cues (and provide specification if any)

that influenced their decisions to take mitigating actions against the transmission of WRZD, or to

specify any triggers or cues to their actions that were not listed in the questionnaire.

3.3. Results

3.3.1. Perceptions of water quality in the villages

Table 3.1 shows the results with respect to farmers’ perceptions about the general

conditions of the water environment in their villages. In both provinces, almost 70% of the SSI

farmers thought that the water environment conditions in their villages were either good or

adequate. A very small percentage thought their water environment was in either excellent or

very poor condition (1.2% and 2.9% respectively). None of the SSI farmers in Thai Binh

considered their water environment as excellent and 4.4% of them thought their water

environment was very poor. Meanwhile, 2.4% of the SSI farmers in An Giang considered their

water environmental condition to be excellent while 1.3% believed it to be very poor (Table 3.1).

In addition, the percentage of SSI farmers in An Giang who thought that their sources of

untreated water (i.e., not boiled or filtered) were safe for drinking were much greater than in Thai

Binh (17% and 1% respectively).

of 340

Table 3.1. Perceived General Condition of the Water Environment in the Villages of Small-

Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam

Scale Both provinces Thai Binh An Giang

Freq. Percent Freq. * Percent Freq. Percent

Excellent 7 1.2 - - 7 2.4

Good 104 17.4 18 6.0 86 28.9

Adequate 275 46.1 164 54.9 111 37.3

Poor 194 32.5 104 34.8 90 30.2

Very poor 17 2.9 13 4.4 4 1.3

Total 597 100.0 299 100.0 298 100.0

(*) indicates significant difference between Thai Binh and An Giang with p = 0.000, two-tailed Fisher's exact test

Most farmers thought there was at least some harm when using untreated sources of

water. In particular, farmers thought that if they consumed untreated drinking water or handled

untreated domestic water, there would be moderate to high levels of harm to their health (Table

3.2), however, that drinking bottled water imposed no or low harm to health. According to the

SSI farmers in both provinces, handling pond water was considered to result in high or extreme

levels of harm to health (Table 3.2).

of 340

Table 3.2. Perceived Harm to Health When Using Different Sources of Water among

Small-Scale (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam

Province/Source of water Obs Mean a Std. Dev.

Both provinces

Untreated drinking water 574 3.8 0.7

Untreated domestic water 576 3.5 0.9

Pond water 511 4.1 0.8

Bottled water 475 1.7 0.8

Thai Binh

Untreated drinking water 280 3.7*** 0.8

Untreated domestic water 282 3.7*** 0.8

Pond water 275 4.3*** 0.6

Bottled water 202 1.8*** 0.7

An Giang

Untreated drinking water 294 3.9 0.7

Untreated domestic water 294 3.3 1.0

Pond water 236 3.9 0.9

Bottled water 273 1.6 0.8 a mean of level of harm with the scale from 1 = No harm, 2 = Low harm, 3 = Moderate harm, 4 = High harm,

to 5 = Extreme harm

(***) indicates a statistical significant difference between Thai Binh and An Giang p < 0.01

Table 3.3 shows that in Thai Binh the two most cited factors for farmers as influences on

their thinking about the quality of a water source were their impressions of water quality such as

the taste of the water and their previous experience with that water source (80% and 67% of the

SSI farmers in Thai Binh respectively). In contrast, in An Giang, previous experience with water

use was the factor that influenced most SSI farmers (73%) in their perception about water

quality; moreover, fewer of them indicated that impressions of water quality such as taste

influenced their perceptions about the quality of water. Media also influenced the perceptions of

SSI farmers about water in Thai Binh, and more so than in An Giang (i.e., 35% and 5%

respectively). Other listed factors in Table 3.3 appeared to influence the general perceptions of

of 340

water quality among a smaller proportion of the SSI farmers in both provinces (i.e., ranging from

2% to 13% of the farmers).

Table 3.3. Factors Influencing Small-Scale Integrated (SSI) Farmers’ Thoughts about

Water Quality in the Provinces of Thai Binh and An Giang, Vietnam

Factors An Giang Thai Binh

Freq. Percent Freq. Percent

Impressions (e.g., taste) 72*** 24 239 80

Previous experiences 208 70 201 67

Risk perceptions 5*** 2 38 13

Attitudes towards water chemical

contaminants

7*** 2 34 11

Contextual cues relating to the

source of water supply

14** 5 33 11

Familiarity with specific water

properties

30 10 28 9

Trust in suppliers/sources 7*** 2 23 8

Media 14*** 5 104 35

(***) significant difference between Thai Binh and An Giang with p = 0.000, Pearson Chi-squared test

(**) significant difference between Thai Binh and An Giang with p = 0.01, Pearson Chi-squared test

3.3.2. Perceptions of threat to health and/or wellbeing

As indicated in the theoretical model of my research (see Appendix A), perceptions of

threat to health and/or wellbeing are one of the constructs of perceptions that may influence

people’s health behaviours. Together, the perceived susceptibility to diseases and perceived

severity of diseases inform the levels of perception of threat. The following sections present the

results of the level of perceived susceptibility as well as the level of perceived severity of key

diseases relating to livestock/poultry production (i.e., HPAI, diarrhea, coliform bacteria, and

parasites) when using various types of on-farm water (i.e., drinking water, handling domestic

water, handling pond water, and handling livestock wastewater) among SSI farmers.

of 340

3.3.2.1.Perceived susceptibility to diseases

With respect to perceptions of susceptibility to diseases relating to SSI livestock

production and water use, Table 3.4 shows the perceived susceptibility among SSI farmers to

highly pathogenic avian influenza (HPAI), diarrhoea, coliform bacteria, and parasites from

untreated water sources. In Thai Binh, none of the SSI farmers thought that drinking and/or

handling untreated water has either a low or no level of exposure30 to HPAI, diarrhoea, coliform

bacteria, or parasites. On the contrary, SSI farmers thought that drinking untreated water and

handling untreated domestic water had a moderate to high level of susceptibility to HPAI,

diarrhoea, coliform bacteria, and parasites (i.e., mean level of susceptibility from 3.4 to 3.8).

Additionally, when it came to issues of handling untreated pond water and livestock wastewater

the SSI farmers in Thai Binh considered this to result in a high and extremely high level of

susceptibility to HPAI, diarrhoea, coliform bacteria, and parasites (i.e., mean level of

susceptibility from 4.2 to 4.6). Among the SSI farmers in An Giang, the perceived levels of

susceptibility to HPAI viruses, diarrhoea, coliform bacteria, and parasites from untreated water

sources were similar to the perceived levels of susceptibility among the Thai Binh SSI farmers.

However, there were a few exceptions. The first was that the An Giang SSI farmers

considered drinking untreated water as being a high susceptibility for and extremely high

susceptibility for diarrhea and parasites respectively (i.e., mean levels of susceptibility of 4.0 and

4.2 respectively). These mean levels of susceptibility were greater than those among the Thai

Binh SSI farmers (i.e., mean level of susceptibility 3.8 and 3.7 respectively). The second

30 This phrase emphasizes on farmers’ perceived susceptibility to HPAI, diarrhea, coliform bacteria, or parasites

from using untreated drinking water, not the act of drinking itself. I expected and observed during the data collection

that farmers’ perceptions of susceptibility to diseases were limited to as simple as exposure and risk of having a

disease, and did not include other aspects of susceptibility to a disease.

of 340

exception was that the SSI farmers in An Giang believed they were less susceptible to HPAI,

diarrhoea, coliform bacteria, and parasites (i.e., mean levels of susceptibility from 3.7 to 3.9)

when handling untreated pond water than did their counterparts in Thai Binh (i.e., mean levels of

susceptibility from 4.2 to 4.5).

3.3.2.2. Perceived severity of diseases

With respect to SSI farmers’ perceived levels about the severity of diseases from HPAI

viruses, diarrhoea, coliform bacteria, and parasites, Table 3.5 shows that SSI farmers indicated

“severe” as the most common level of severity for HPAI, diarrhoea, coliform bacteria, and

parasites. The percentages of SSI farmers who considered bird flu very severe if contracted were

similar between the two provinces (i.e., 28%). When combining the responses in the perceived

severe and the perceived very severe levels, the results showed that about 80% and greater of SSI

farmers in both Thai Binh and An Giang considered HPAI, diarrhoea, and coliform diseases as

being severe and very severe if they had these diseases. Less than 12% of them considered it

mild or trivial if they contracted HPAI, diarrhoea, and coliform disease. None of the SSI farmers

in either province considered diarrhea or coliform diseases as being trivial.

of 340

Table 3.4. Perceived Susceptibility to Highly Pathogenic Avian Influenza (HPAI),

Diarrhoea, Coliform Bacteria, and Parasites from Untreated Water Sources among Small-

Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang, Vietnam

Perceived susceptibility to Thai Binh An Giang

Mean a Std. Dev. Mean Std. Dev.

HPAI from b

Drinking water 3.4* 0.9 3.9* 0.9

Handling domestic water 3.5* 0.9 3.6 0.9

Handling pond water 4.2* 0.8 3.9* 0.9

Handling livestock wastewater 4.5* 0.6 4.1* 1.0

Diarrhea from c

Drinking water 3.8*** 0.7 4.0*** 0.7

Handling domestic water 3.8*** 0.7 3.6*** 1.0

Handling pond water 4.4*** 0.6 3.9*** 0.9

Handling livestock wastewater 4.5*** 0.6 4.2*** 0.8

Coliform bacteria from d

Drinking water 3.7** 0.6 3.9** 0.6

Handling domestic water 3.7** 0.6 3.6** 0.8


Handling livestock waste water 4.6*** 0.5 4.1*** 0.8

Parasites e

Drinking water 3.7*** 0.6 4.2*** 0.7

Handling domestic water 3.8 0.7 3.9 0.9


Handling livestock wastewater 4.6*** 0.5 4.3*** 0.8 a scaled as 5 = extremely high susceptibility; 4 = high susceptibility; 3 = moderate susceptibility; 2 = low

susceptibility; and 1 = no susceptibility; b n = 456; c n = 498; d n = 207; e n = 382; (*) significant difference between

Thai Binh and An Giang p < 0.1; (**) significant difference between Thai Binh and An Giang p < 0.05; (***)

significant difference between Thai Binh and An Giang p < 0.01;

of 340

Table 3.5. Perceived Severity of Diseases from Highly Pathogenic Avian Influenza (HPAI),

Diarrhoea, Coliform Bacteria, and Parasites among Small-Scale Integrated (SSI) Farmers

in the Provinces of Thai Binh and An Giang, Vietnam

Severity a Thai Binh An Giang


HPAI

Very Severe 77 28 67 28

Severe 184 67 153 64

Moderate 13 5 16 7

Mild - - 2 1

Trivial 1 0 - -

Total 275 100 238 100

Diarrhea

Very Severe 8* 3 40* 14

Severe 205* 77 191* 65

Moderate 50* 19 59* 20

Mild 3* 1 5* 2

Trivial - - - -

Total 266 100 295 100

Coliform


Severe 121* 70 27* 63

Moderate 44* 25 3* 7

Mild - - 2* 5

Trivial - - - -

Total 173 100 43 100

Parasites


Severe 131* 54 72* 43

Moderate 101* 42 56* 34

Mild 5* 2 18* 11

Trivial - - 2 1

Total 241 100 166 100 a

scaled as 5 = very severe; 4 = severe; 3 = moderate; 2 = mild; and 1 = trivial


of 340

3.3.3. Expectations

The study results relating to farmers’ expectations included three sub-constructs: 1)

farmers’ perceived barriers that prevent them from taking mitigating actions against transmitting

WRZD; 2) farmers’ perceived benefits that encourage them to either take or not take mitigating

actions against the transmission of WRZD; and 3) farmers’ perceived self-efficacy to take

mitigating actions against the transmission of WRZD.

3.3.3.1. Perceptions of barriers to taking mitigating actions

The first sub-construct of perceptions of threat is perceptions of barriers to actions to

mitigate transmission of WRZD. Table 3.6 below shows the results of barriers to SSI farmers for

taking mitigating actions against the transmission of WRZD. The most common barriers to

actions were the responses of “do not know” and “cannot afford the cost” in Thai Binh and An

Giang respectively. In both provinces, “neighbours do not take actions” was indicated as a

barrier to taking mitigating actions by a small proportion of the SSI farmers in each province

(4% and 5% respectively). Beyond the above possible barriers to taking mitigating actions as

response options for SSI farmers, the only other barrier that SSI farmers specified was “lack of

land,” yet only three out of 598 SSI farmers indicated this.

3.3.3.2. Perceptions of benefits of taking or not taking mitigating actions

The second sub-construct of perceptions of threats is the perception of benefits from

either taking or not taking actions against WRZD transmission. Table 3.7 below shows five key

benefits of taking mitigating actions against WRZD transmission as indicated by SSI farmers in

both An Giang and Thai Binh. Farmers indicated the most common benefit in both Thai Binh

and An Giang as “economic gains” (i.e., 37% and 42% respectively). As for An Giang, only the

response “being a good model for others to follow” was indicated as a benefit for taking

of 340

mitigating actions by less than 1% of the SSI farmers. On the other hand, 26% and 18% of the

SSI farmers in Thai Binh and An Giang respectively indicated that there were no benefits to

taking mitigating actions against WRZD transmission. Most SSI farmers in Thai Binh (98%)

indicated that there were no benefits from not taking any mitigating actions. However, in An

Giang, 49% and 42% of the SSI farmers respectively indicated that they received “no benefits”

but had “economic gains” from not taking mitigating actions against the transmission of WRZD.

Table 3.6. Perceived Barriers to Taking Actions to Mitigate Transmission of WRZD among


Vietnam

Barriers to taking mitigating

actions Thai Binh An Giang


Cannot afford the costs 95*** 32 144 48

Do not know how 114 38 118 40

Lack of understanding 90*** 30 31 10

Neighbors do not take actions 13 4 15 5

Too busy 106*** 35 47 16 (***) significant difference between Thai Binh and An Giang with p = 0.000, Pearson Chi-squared test

of 340

Table 3.7. Perceived Benefits that Encourage Small-Scale (SSI) Farmers in the Provinces of

Thai Binh and An Giang, Vietnam, to Either Take or Not Take Actions to Mitigate

Transmission of WRZD

Benefits Thai Binh An Giang


Benefits of taking mitigating actions a

Disease prevention for humans (a) 41* 14.0 54 18.9

Economic gains (b) 108* 36.7 121 42.3

Environment/water protections (c) 35* 11.9 25 8.7

(a)+(b)+(c) 28* 9.5 29 10.1

No benefit 77* 26.2 51 17.8

Disease prevention for animals 5* 1.7 4 1.4

Being good models for others to follow - - 2 0.7

Benefits not taking mitigating actions b

No benefit 295** 98.3 121 48.8

Time (d) 1** 0.3 19 7.7

Economic gains (e) 4** 1.3 103 41.5

Avoid being/feeling strange - - 1 0.4

(d) +(e) - - 3 1.2

Convenience - - 1 0.4 a Thai Binh (n = 294) and An Giang (n = 286),

b Thai Binh (n = 300) and An Giang (n = 248)


(**) significant difference between Thai Binh and An Giang with p = 0.000, two-tailed Fisher's exact test

Further to exploring farmers’ perceptions about the benefits of taking mitigating actions

against the transmission of WRZD, Table 3.8 shows SSI farmers’ perceptions of the value of

water and wastewater for their livestock production. Concerning the value of their water, about

60% of the SSI farmers in both Thai Binh and An Giang indicated that water has moderate to

high value for their livestock production. Moreover, double the percentage of SSI farmers in

Thai Binh, compared to An Giang, considered water as being of extremely high value (13% and

6% respectively). The rest of the SSI farmers thought that water has low or no value to their

livestock production. Regarding wastewater, 30% and 77% of the SSI farmers in Thai Binh and

of 340

An Giang respectively considered that the wastewater from livestock production has no value for

their livestock production (Table 3.8).

Table 3.8. Perceptions of Water and Wastewater Value for Livestock Production among


Vietnam

Value levels a

Thai Binh An Giang


Water b

Extremely high value 21* 13 17 6

High value 29* 17 107 39

Moderate value 74* 45 64 23

Low value 12* 7 25 9

No value 30* 18 63 23

Wastewater c

Extremely high value 2** 1 2 1

High value 40** 14 8 3

Moderate value 122** 41 31 12

Low value 43** 15 21 8

No value 89** 30 207 77 a scaled as 1 = no value, 2 = low value, 3 = moderate value, 4 = high value, and 5 = extremely high value

b Thai Binh (n = 166) and An Giang (n = 276),

c Thai Binh (n = 296) and An Giang (n = 269)

(*) significant difference between Thai Binh and An Giang with p = 0.000, Pearson Chi-squared test

(**) significant difference between Thai Binh and An Giang with p = 0.000, two-tailed Fisher's exact test

In addition, farmers’ perceived levels of importance about other issues relating to on-

farm water in reducing WRZD transmission are presented in Table 3.9. A total of 95% of SSI

farmers in both An Giang and Thai Binh indicated that on farm water-related issues31 were of at

least some importance in reducing risk factors for the transmission of WRZD (i.e., from low

important to extremely important). The response “highly important” was the most frequently

indicated level of importance of the on-farm water-related issues in reducing transmission of

WRZD (i.e., ranging from 44% to 71%). There was an exception to the issue of “rainfall”, which

31 Including water sources, sharing water between animals and humans, water sources-animals distance, lack of

water, rainfall, and water quality

of 340

the SSI farmers in An Giang considered as being a “moderately important” issue instead of a

“highly important” issue in reducing risk factors for the transmission of WRZD.

3.3.3.3.Perceptions about self-efficacy in performing mitigating actions

Table 3.10 below shows farmers’ perceived actual capacity to take mitigating actions

against the transmission of WRZD (i.e., self-efficacy). Most SSI farmers in both provinces

considered they had a moderate to high ability to perform all of the key clusters of mitigating

actions. Moreover, about 10% of them thought they had a very high ability to do so, whereas

none of them considered themselves to have a weak ability to do so. No SSI farmers in either

province thought they had a weak or very weak capacity to manage livestock health. Less than

15% of the SSI farmers indicated that they were weak in undertaking mitigating actions at the

levels of protection of sources of water, water storage, treatment, and distribution, and point of

use/household level.

of 340

Table 3.9. Levels of Importance for Water Use Issues in Reducing Transmission of WRZD

among Small-Scale Integrated (SSI) Farmers in the Provinces of Thai Binh and An Giang,

Vietnam

Water use issues/Importance Thai Binh An Giang


Water sources a

Extremely important 33* 12 27 9

Highly important 164* 59 209 71

Moderately important 75* 27 47 16

Low important 4* 1 7 2

Not important 3* 1 3 1

Sharing water between animals and humans b






Water sources-animals distance c






Lack of water d





Not important 10 4

Rainfall e

Extremely important 16* 6





Water quality f





Not important 1* 0 10 4 a Thai Binh (n = 279) and An Giang (n = 293); b Thai Binh (n = 273) and An Giang (n =288) c Thai Binh (n = 270) and An Giang (n = 283); d Thai Binh (n = 261) and An Giang (n =256) e Thai Binh (n = 260) and An Giang (n =245); f Thai Binh (n = 255) and An Giang (n =247)


of 340

Table 3.10. Perceived Self-efficacy to Perform Mitigating Actions among Small-Scale

Integrated Farmers in Thai Binh and An Giang, Vietnam

Perceived self-efficacy Thai Binh An Giang


In livestock health management a

Very high 29 10 35 12

High 122 41 77 26

Moderate 125 42 149 50

Weak 19 6 37 12

Very weak - - - -

Protection of sources of water b

Very high 23 8 32 11

High 124 41 76 26

Moderate 115 38 157 53

Weak 36 12 32 11

Very weak 1 0 - -

In water storage, treatment and distribution c

Very high 19 6 29 10

High 121 41 70 24

Moderate 112 38 166 56

Weak 45 15 29 10

Very weak 1 0 - -

Taking mitigating actions at point of use d

Very high 44 15 32 12

High 94 32 80 31

Moderate 152 51 127 49

Weak 6 2 21 8

Very weak 1 0 - - a Thai Binh (n = 295), An Giang (n = 298), and significant difference between Thai Binh and An Giang with p =

0.000, Pearson Chi-squared test b Thai Binh (n = 299), An Giang (n =297); significant difference between Thai Binh and An Giang with p = 0.000,

two-tailed Fisher's exact test c Thai Binh (n = 298), An Giang (n = 294), and significant difference between Thai Binh and An Giang with p =

0.000, two-tailed Fisher's exact test d Thai Binh (n = 297) and An Giang (n =260), significant difference between Thai Binh and An Giang with p =

0.000, two-tailed Fisher's exact test

of 340

3.3.4. Cues to actions

Figure 3.2 shows the common triggers/cues that SSI farmers in both provinces indicated

prompt them to take mitigating actions against the transmission of WRZD. In An Giang, the two

most frequently cited triggers/cues to action were “radios/newspapers” and “advice from

health/veterinary workers,” whereas in Thai Binh it was “worry about being infected or affected

with WRZD” and “radios/newspapers”. In contrast, the two least frequently cited triggers/cues to

action in Au Giang were “not busy with other tasks” and “lose income if do not take actions”

whereas in Thai Binh they were “not busy with other tasks” and “neighbors do the same

mitigating actions”. Worry about being affected by the transmission of WRZD was the most

common cue for taking mitigating actions in Thai Binh, while it was the third most common cue

in An Giang.

Figure 3.2. Triggers/Cues to Taking Mitigating Actions against Transmission of Water

Related Zoonotic Diseases (WRZD)

46

148

208

10

263

113

55

147

197

26

111

52

050

100

150

200

250

Thai Binh An Giang

Neighbors do the same mitigating actions (p = 0.308) Health workers' advice (p = 0.999)

Media - radios/newspapers (p = 0.399) Not busy with other tasks (p = 0.001)

Worried about being infected/affected by EID (p = 0.000) Lose income (p = 0.000)

Num

bers

of f

arm

ers

Graphs by Provinces

Trigger/cues to taking mitigation actions

of 340

3.4. Discussion

The results of farmers’ perceptions of risk factors for the transmission of WRZD

supported my hypothesis formulated prior to conducting the survey that there are differences in

the perceptions of risk factors for transmission of WRZD between SSI farmers of the provinces

of Thai Binh and An Giang. They also appeared to have a good sense of their water environment

conditions. For example, SSI farmers in Thai Binh appeared to be more skeptical than those in

An Giang when assessing the conditions of their water environment. Meanwhile, other studies

have shown some differences in the density of livestock production and water resources between

Thai Binh and An Giang. For instance, the number of poultry produced in Thai Binh per year

(8,899,000) was almost double that in An Giang (4,067,000) (GSO, 2012a). The total number of

rural households using aquaculture land in Thai Binh was also about 10 times greater than that in

An Giang (GSO, 2012b). As well, the total area of surface water for aquaculture in Thai Binh

was much greater than that in An Giang (13.4 and 1.9 thousand hectares respectively) (GSO,

2010). Interestingly, the total number of communes in Thai Binh that have a common sewage

system was about half to those in An Giang (GSO, 2011). Therefore, it is likely that the general

water environment conditions in Thai Binh are not as safe as those in An Giang. I suggest that

SSI farmers with greater densities of livestock production and aquaculture land resources may be

more skeptical of their water environment conditions. However, further studies are needed to

assess the SSI water environment conditions and relate them to the corresponding perceptions of

SSI farmers.

The levels of harm to health from using various untreated sources of water as stated by

the SSI farmers in the survey are in line with my expectations before the survey. SSI farmers

were aware of the low quality as well as the potential harm caused by using untreated on-farm

of 340

water sources. Nonetheless, SSI farmers shared in in-depth interviews and focus group

discussions that bottled water is expensive and not always affordable for them. Given the

relatively high price and high level of technology used to produce bottled water compared to

other sources of on-farm water, it is understandable why SSI farmers believed in the high quality

and low level of harm of bottled water.

In my previous experience in the studied area, SSI farmers judged their water quality

based on subjective factors such as water turbidity or colour, its taste in tea, or simply their

feeling after using the water for some time. Thus, it is not surprising that most SSI farmers in the

survey assessed their water quality according to their own impressions such as how it tastes,

rather than using any objective assessments by themselves or by local health authorities. I also

observed32 that SSI farmers had a great need for the water quality on their farms to be assessed.

However, they did not know how to do this, nor could they afford to have their water assessed.

Meanwhile, the local government agencies in Thai Binh and An Giang primarily monitor and

assess, to some extent, the water quality provided by industrial suppliers, but not the water

quality available at the farm/household level. As a result, there exists a dire need to introduce

simple and affordable on-farm water quality testing and to improve the capacity of local health

authorities to conduct rural water quality monitoring, assessment, and validation. Simple and

affordable on-farm water quality testing tools may help change the water-related behaviour of

farmers by enhancing visualization and ownership of water testing for farmers. Alternatively,

engaging the private sector may be a solution for delivering simple and affordable on-farm water

quality testing tools together with affordable options for on-farm water treatment and storage.

32 Field investigator/enumerators observed through questions asked in interviews and focused group discussion

of 340

Nonetheless, more studies are needed to explore the concept of the objective assessment of on-

farm water using basic, no-frills, inexpensive on-farm water quality testing tools or other low-

priced options for on-farm water treatment and storage.

One outcome that I was not expecting in the results was the level of influence of media

such as radio and television, and print media such as newspapers on farmers’ perceptions in Thai

Binh, which was much greater than in An Giang. This may reflect regional differences in

lifestyles and cultures between the two river deltas. As a result, it may be necessary to consider

the differing levels of media influence on farmers’ perceptions in the two river deltas when

developing relating policies and programs.

Given the insights about farmers’ perceptions of the general conditions of their on-farm

water and the limited ways farmers engage to assess the quality and the level of harm of their on-

farm water, it is apparent that field-based objective assessments of on-farm water quality and/or

the risk of disease from on-farm water use are needed. Such field-based assessments may help to

quantify the quality of on-farm water and help better understand differences in the overall

perceptions and assessment of water quality among farmers in Thai Binh and An Giang. Further

studies in not only health but also communication may also be needed to explore the role of

media in influencing farmers’ perceptions in different regions of Vietnam.

Concerning the perceived threat of disease related to water and livestock production, the

results reflected my expectations that farmers were aware of and noticed the susceptibility to and

severity of diseases. Their levels of perceived susceptibility and perceived severity of diseases

were consistent with their concern for on-farm water quality. With understanding their perceived

level of susceptibility to disease, it may be useful to explore the association, if any, between

farmers’ perceptions and their mitigating strategies against transmitting WRZD. However, this

of 340

research did not explore other indicators such as immunity and age that may influence the level

of susceptibility to diseases. Therefore, this research did not aim at assessing the correctness of

farmers’ perceived level of susceptibility to diseases. Moreover, given the various waves of the

H5N1 epidemic that have occurred in Vietnam since 2003, I expected my research to show a

high percentage of SSI farmers who perceived HPAI as a severe infectious disease. However,

their percentage was slightly lower than for the percentage of urban consumers (75%) who

considered HPAI a serious threat to them during the first peak of the outbreak in 2004 (Figuie

and Fournier, 2008). Therefore, key differences between rural and urban populations such as

demographic and socioeconomic characteristics, population density, and access to information

may be important contributing factors to people’s perception of the severity of various diseases

that require further study.

With respect to farmers’ expectations of conducting mitigating actions, it is

understandable and was expected that farmers considered economic factors important in taking

mitigating actions. Economic factors were perceived by SSI farmers not only as barriers for

taking mitigating actions (e.g., cannot afford the costs), but also as important benefits if they did

not take mitigating actions against the transmission of WRZD. This result is consistent with

other research which also reported the importance of overall economic impact on how farmers

assess mitigating actions against WRZD (Keraita et al., 2010). Farmers’ expectations of how

they perceive self-efficacy or their perceived actual capacity in taking mitigating actions may

also influence their health behaviours around control of water use and water quality (Carr et al.,

2011). The large proportion of SSI farmers who indicated having a moderate to high ability to

perform mitigating actions could possibly be explained by the fact that the self-efficacy

questions referred to the key aspects of traditional small-scale integrated farming practices.

of 340

Self-efficacy is a person's belief in his/her ability to make a change or take an action

(Rosenstock et al., 1988). Perceived self-efficacy is indicated in the theoretical model of this

research as one of the constructs of perception that may influence people’s health behaviours. In

particular, perceived self-efficacy in this study was referred to as a farmer’s perceived actual

capacity in taking mitigating actions against the transmission of WRZD. Yet, the level of self-

efficacy may change when considering mitigating actions that go beyond or are out of the

context of traditional small-scale integrated farming practices. For example, farmers’ perceived

level of self-efficacy for private adaptive measures to climate changes (e.g., migration, changing

from farming to non-farming activities, and moving from crop to livestock) were reported as

being fairly low (Le et al., 2014a). Therefore, caution should be taken when generalizing the

levels of self-efficacy among farmers in the context of mitigating actions that meet certain

standards/recommendations rather than in the context of traditional small-scale integrated

farming.

In addition to perceptions of threats and expectations, triggers/cues to taking mitigating

actions against the transmission of WRZD were a construct in the theoretical model of this

research. Triggers/cues to action are considered as “a stimulus needed to trigger the decision-

making process to accept a recommended health action” (BUMC, 2013). Cues to actions (e.g.,

media) together with perceptions of threats and expectations are indicated in the theoretical

model of this research as one of the key constructs that may influence farmers’ health-related

behaviours. Given the huge impact of the H5N1 epidemics on the socioeconomic and health

conditions in Vietnam, the research results are reasonable in showing “worrying about being

affected by the diseases,” “radios/newspapers,” and “advice from CHWs/CVWs” as the top three

factors influencing SSI farmers’ reported mitigating actions. However, identifying triggers/cues

of 340

to taking mitigating actions other than health-related benefits may be needed to influence

farmers’ adoption of safer practices (Keraita et al., 2010).

One interesting observation is that the SSI farmers indicated economic factors (e.g., loss

of income) as the fourth cue to taking mitigating actions. This is different from the results of

farmers’ expectations (i.e., farmers’ perceived barriers and perceived benefits of taking

mitigating actions to reduce transmitting WRZD) in which economic factors were indicated as

the top benefit as well as the top barrier to taking mitigating actions. This difference may relate

to the urgency caused by the rapid transmission of WRZD. Therefore, this difference may be an

important factor to consider when developing long-term or short-term programs and/or policies

that address mitigating actions against the transmission of WRZD. In addition, the striking

difference between the number of SSI farmers in An Giang and Thai Binh who considered

“worry about being affected by the transmission of WRZD” as a trigger/cue to taking mitigating

actions is not surprising when considering the differences in lifestyle and economic conditions

between Northern and Southern Vietnam. This may have implications for developing strategies

for intervention and for behavioural changes between the provinces of Thai Binh and An Giang.

3.5. Conclusions

This chapter presented and discussed the research results of SSI farmers’ perceptions

about the following: their overall on-farm water quality; the threat of transmitting WRZD

including the perceived susceptibility for and perceived severity of transmitting WRZD; the

barriers and benefits to taking mitigating actions and the belief in their ability to take an action;

and the triggers for their actions to mitigate the transmission of WRZD.

SSI farmers perceived the overall on-farm water condition in their villages to be

adequate. They were also aware of the potential harmful health consequences of using untreated

of 340

water sources, and assessed water quality subjectively and primarily based on their own

impressions and previous experience. Most SSI farmers considered drinking untreated water and

handling untreated domestic water as being of moderate to high risk for the transmission of

HPAI, diarrhea, coliform bacteria, and parasites. If then infected, they believed that HPAI,

diarrhea, and coliform bacteria would be severe or very severe.

Most of the SSI farmers believed that they have moderate or high ability to conduct

mitigating actions against the transmission of WRZD. Moreover, economic factors were

perceived by most of the SSI farmers as both barriers and benefits to taking mitigating actions

against the transmission of WRZD. However, when it came to triggers/cues for taking mitigating

actions, economic factors were ranked as only the fourth factor that triggered action by most of

the SSI farmers. The top three factors were influence from the media, worry about being

infected, and advice from health workers.

The results presented in this chapter confirm my priori expectations that SSI farmers are

aware of the general quality of the water environment in their villages. They have some

impressions of risk factors for transmission of WRZD specifically moderate threats of WRZD to

their health and wellbeing, economic related factors as farmers’ common expected barriers and

benefits, and perceived moderate ability to taking actions to reduce transmission of WRZD.

Furthermore, this chapter also confirms my expected differences in perceptions of risk factors for

transmission of WRZD between SSI farmers of the two provinces. It will be useful to relate the

results of this chapter with objective assessments of on-farm water quality (see Chapter Four)

and to explore how farmers assess and make decisions about their mitigating strategies or how

these perceptions influence farmers’ choices about mitigating strategies (see Chapter Five).

of 340

References

Aburto, N. J., Pevzner, E., Lopez-Ridaura, R., Rojas, R., Lopez-Gatell, H., Lazcano, E. et al.

(2010). Knowledge and Adoption of Community Mitigation Efforts in Mexico During the

2009 H1N1 Pandemic. American Journal of Preventive Medicine, 39, 395-402.

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. Prentice-Hall

Englewood Cliffs, NJ.

Becker, G. S. (1974). Rational choice theory: A Theory of Social Interactions. Journal of

Political Economy, 82, 1063-1093.




Carr, G., Potter, R. B., and Nortcliff, S. (2011). Water reuse for irrigation in Jordan: Perceptions

of water quality among farmers. Agricultural Water Management, 98, 847-854.


Publishing.

EPA (1993). Method 180.1: Determination of Turbidity by Nephelometry. EPA [On-line].

Available:

http://water.epa.gov/scitech/methods/cwa/bioindicators/upload/2007_07_10_methods_me

thod_180_1.pdf

Figuie, M. and Fournier, T. (2008). Avian Influenza in Vietnam: Chicken-Hearted Consumers?

Risk Analysis, 28, 441-451.


http://water.epa.gov/scitech/methods/cwa/bioindicators/upload/2007_07_10_methods_method_180_1.pdf

http://water.epa.gov/scitech/methods/cwa/bioindicators/upload/2007_07_10_methods_method_180_1.pdf

of 340

GBD (2014). Global Burden of Disease Compare. Institute for Health Metrics and Evaluation

(IHME) [On-line]. Available: http://vizhub.healthdata.org/gbd-compare/

GSO (2010). Summary results of the Vietnam household living standard survey 2010 Vietnam

General Statistics Office (GSO).

GSO (2011). The 2009 Vietnam Population and Housing Cencus General Statistics Office,

Ministry of Planning and Invesment Vietnam.

GSO (2012a). Number of farms by province General Statistics Office.

GSO (2012b). Results of the 2011 Rural, Agricultrure and Fishery Census: Part 4 - Households

using aquacultural land by the size of land used and province Hanoi: Statistical

Publishing House - General Statistics Office.

Hall, D. C. and Le, Q. B. (2009). The midterm evaluation of the Government of Vietnam-UN

Joint Programme to Fight Highly Pathogenic Avian Influenza 2007-2010: Strengthening

the Management of Public Health Emergencies in Vietnam – with a focus on the

Prevention and Control of Diseases of Epidemic Potential including Highly Pathogenic

Avian Influenza (HPAI). The Government of Vietnam-UN Joint Programme to Fight

Highly Pathogenic Avian Influenza.




Janz, N. K. and Becker, M. H. (1984a). The Health Belief Model: A Decade Later. Health


http://vizhub.healthdata.org/gbd-compare/

of 340

Janz, N. K. and Becker, M. H. (1984b). The Health Belief Model: A Decade Later. Health


Keraita, B., Drechsel, P., and Konradsen, F. (2008a). Perceptions of farmers on health risks and

risk reduction measures in wastewaterirrigated urban vegetable farming in Ghana.

Journal of Risk Research, 11, 1047-1061.






Ltd.

Keraita, B., Jim'nez, B., and Drechsel, P. (2008b). Extent and implications of agricultural reuse

of untreated, partly treated and diluted wastewater in developing countries. Agr Vet Sci

Nutr Nat Resour, 3, 15.

Knudsen, L. G., Pham, D. P., Nguyen, H. T., Helle, S., Jensen, P. K., Dalsgaard, A. et al.

(2008a). The fear of awful smell: risk perceptions among farmers in Vietnam using

wastewater and human excreta in agriculture. The Southeast Asian journal of tropical

medicine and public health, 39, 341-352.

Knudsen, L. G., Pham, D. P., Nguyen, T. H., Samuelsen, H., Jensen, P. K., Dalsgaard, A. et al.

(2008b). The fear of awful smell: risk perceptions among farmers in Vietnam using

wastewater and human excreta in agriculture. Southeast Asian Journal of Tropical

Medicine and Public Health, 39(2), 341-352.

of 340

Le, D. H., Li, E., Nuberg, I., and Bruwer, J. (2014a). Farmers's assessments of private adaptive

measures to climate change and influential factors: a study in the Mekong Delta,

Vietnam. Nat Hazards, 71, 385-401.

Le, Q. B., Hall, D. C., and Cork, S. C. (2014b). Vietnam National Rural Clean Water Supply and

Sanitation Strategy (2020): Implications for the reduction of waterborne diseases among

small scale farmers and their animals. In GRF One Health Summit 2112: One Health -

One Planet - One Future Risks and Opportunities Global Risk Forum Davos.

Le, Q. B. and Hall, D. C. (2011a). Stakeholder cooperation and ecohealth to control Highly

Pathogenic Avian Influenza (HPAI) in Vietnam. Canadian Public Health Association

Annual Conference 2011, Montreal Quebec, 2011 [On-line]. Available:

http://resources.cpha.ca/CPHA/Conf/Data/2011/A11-270ae.pdf

Le, Q. B. and Hall, D. C. (2011b). Using an ecohealth approach to livestock production:

community surveillance of newly emerging infectious diseases (nEIDs), International

Meeting on Emerging Diseases and Surveillance (IMED 2011.

Levy, B. R. and Myers, L. M. (2004). Preventive health behaviors influenced by self-perceptions

of aging. Preventive Medicine, 39, 625-629.

MARD (2012). Provincial rural and agriculture database Vietnam: Planning Department,

Ministry of Agriculture and Rural Development.

MoC and MARD (2000a). National Rural Clean Water Supply and Sanitation Strategy up to the

Year 2020 Hanoi, Vietnam: Ministry of Construction and Ministry of Agriculture and



of 340

MoC and MARD (2000b). National Rural Clean Water Supply and Sanitation Strategy up to the

year 2020 Hanoi, Vietnam: Ministry of Construction and Ministry of Agricultrue and





MoH (2009a). National technical regulation on domestic water quality (QCVN 02: 2009/BYT).



MoH (2009b). National technical regulation on drinking water quality (QCVN 01: 2009/BYT).



OECD (2012). "Determinants of farmer behavioural change", in Farmer Behaviour, Agricultural

Management and Climate Change. In ( OECD Publishing.

Owusu, V., Bakang, J. E., Abaidoo, R., and Kinane, M. (2011). Perception on untreated

wastewater irrigation for vegetable production in Ghana. Environment, Development and

Sustainability, 1-16.

Payne, N., Jones, F., and Harris, P. (2002). The impact of working life on health behavior: The

effect of job strain on the cognitive predictors of exercise. Journal of Occupational

Health Psychology, 7, 342-353.

Quinlan, J. A. (2005). The use and perception of drinking water in a Canadian community. The

Universitiy of Guelph.

of 340

Rosenstock, I. M., Strecher, V. J., and Becker, M. H. (1988). Social Learning Theory and the


Weldesilassie, A. B., Boelee, L. I. N., Drechel, P. A. Y., and Dabbert, S. T. E. P. (2011).

Wastewater use in crop production in peri-urban areas of Addis Ababa: impacts on health

in farm households. Environment and Development Economics, 16, 25-49.


WHO (2011). Guidelines for drinking-water quality. (Fourth ed.) World Health Organisation.

WHO (2014a). Avian influenza in humans World Health Organization.

WHO (2014b). Diseases and injury country estimates: Burden of disease. World Health

Organization (WHO) [On-line]. Available:

http://www.who.int/healthinfo/global_burden_disease/estimates_country/en/

http://www.who.int/healthinfo/global_burden_disease/estimates_country/en/

of 340

Chapter Four: Water Quality and Public Health among Small-Scale Integrated (SSI)

Farms in the Provinces of Thai Binh and An Giang in Vietnam

4.1. Introduction

4.1.1. Rationale

Water supply and water quality have a profound influence on public health(WHO, 2011b;

WHO, 2011d). It is also estimated that water is the vehicle for about 80% of all water-related

zoonotic diseases (WRZD) in people (Ford and Colwell, 1996). Specifically, in the Red River

and Mekong River Deltas in Vietnam, water is considered to be “the heart of development,” yet

this location was also the major focus for the highly pathogenic avian influenza (HPAI) epidemic

waves that took place between 2003 and 2005 (Gilbert et al., 2008a; Henning et al., 2009; Phan,

2010; Sims et al., 2005; Soares et al., 2010). The distribution of early HPAI outbreaks during the

2003-5 HPAI epidemic in Vietnam was concentrated at the farm level in the Red River Delta and

the Mekong River Delta. This distribution was thought to be associated with bodies of water on

farms (Morris and Jackson, 2005). Avian influenza viruses (AIV) can persist for extended

periods of time in water depending on water temperature and other environmental factors (Boere

et al., 2006; Morris and Jackson, 2005; Webster et al., 2006; WHO, 2007a). Moreover, highly

Pathogenic Avian Influenza (HPAI) viruses can persist in water over a 60-day period and

constitute a potential source of waterborne AIV transmission for wild birds and domestic ducks

(Brown et al., 2007; Domanska et al., 2010; Nazir et al., 2010; Roche et al., 2009).

Environmental transmission of viruses in general can occur due to consuming virus-

contaminated drinking water, having contact with virus-contaminated water bodies, and being

exposed to poorly managed excreta and wastewater (WHO, 2005; WHO, 2007b). For example,

wild birds carry and shed AIV (Costa et al., 2011; Ip et al., 2008). Factors that might facilitate

of 340

AIV transmission between domestic and wild birds include: poultry operations located in wild

migratory bird flyways; watering and feeding areas open to access by wild birds and domestic

poultry; and waste runoff from domestic poultry operations that ends up in wetlands used by

wild birds (Rapport et al., 2006). These factors create favourable conditions for the co-infection,

re-assortment, and evolution of viruses among host species (Stallknecht et al., 1990; Ward et al.,

2009; Webster et al., 2006).

Infected wild birds, humans, and domestic animals shed AIV in their faeces and saliva

which could enter water environments (WHO, 2007a). High densities of livestock can also

pollute surface water such as lakes, ponds, streams, and rivers (Bui, 2010; Trinh, 2001).

Resultant water pollution levels can be 10 - 100 times greater than the acceptable level for use in

agriculture (Vu, 2001; Yajima and Kurokura, 2008). As a result, farmers need appropriate risk

mitigating measures because the direct use of untreated contaminated water can result in a

disease risk for humans residing in the areas (Dalsgaard, 2001; Guan and Holley, 2003; Toews,

2009; Trang et al., 2007a; Yajiam and Kurokura, 2008). Moreover, duck farms with a shared

water supply are often not fenced nor are they separated from streams and ponds and these farms

appear to be instrumental in AIV transmission in domestic ducks (Hubbard et al., 2004; Lin et

al., 2010; Markwell and Shortridge, 1982).

In Vietnam, farmers traditionally consider water to be the most important factor in

agriculture and this is commonly expressed in the proverb “nhất nước, nhì phân, tam cần, tứ

giống” meaning “water is the most important factor, and then comes fertilizer, industriousness,

and breeds” (Bach, 2007). However, over 70% of small-scale integrated (SSI) farmers use

contaminated water for drinking and other on-farm purposes and have no access to hygienic

of 340

latrines. Meanwhile, they have very limited awareness about water quality or environmental

sanitation (MoC, 2000; MoC and MARD, 2000). SSI farming also presents a number of risk

factors for WRZD transmission. Some examples of these factors for HPAI transmission include

the proximity of poultry flocks to water sources and their increased accessibility to them, an

increased density of ponds and streams, and temporal flooding (Gilbert et al., 2008b; Nguyen et

al., 2011; Pfeiffer et al., 2009).

In contrast, appropriate densities for livestock production and rational water use33 can

help reduce environmental problems, control nutrient loss, and reduce the release of pathogens

into agricultural runoff (Hooda et al., 2000). Yet, the inappropriate use of water and its

consequences to the health of SSI farmers and their animals has not been well studied. Therefore,

prior to undertaking a public health intervention to reduce the incidence of WRZD in rural

communities, it is important to understand the status of water and its management on individual

farms. This includes developing an understanding of water quality on farms, uses of water

sources, farmers’ awareness and perceptions of water quality, and water use with respect to

microbial infection (Dinh et al., 2004; Huynh et al., 2006; Lye, 2002; Nguyen et al., 2008; Tran

et al., 2010; Trang et al., 2007b; Trinh and Nguyen, 2009). The basic recommended indicators to

assess rural water quality include pH, turbidity, and the presence of coliform bacteria.

Escherichia coli (E. coli) is also an indirect indicator organism of other faecal microbes in water.

Meanwhile, levels of pH and turbidity have positive correlations with the pathogen removal in

33 Rational use of water in the context of livestock production may include good livestock management activities

such as hygienic rearing condition for livestock, adopting water quality standards for livestock, and treating

livestock waste at source (Carr et al., 2004)

of 340

water (WHO and OECD, 2003a; WHO, 2011a; WHO, 2011c; WHO, 2011b). Nevertheless,

methods for detecting viruses in rural water are limited (WHO, 2004).

Given the above, this research study had two objectives. First was to describe sources and

quality of water used on participating SSI farms. The second was to look for associations

between the quality of these water sources and farmers’ perceptions of risk factors for the

transmission of WRZD, which were presented in Chapter Three.

4.1.2. Research questions and hypotheses

4.1.2.1.Research questions:

This chapter addresses the fourth and fifth questions of my research, which are: 1) What

sources of water do SSI farms use for drinking and domestic purposes 34 and what are the

corresponding basic microbial and related indicators35 of water quality for those sources in the

provinces36 of Thai Binh and An Giang?, and 2) How does the quality of on-farm water SSI

farmers use relate to both their perceptions about water quality and risk factors for the


In Figure 4.1, I present a flow chart outlining the structure of Chapter Four.

4.1.2.2.Hypotheses:

1. The frequency of use of water sources for drinking and domestic purposes is different

among SSI farmers in the province of Thai Binh than in the province of An Giang.

34 In my research, drinking water is defined as “water used for direct drinking or food processing.” Domestic water

is defined as “water use for domestic purposes but nor for direct drinking or food processing” (MoH, 2009b; MoH,

2009a) 35 The WHO recommends Escherichia coli (E. coli) colony forming unit (cfu), turbidity, and pH as basic microbial

and related indicators for assessing rural water quality (WHO, 2012) 36 Vietnam is divided into 64 provinces. Each province is further sub-divided into districts, then communes and

villages.

of 340

2. SSI farms in the provinces of Thai Binh and An Giang are faced with low quality

water sources for drinking and domestic purposes.37

3. SSI farmers’ socioeconomic status and perceptions of on-farm risk factors for WRZD

transmission are associated with the differing microbial content of water on farms.

4.2. Methods

In this section, I focus on presenting the methods that were used to: describe the water

sources used on farms; objectively assess on-farm water quality in the studied areas; and to

analyze the relationship between the objectively assessed on-farm water quality (i.e., pH,

turbidity, and E. coli) and farmers’ subjective assessments of their on-farm water quality

(presented in Chapter Three). A more detailed description of the research methods used in this

chapter is presented in Appendices 4.6.1-2. Other information concerning the research methods

such as sample size, studied areas, research questionnaires, and research ethics were presented in

Chapter Two.

I used questionnaires to ask participating farmers to indicate the water sources they have

access to and their corresponding frequency of use for drinking and domestic purposes. In-depth

interviews, farm visits, and focus group discussions were also employed to obtain more insights

37 These terms are defined in section 4.2 Methods.

of 340

Figure 4.1. Chapter Four’s Outline

Used to assist modeling factors associated with SSI

farmers’ engagement in mitigating strategies to reduce

transmission of WRZDs (Chapter 5) in the provinces

of Thai Binh and An Giang in Vietnam


What are sources and the basic microbial and related quality of

water used by SSI farmers for drinking and domestic purposes in

Thai Binh and An Giang provinces of Vietnam?

Factors associated with the microbial quality of

water used on SSI farms for drinking and domestic

purposes

Microbial

quality of water

used on SSI

farms for

drinking

(assessed using

Coliplate)

Microbial and

related quality of

water used on SSI

farms for drinking

and domestic

purposes (assessed

using laboratory

protocol)

Sources of water and

frequencies of use on

SSI farms

Chapter 4:

Presents the water quality among SSI farms in Thai Binh and An Giang provinces of Vietnam, including


How does the quality of water used by SSI farmers relate to their

perceptions of water quality and risk factors for transmission of

WRZD?

Socioeconomic of

SSI farms (from

Chapter 2)

Perceptions of SSI farmers

toward risk factors for

WRZD transmission (from

Chapter 3)

of 340

about on-farm water sources and frequencies of use. Drinking water was defined as “water used

for direct drinking or processing food,” and domestic water was defined as “water used for

domestic use but not for direct drinking or processing food” (MoH, 2009b; MoH, 2009a). I used

two protocols to assess the basic objective quality indicators for on-farm water for drinking and

domestic purposes (i.e., pH, turbidity, and E. coli). These protocols were 1) a protocol for

laboratories to assess on-farm water quality and 2) a protocol for community health/veterinary

workers (CHWS/CVWs) to quantify E. coli in on-farm water. These protocols followed WHO’s

recommendations for assessing the basic indicators of rural water quality (WHO, 2011b; WHO,

2011c; WHO and OECD, 2003b) and conformed with the Vietnamese national technical

regulations on drinking and domestic water quality to assess the basic indicators of the quality of

on-farm water (MoH, 2009b; MoH, 2009a). Detailed descriptions of these protocols are provided

in Appendix F2.

Descriptive and analytic statistics were used to analyze and present the results of the

water sources, frequencies of use, and objective assessments of on-farm water quality. Two sided

two-sample t-tests, one for drinking water and one for domestic water, using province

(DProvince) as a group variable were used to test the first hypothesis. In this hypothesis, the null

hypothesis indicated there were no differences between the mean frequencies of use of a water

source for drinking or domestic purposes between farmers in Thai Binh and An Giang (i.e., Ho:

diff = 0). The alternative hypothesis stated that the difference was not “0” (i.e., Ha: diff ≠ 0).

With respect to the second hypothesis, I used one-sided, one-sample t tests to compare the means

of pH, turbidity, and E. coli levels of water on farms with the standard levels of pH, turbidity,

of 340

and E. coli in drinking and domestic water set by the Vietnamese Ministry of Health (MoH,

2009b; MoH, 2009a).

Regression analysis (Dooho et al., 2009) was used to model and test the relationship

between the objectively assessed on-farm water quality and farmers’ perceived on-farm water

quality and risk factors for WRZD transmission (i.e. the third hypothesis). Only E. coli test

results, determined using standard laboratory protocols, were used for testing this hypothesis.

Descriptions of variables used in the regression analysis are presented in section 4.3.4.1. An

overall strategy and four specific strategies were used to model factors that are associated with

on-farm water quality.

The overall strategy for modeling associated factors of on-farm water quality was to start

with a Multiple Linear Regression (MLR) model then use a Seemingly Unrelated Regression

(SUR) model. The MLR model included a number of separate single regression equations in

which E. coli cfu in various sources of water for drinking and domestic purposes were dependent

variables. The SUR model included two regression equations simultaneously for E. coli cfu in

both drinking and domestic water. The purpose of using both MLR and SUR models was to

provide further insights about the complexity of on-farm water quality.

The first specific strategy was specifying variables for a maximum model. E. coli cfu in

each source of on-farm drinking water samples were specified as the dependent variables (DVs)

for the MLR models. E. coli cfu in both drinking and domestic water were specified as the DVs

for the SUR model. A full set of potential independent variables (IVs) was specified based on the

theoretical model (See Section 3.2, Chapter Three), the hypothesis, the literature for this

of 340

research, and my knowledge of on-farm water public health (See Section 3.4.1 for more detailed

description of IVs). The second specific strategy was to use two sets of considerations for

retaining IVs in the models. The first set of considerations was to include variables of my

primary interest in the research, variables known to be a potential confounder of primary factors

(e.g., farmers’ age), and evidence of change in the coefficients of primary factors when removing

a variable. The second set of considerations was to use p-values of the coefficients of individual

IVs and an adjusted R2 for the number of associated factors in the model to assess the model’s

goodness of fit. The third specific strategy was to fit models using various variable selecting

procedures (e.g., stepwise selection and forward selection). The final specific strategy was to use

statistical software (i.e., Stata 13) and my knowledge and experiences of water public health in

Vietnam to analyze data and interpret the results.

The MLR model and SUR model are written as follows and detailed descriptions of

model building strategies are presented in Appendix 4.6.3:

The MLR model:

Yi = β0+ βi Xi + ɛi; with i = 1, 2, 3... n

Yi is a continuous dependent variable representing on-farm water quality, considering

bundle i. Each Y represents the numbers of E .coli cfu in 100 ml of a source of on-farm

water.

Xi is a vector of IVs (continuous and/or dichotomous) believed to be relevant to the

consideration of ith equations (i = 1, 2, 3... n).

βo is a intercept or constant to be estimated,

of 340

βi is a vector of unknown coefficients describing the direction and magnitude of the effect

of an Xi on Yi after controlling for the effects of the other Xi; βi to be estimated,

ɛi is the error term

The SUR model:

The SUR model includes two equations for 1) E. coli cfu in on-farm drinking water

versus a set of potential IVs; and 2) E. coli cfu in on-farm domestic water versus another set of

IVs. The two equations are written below with j = 1… m MLR equations:

yj = x’j βj + uj

These two equations are stacked into an SUR model:

[

𝑦1..

𝑦𝑚

] = [𝑥1 0 ⋯ 0⋮ ⋱ ⋮0 ⋯ 𝑥𝑚

] [

β1..

βm

] + [

u1..

um

]

The error terms are assumed to have a zero mean and to be independent across farmers and

homoscedastic. For a given farm, the errors are correlated across equations:

E(uij uij’|𝑋) = σjj’ and σjj’ ≠ 0 where j ≠ j’

Cross-equation restrictions were tested, and constraints were imposed for IVs that were

identified in the fitting of both equations and that were not significantly different from each

other. The SUR model was then re-estimated.

4.3. Results:

4.3.1. Sources of water and frequencies of use in participating farms

SSI farmers participating in the research study in both provinces depended on multiple

sources of water for both drinking and domestic purposes (Table 4.1). Most farmers in Thai Binh

of 340

and An Giang used water from drilled well or river/canal respectively. Most farmers in An Giang

reported that they rarely or almost never used dug wells or drilled water wells as one of their

drinking water sources - 217/298 (73%) and 219/298 (73%) respectively. In Thai Binh, 143/300

(48%) of the farmers indicated that pond water or river/canal water were never or rarely used as

a source of drinking water. Farmers in both provinces also used combinations of different water

sources for domestic purposes. Among these, drilled wells and rivers/canals were frequently used

by 251/300 (84%) and 288/298 (97%) of the farmers in Thai Binh and An Giang respectively. In

both provinces, participating farmers rarely or never used water from bottles, pipes, or dug wells

for domestic purposes. The frequencies of use of water sources are statistically different between

two provinces. This supports one of the hypotheses tested in this chapter.

4.3.2. Results of on-farm water quality assessed using the Vietnamese national laboratory

protocol

Quality of on-farm water used for drinking purposes

Well water used for drinking on An Giang farms had a higher mean pH compared to

farms in Thai Binh (Table 4.2). The mean turbidity in rain and piped water on An Giang farms

was almost double that found in Thai Binh farms; however, mean numbers of E. coli cfu in An

Giang in those same two water sources were much lower than those in Thai Binh. Conversely,

mean E. coli cfu numbers in well water used for drinking in An Giang farms was about four

times higher than those on Thai Binh farms. For more results of the quality of drinking water, see

Appendix F 4.1.

Quality of on-farm water used for domestic purposes

of 340

There were slight differences between the pH levels of domestic water between the two

provinces. The mean turbidity levels in these sources of water in An Giang were two to three

times greater than in Thai Binh (Table 4.3), yet the mean numbers of E. coli cfu in well water in

both Thai Binh and An Giang were similar. Pond water and river water used for domestic

purposes in both Thai Binh and An Giang had much greater mean numbers of E. coli cfu

compared to well water. For more results about the quality of drinking and domestic water,

please see Appendix F.4.2.

4.3.3. Quality of on-farm water assessed by community health/veterinary workers

(CHW/CVWs)

This section presents the results of the prevalence of E. coli in on-farm drinking water by

CVWs in the province of Thai Binh and CHWs in the province of An Giang using ColiplateTM

tests. Seventy-five percent (224/299) and 81% (242/298) of water samples used for drinking

were positive in E. coli tests in participating farms in Thai Binh and An Giang respectively

(Table 4.4). Farmers in Thai Binh used stored rainwater, piped water, well water, and bottle

water for drinking; those in An Giang also used flocculated river water and piped water stored in

reservoirs (For more detailed information of the quality of on-farm water assessed by

CHWs/CVWs, please see Appendix F.4.3).

4.3.4. Factors associated with the auality of on-farm water

This section includes a description of DVs representing E. coli levels in individual and in

all water sources and the potential IVs that are associated with the DVs (Subsection 4.3.4.1).

Pairwise correlation among the variables used for modelling is provided in Appendix F.3.2.

of 340

Subsections 4.3.4.2 – 5 present the results of the best fittings of the regression model for each

DVs and the SUR model respectively.

4.3.4.1. Description of variables used for modelling

Table 4.5 shows the summary statistics of the dependent variables (DVs) used for

modelling. These DVs represent E. coli cfu in various on-farm water sources for drinking and

domestic purposes. Table 4.6 summarizes the possible IVs for modelling the DVs. On-farm

water quality may be associated with demographic factors (e.g., age), socioeconomic status (e.g.,

years of attending school), and subjective assessments of on-farm water quality (i.e., farmers’

perceptions of water quality). Most pairs of IVs were not significantly highly correlated (i.e.,

correlation coefficients greater than 0.5 with p-value< 0.05). These pairs of variables did not

essentially contain the same information (see Appendix F.3.2 for more details).

4.3.4.2.MLR Model - E. coli cfu in sources of water used for drinking (DV) vs. a set of IVs

Table 4.7 shows the results for the MLR regression analysis in which the number of E.

coli cfu in four sources of drinking water including rain, pipe, well, and river (i.e., the dependent

variables), each of which was regressed against a set of independent variables (IVs). The model

for E. coli cfu in stored rainwater for drinking was significant overall with p < 0.001 and

adjusted R2 = 0.17. Three out of ten variables were significant for the associated factors of E. coli

cfu in stored rainwater at p < 0.1. These significant associated factors included on-farm income

per person per year (Income; p < 0.09), fish production (Fish prod; p < 0.01), and satisfaction

with drinking water sources (DSatisfy-Drink, p < 0.01).

of 340

Table 4.1. Mean Frequency of Use for On-Farm Water Used for Drinking and Domestic Purposes in Both Provinces of Thai

Binh and An Giang, Vietnam

Province/Sources of

water

Mean frequency of use for drinking by source of water (a)

Drinking use Domestic use

Obs Mean Std. Dev. Obs Mean Std. Dev.

Thai Binh

Rain 196 3.4*** 1.3 256 2.5*** 0.9

Drilled well 231 3.2*** 1.3 251 3.5*** 1.1

Bottled 169 2.4 1.1 108 1.2*** 0.5

Pipe 61 2.4 1.4 52 1.8*** 0.9

Dug Well 40 1.6*** 1.0 38 1.9*** 1.1

Pond 143 1.0*** 0.2 209 2.9*** 1.4

River/Canal 143 1.0*** 0.2 251 3.2*** 1.2

An Giang

River/Canal 297 3.0 1.3 288 3.8 1.0

Pipe 243 2.6 1.7 245 1.3 0.9

Bottled 280 2.5 1.2 254 1.5 0.8

Rain 277 2.4 1.0 267 2.1 1.0

Pond 271 1.3 0.6 260 1.4 0.7

Drilled well 219 1.0 0.2 227 1.1 0.6

Dug Well 217 1.0 0.1 226 1.0 0.1 (a) 1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, to 5 = Very often

*** indicating values that are significant different between two provinces at p<0.01

of 340

Table 4.2. Comparison of Mean pH, Turbidity, and E. coli cfu of on-Farm Water Used for Drinking, Assessed Using Samples

Tested by National Laboratories, in the Provinces of Thai Binh and An Giang, Vietnam

Variable/Province Rain water Pipe water Well water e Bottled water

River water

(fl.)

pHa

Thai Binh 6.4** 6.4** 6.5** - -

An Giang 7.9** 7.5** 6.7** 9.4 7.0

Turbidity (measured in Nephelometric Turbidity Units) b

Thai Binh 0.8** 0.9* 3.6 - -

An Giang 1.3** 2.0* 3.9 0.8 7.6

E. coli (measured in number of E. coli colony-forming units per 100 mL of sample water)c

Thai Binh 61.8** 19.5** 27.0** - -

An Giang 11.1** 7.1** 107.1** 8.0 12.8

Frequency of used

Thai Binh 3.4 2.4 2.4 2.4 1.0

An Giang 2.4 2.7 1.0 2.5 3.0 a = means pH of all sources of water were not smaller than “6.5” nor significantly greater than “8.5” at p < 0.1

b = means of turbidity of well and river water were significantly greater than “2” at p < 0.01

c = means of E. coli in all sources of water were significantly greater than “0” at p<0.01

d = Frequency of use of the source water for drinking using (1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, and 5 = Very often).

e =both drilled well and dug well

** indicating values that are significantly different between the two provinces at p<0.05; * indicating values that are significantly different between the two

provinces at p<0.1

of 340

Table 4.3. Comparison of Mean pH, Turbidity, E. coli cfu of On-Farm Water Used for Domestic Purposes, Assessed Using

Samples Tested by National Laboratories, in the Provinces of Thai Binh and An Giang, Vietnam

Variable/Province Well water e Pond water River water

pH a

Thai Binh 7.1* 6.7** 7.1

An Giang 6.1* 7.0** 7.3

Turbidity (measured in Nephelometric Turbidity Units) b

Thai Binh 10.2 34.4*** 25.3*

An Giang 31.5 59.2*** 62.1*

E. coli (measured in number of E. coli colony-forming units per 100 mL of sample water) c

Thai Binh 356.5 2,494.7 2,757.5***

An Giang 371.4 2,950.1 780.7***

Frequency of used

Thai Binh 2.2 2.9 3.2 An Giang 1.1 1.4 3.8

a = means pH of all water were not significantly smaller than “6.0” nor greater than “8.5” at p < 0.1 b = means of turbidity of well and river water were significantly greater than “5” at p < 0.01

c = means of E. coli in all sources of water were significantly greater than “20” at p < 0.01 d = Frequency of use of the source water for domestic purposes (1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, to 5 = Very often). e = both drilled well and dug well

*** indicating values that are significantly different between the two provinces at p < 0.01; ** indicating values that are significantly different between the two

provinces at p < 0.05; * indicating values that are significantly different between the two provinces at p<0.1

of 340

Table 4.4. Prevalence of E. coli in On-Farm Water Tests (using ColiplateTM) Used for

Drinking by Source in Thai Binh and An Giang, Vietnam

Water

sources

Presence of E. coli in on-farm water for drinking

An Giang Thai Binh

Positive Total Prevalence Positive Total Prevalence

River with

flocculation 124 131 94.7 - - -

Rain 24 27 88.9 67 92 72.8

Pipe 61 82 74.4 19 36 57.8

Well 3 5 60.0 137 165 83.0

Bottle 19 33 57.6 1 6 16.7

Pipe (stored

in reservoir) 11 20 55.0 - - -

Total 242 298 81.2 224 299 74.9

Table 4.5. Summary Descriptive Statistics for the Dependent E. coli Variables Used in the

Regression Models

Variable

name

Description - Number of E.

coli cfu in: Obs Mean sd Iqr Max a

NE.coli

Drink-rain Rain water used for drinking 150 46.9 74.0 61 232

NE.coli

Drink-pipe Pipe water used for drinking 117 10.1 17.1 13 50

NEcoli

Drink-well Well water used for drinking 156 24.1 40.0 25 213

NE.coli

Drink- bottle

Bottled water used for

drinking 42 8.0 20.1 0 110

NE.coli

Drink-river River water used for drinking 117 12.8 30.5 2 260

NE.coli

Dom-well

Well water used for domestic

purposes 29 360.1 542.3 55 1,800

NE.coli

Dom-pond

Pond water used for

domestic purposes 353

2,596.

6

3,237.

5 1,200

12,40

0

NE.coli

Dom-river

River water used for

domestic use 213 817.8

1,077.

9 300 3,200

NE.coli

Drink

All sources of water used for

drinking 529 12.7 21.0 2 98

NE.coli Dom

All sources of water used for

domestic purposes 576

1,506.

3

1,994.

6 560

10,06

0 Notes: Dom = Domestic, N = Number, a Min values of these variables equal “0”

of 340

Table 4.6a. Summary Statistics for the Independent Variables (IVs) Considered for Use in Regression Models

Variable name Description N Mean Sd Min Max Count (1) % (1)

Age Age of the farmers 597 45.9 11.3 17 85 - -

DGender Gender of the farmers (Male = 1; Female =0) 598 - - - - 428 71.6

Schooling Years of attending school 598 6.9 3.2 0 18 - -

DIncome Income under poverty (Poor =1; No poor = 0) 590 - - - - 118 20.0

Children Number of household member under 18 years old (Yes

=1; No = 0) 598 - - - - 421 70.4

Experience Years of farming 598 9.6 8.3 0 50 - -

DOff-farm job Had off-farm jobs (Yes =1; No = 0) 593 - - - - 293 49.4

DFPR-Livestock-

Man

Man primarily responsible for livestock production (Yes

=1; No = 0) 589 - - - 244 41.4

DFPR-Livestock-

Woman

Woman primarily responsible for livestock production

(Yes =1; No = 0) 589 - - - - 91 15.5

DFPR-Fish-

Woman

Woman primarily responsible for fish production (Yes

=1; No = 0) 598 - - - 41 6.9

DFPR-Fish-Man

Man primarily responsible for fish production (Yes =1;

No = 0) 598 - - - - 187 31.3

DFPR-Health-Man

Man primarily responsible for family health production

(Yes =1; No = 0) 591 - - - 131 22.2

Province Thai Binh =1; An Giang = 0) 598 - - - - 300 50.2

DFish prod Having fish production (Yes =1; No = 0) 598 - - - 388 64.9

DRain for animals Using rain water for livestock production (Yes =1; No =

0) 598 - - - - 392 65.6

DPond for animals Using pond water for livestock production (Yes =1; No

= 0) 598 - - - 211 35.3

DPipe for animals Using pipe water for livestock production (Yes =1; No =

0) 598 - - - - 67 11.2

DMulti-agri

activities

Engage in more than three agricultural activities (Yes

=1; No = 0) 503 - - - - 436 86.7

DPoultry AI Poultry infected with AI in the past (Yes =1; No = 0) 598 - - - - 50 8.3

Npigs/yr Number of pigs produced on-farm per year 491 11.3 16.9 0 70 - -

Note: AI = Avian influenza, D = Dummy, FPR = Farmers’ perceived responsibility, and N = Number

of 340

Table 4.6b. Summary Statistics for the Independent Variables (IVs) Considered for Use in Regression Models (cont’d)

Variable name Description N Mean Sd Min Max Count (1) % (1)

Independent variables relating to farmers’ perceptions of:

DFPWater quality general water quality in the villages (Good=1; Not good = 0) 597 - - - - 486 81.4

DFPBarr-Cost cost as barriers of WRZD mitigating actions (Yes =1; No = 0) 598 - - - - 239 40.0

DFPRisk rain water

Risk of WRZD transmission from untreated rain water (Yes

=1; No = 0) 550 - - - - 517 94

DFPS-AI-Drink susceptibility to AI from UDW (Yes =1; No = 0) 496 - - - - 469 94.6

DFPS-Diarrhea-Drink susceptibility to diarrhea from UDW (Yes =1; No = 0) 559 - - - - 554 99.1

DFPS-Coliform-Drink susceptibility to coliform from UDW (Yes =1; No = 0) 213 - - - - 213 100.0

DFPS-Parasite-Drink susceptibility to parasites from UDW (Yes =1; No = 0) 404 - - - - 402 99.5

DFPS-AI-Dom susceptibility to AI from UDomW (Low =1; High = 0) 497 - - - - 289 58.2

DFPS-Coliform-Dom susceptibility to coliform from UDomW (Low =1; High = 0) 213 - - - - 146 68.5

DFPS-Diarrhea-Dom susceptibility to diarrhea from UDomW (Low =1; High = 0) 561 - - - - 354 63.1

DFPS-Parasite-Dom Susceptibility to parasites from UDomW (Low =1; High = 0) 405 - - - - 293 72.4

DSatisfy-Drink satisfaction with source of drinking water (Low =1; High = 0) 596 - - - - 244 40.9

DSatisfy-Dom satisfaction with domestic water (Low =1; High = 0) 580 - - - - 398 68.6

DAI heard heard of AI or not (Yes =1; No = 0) 598 - - - - 541 90.5

DFPHarm-Dom harm of UDomW (Low =1; High = 0) 576 - - - - 332 57.6

DFPHarm-Drink harm of UDW (Yes =1; No = 0) 574 - - - - 565 98.4

Other Independent variables:

DDump Waste

Dump waste of livestock production to sources of water for

domestic (Yes =1; No = 0) 547 - - - - 175 32.0

DFE-Livestock

Management Engage in livestock management (Yes =1; No = 0) 598 - - - - 329 55.0

DTest-dom Have tested domestic water in the past (Yes =1; No = 0) 559 - - - - 40 7.2

Notes: AI = Avian influenza, D = Dummy, Dom = Domestic, FP = Farmers’ perceptions of, FPBarr = Farmers’ perceived barriers to taking mitigating

actions, FPS = Farmers’ perceived susceptibility to, UDW = untreated drinking water, UDomW = untreated domestic water

of 340

The model for E. coli cfu in pipe water was statistically significant with p < 0.01 and

adjusted R2 = 0.20. Two variables were significantly associated with of E. coli in piped drinking

water including the number of chickens produced (Npoultry/yr, positive, p < 0.05) and the

number of ducks produced (Nducks/yr, p < 0.01). The most interesting model fitting for E. coli

in well water was significant at p < 0.01 and adjusted R2 = 0.35. This model included three

variables that were significantly associated with E. coli cfu in well water used for drinking:

income (DIncome, p<0.05), fish production (DFish prod, p<0.01), and perceived susceptibility to

parasites from using untreated drinking water (DFPS-Parasite-Drink, p < 0.1). The model for

associated factors of E. coli cfu in river water was significant overall with p < 0.05 and adjusted

R2 = 0.07. Three variables were significantly associated with E. coli cfu in river water including

gender (DGender, positive, p < 0.1) and the number of chickens produced (Nducks/yr, p < 0.01)

(Table 4.7).

4.3.4.3. MLR Model - E. coli cfu in water sources for domestic use (DV) vs. a set of IVs

In Table 4.8, the model fitting results of E. coli cfu in pond water for domestic purposes

was significant overall with p < 0.01 and adjusted R2 = 0.05. Three variables were significant

associated factors of E. coli cfu in pond water, specifically: education (Schooling, p < 0.05),

perceived susceptibility to HPAI when using untreated domestic water (DFPS-AI-Dom, p < 0.1),

and satisfaction with current sources of domestic water (DSatisfy-Dom, p < 0.1). The model

fitting results of E. coli cfu in river water for domestic purposes was significant overall with p <

0.01 and adjusted R2 = 0.12. Two variables were positive significantly associated factors of E.

coli cfu in river, specifically the following: occupation other than farming (DOff-farm job, p <

0.1) and number of pigs produced (Npigs/yr, p < 0.01). Three factors negatively and significantly

of 340

associated with E. coli in river water for domestic purposes were gender (DGender, p < 0.1),

male farmers playing major roles in livestock production (DFPR-Livestock-Man, p < 0.05), and

perceived susceptibility to parasites from using untreated domestic water (DFPS-Parasite-Dom, p

< 0.01).

4.3.4.4.SUR Model - E. coli cfu in sources of on-farm water used for domestic purposes

Tables 4.9a-c show a SUR model fit of four associated factors of the level of E. coli cfu

in sources of on-farm drinking water in which one variable was significant. This variable was the

perceived level of harm from untreated drinking water (DFPHarm-Drink, negative, p < 0.05).

With respect to E. coli cfu in sources of on-farm domestic water, a more complex model fit was

found with 25 associated factors of E. coli cfu in sources of on-farm domestic water. Only four

factors were significantly associated with the level of E. coli cfu in sources of on-farm domestic

water. These four factors consisted of the following: number of years attended school

(Schooling, p < 0.01), production of fish (DFish prod, p < 0.05), ever had poultry infected with

bird flu (DPoultry AI, p < 0.01), and has heard of bird flu (DAI heard, p < 0.1). Overall, the SUR

model is significant with the p = 0.03 for the Breusch-Pagan test (Table 4.9b).

of 340

Table 4.7a. Summary of the Best MLR Model Fittings for E. coli cfu in Each of the Four Sources of Drinking Water in the


Independent variables a Rain Pipe Well River

Coef P>|t| Coef P>|t| Coef P>|t| Coef P>|t|

DGender - - -5.64 0.19 -14.64 0.11 14.12* 0.05

Age 0.76 0.17 0.19 0.32 - - - -

Schooling -2.86 0.23 -0.69 0.26 -1.96 0.17 -1.61* 0.07

Experience -0.89 0.22 0.28 0.36 - - - -

DIncome -34.52* 0.09 6.86 0.11 37.48** 0.03 8.55 0.16

Children - - 6.67 0.14 - - - -

DOff-farm job - - 1.03 0.81 -9.85 0.22 - -

Npoultry/yr - - 0.27** 0.03 - - 0.75** 0.01

Nducks/yr - - 0.08*** 0.00 - - -0.02 0.15

Npigs/yr - - 0.10 0.53 - - - -

DFish prod 50.07*** 0.00 - - -70.33*** 0.00 -

DFPR-Livestock-Man - -1.28 0.78 - - - -

DFPR-Livestock-Woman 17.93 0.24 -2.12 0.66 17.83 0.10 - -

DFPWater quality - - -0.52 0.91 - - - -

DFPHarm-Drink 20.33 0.20 6.11 0.21 - - - -

No. of significant variables 3 - 2 - 3 - 3 - Notes: AI = Avian influenza, D = Dummy, FP = Farmers’ perceptions of, FPR = Farmers’ perceived responsibility, FPS = Farmers’ perceived susceptibility to, N

= Number, a Independent Variables described in Table 4.6; (*) significant at p<0.1; (**) significant at p<0.05; (***) significant at p<0.01

of 340

Table 4.7b. Summary of the Best MLR Model Fittings for E. coli cfu in Each of the Four Sources of Drinking Water in the

Provinces of Thai Binh and An Giang, Vietnam (cont’d)

Independent variables a Rain Pipe Well River


DFPS-AI-Drink - - 7.27 0.11 -12.48 0.16 - -

DFPS-Diarrhea-Drink -23.75 0.18 -0.94 0.86 16.19 0.10 - -

DFPS-Coliform-Drink -19.00 0.20 0.77 0.88 - - - -

DFPS-Parasite-Drink - - -7.59 0.10 -17.41* 0.07 - -

DSatisfy-Drink -47.07*** 0.00 2.34 0.57 - - -

constant 28.34 0.43 -12.80 0.36 129.68*** 0.00 1.79 0.82

Obs 135 - 90 - 138 - 111 -

Prob > F 0.00*** - 0.01** - 0.00*** - 0.03** -

R-squared 0.24 - 0.37 - 0.39 - 0.11 -

Adj R-squared 0.17 - 0.20 - 0.35 - 0.07 -

Root MSE 66.82 - 14.16 - 44.24 - 30.15 -

No. of significant variables 3 - 2 - 3 - 3 - Notes: AI = Avian influenza, D = Dummy, FP = Farmers’ perceptions of, FPR = Farmers’ perceived responsibility, FPS = Farmers’ perceived susceptibility to, N

= Number, a Independent Variables described in Table 4.6; (*) significant at p<0.1; (**) significant at p<0.05; (***) significant at p<0.01

of 340

Table 4.8. Summary of the MLR Best Model Fittings for E. coli in Pond Water and River

Water for Domestic Purposes in the Provinces of Thai Binh and An Giang, Vietnam

Independent variables a

Pond River

Coef P>|t| Coef P>|t|

DGender - - -338.80* 0.08

Age - - 11.06 0.11

Schooling 213.66** 0.02 - -

Experience - - - -

DIncome - - -264.47 0.10

Children - - - -

DOff-farm job - - 264.33* 0.08

Npoultry/yr -17.79 0.10 - -

Nducks/yr -1.03 0.20 - -

Npigs/yr 14.78 0.11 36.97** 0.01

DFish prod - -

DFPR-Livestock-Man - - -368.40** 0.03

DFPR-Livestock-Woman - - - -

DFPWater quality - - - -

DFPHarm-Dom - - -278.42 0.10

DFPS-AI-Dom 1,044.65* 0.05 279.94 0.16

DFPS-Diarrhea-Dom - - - -

DFPS-Coliform-Dom - - - -

DFPS-Parasite-Dom - - -482.03** 0.01

DSatisfy-Dom 1,028.56* 0.09 - -

constant 210.21 0.81 824.62** 0.03

Obs 197 - 188 -

Prob > F 0.02** - 0.00*** -

R-squared 0.08 - 0.16 -

Adj R-squared 0.05 - 0.12 -

Root MSE 3,433.80 - 990.43 -

No. of significant variables 3 - 5 - Notes: AI = Avian influenza, D = Dummy, Dom = Domestic, FP = Farmers’ perceptions of, FPR =

Farmers’ perceived responsibility, FPS = Farmers’ perceived susceptibility to, a Independent Variables

described in Table 4.6; (*) significant at p<0.1; (**) significant at p<0.05; (***) significant at p<0.01

of 340

Table 4.9a. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints

[NE.coli Drink]Nducks/yr - [NE.coli Dom]Nducks/yr = 0

chi2( 1) = 1.52; Prob > chi2 = 0.22

Iteration 1: tolerance = 4.18

(…results of iteration 2-5 omitted)


Seemingly unrelated regression, iterated

Equation Obs Parms RMSE R-sq chi2 P

NE.coli Drink 243 4 14123.37 0.03 6.74 0.15

NE.coli Dom 243 26 2761.33 0.17 50.29 0.00 Notes: N = Number, Dom = Domestic

Table 4.9b. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints (cont’d)

Correlation matrix of residuals:

NE.coli Drink NE.coli Dom

NE.coli Drink 1

NE.coli Dom 0.14 1

Breusch-Pagan test of independence: chi2(1) = 4.667, Pr = 0.03 Notes: N = Number, Dom = Domestic

of 340

Table 4.9c. SUR Model Estimation of the Two Equations of E. coli cfu in Water Used for

Drinking with Cross-Equation Constraints (cont’d)

Coef. Std. Err z P>z [95% Conf. Interval]

NE.coli Drink

DOff-farm job 2,984.05 1,873.74 1.59 0.11 -688.41 6,656.50

Nducks/yr 0.13 0.98 0.13 0.89 -1.80 2.06

DSatisfy-Drink -1,362.50 1,841.32 -0.74 0.46 -4,971.42 2,246.42

DFPHarm-Drink -4,640.21** 2,020.48 -2.30 0.02 -8,600.27 -680.15

_cons 3,334.69 1,966.01 1.70 0.09 -518.61 7,187.99

NE.coli Dom

Age -0.44 19.03 -0.02 0.98 -37.73 36.84

DGender 155.06 493.96 0.31 0.75 -813.09 1,123.20

Schooling 195.10*** 70.11 2.78 0.01 57.68 332.52

Children -354.82 457.96 -0.77 0.44 -1,252.41 542.76

Experience -44.13 29.32 -1.50 0.13 -101.61 13.34

DFPR-Livestock-Woman -947.35 737.80 -1.28 0.20 -2,393.41 498.72

DFPR-Livestock-Man -991.06 543.12 -1.82 0.07 -2,055.56 73.44

DFPR-Fish-Woman -586.72 930.27 -0.63 0.53 -2,410.00 1,236.57

DFPR-Fish-Man 313.86 596.17 0.53 0.60 -854.60 1,482.33

DFPR-Health-Man 291.22 646.41 0.45 0.65 -975.71 1,558.15

DProvince 19.02 697.20 0.03 0.98 -1,347.46 1,385.51

DFish prod 1,210.97** 611.32 1.98 0.05 12.80 2,409.14

DRain for animals 200.38 548.55 0.37 0.72 -874.76 1,275.51

DPond for animals 3.86 451.08 0.01 0.99 -880.25 887.96

DPipe for animals -368.21 687.67 -0.54 0.59 -1,716.01 979.59

DMulti-agri activities -523.38 640.58 -0.82 0.41 -1,778.90 732.14

DPoultry AI 2297.35*** 848.80 2.71 0.01 633.73 3,960.98

Nducks/yr 0.13 0.98 0.13 0.89 -1.80 2.06

Npoultry/yr -5.63 7.98 -0.71 0.48 -21.27 10.01

DFPWater quality -714.74 523.78 -1.36 0.17 -1,741.33 311.85

DFPBarr-Cost 251.82 479.81 0.52 0.60 -688.60 1,192.24

DFPRisk rain water -1,261.25 1,153.12 -1.09 0.27 -3,521.33 998.83

DFPS-Parasite-Dom 57.65 586.64 0.10 0.92 -1,092.14 1207.43

DAI heard -2,676.05* 1,693.47 -1.58 0.11 -5995.18 643.08

DFPHarm-Dom 490.64 493.53 0.99 0.32 -476.66 1457.94

DFE-Livestock Management -358.27 548.16 -0.65 0.51 -1432.65 716.11

_cons 5,502.15*** 2,344.63 2.35 0.02 906.76 10,097.55

Notes: AI = Avian influenza, D = Dummy, Dom = Domestic, FP = Farmers’ perceptions of, FPBarr = Farmers’

perceived barriers to taking mitigating actions, FPR = Farmers’ perceived responsibility, FPS = Farmers’

perceived susceptibility to, * significant at p < 0.05; ** significant at p < 0.1

of 340

4.4. Discussion

4.4.1. Sources and frequency of use of on-farm water for drinking and domestic purposes

Farmers had to depend on various sources of water for drinking and domestic purposes.

This result is consistent with other reports on sources of water used in rural areas in Vietnam.

These sources are pond/lake, river/stream/canal, rainwater collection, piped water, hand dug

wells, drilled wells, and bottled water (MoC and MARD, 2000; MoH, 2009b; MoH, 2009a).

Farm observations (visits) and in-depth interviews of the participating farmers supported this

result.

The percentages of farmers who had access to improved sources38 of water for drinking

were 96% in Thai Binh and 99% in An Giang. These percentages were slightly greater than the

World Bank published Vietnamese national percentage of 95% of the rural population having

access to improved water. Moreover, although farmers in the research study used multiple

sources of water for drinking, their frequencies of use were varied among sources of water. This

variation needs to be considered when developing rural water interventions in the studied areas.

In the following paragraphs, I further discuss by province the various sources of on-farm water

used.

In Thai Binh, observations of farm settings and in-depth interviews showed that rain

water was considered a precious high quality source of water. Compared to other sources of

water, stored rainwater was the most used source by farmers for drinking water. Although the

number of Thai Binh farms that reported using drilled well water for drinking was highest

compared to other sources of water, drilled well water was second to stored rainwater in terms of

38 “improved sources of water” refers to pipe water into dwelling, pipe water to yard/plot, public pipe or standpipe,

tube well or borehole, protected dug well, and rainwater (JMP, 2014)

of 340

frequency of use. Therefore, rural water interventions should prioritize stored rainwater and

drilled well water as most Thai Binh farmers frequently used these sources. Moreover, they did

not use dug wells very often, and farmers indicated that it was becoming harder and harder to dig

a well that had good quality water. Even for drilled wells, the depth of the wells should be about

70 meters to reach underground water, which is quite deep. The other reason participating

farmers did not favour dug wells was the low quality of water found in them (e.g., bad smelling

water was perceived as being low quality water) if they did not dig deeper wells. Farmers were

also afraid that dug wells were likely to be contaminated with pollution from livestock waste and

pesticides.

In An Giang, on the other hand, the most frequently used source for drinking water was

river/canal water. This reflects the more comprehensive network of river branches and canals in

the Mekong River Delta in An Giang compared to the Red River Delta in Thai Binh (Nguyen,

2007). Specifically, the number of SSI farmers in An Giang who had piped water for drinking

was four times greater versus those in Thai Binh. However, these SSI farmers rarely used water

services due to their economic barriers (i.e. not being able to pay for connecting to the piped

water network or for the monthly fees), as well as the availability of other sources for on-farm

drinking water. Another study also showed limited use of piped water by rural households in the

Mekong Delta due to similar reasons (Wilbers et al., 2014) . In contrast, dug wells and drilled

wells were still common in An Giang with more than two-thirds of the participating research

study farmers reporting having access to and use of both dug and drilled wells for drinking

water. However, the mean frequencies of use for drinking water from either dug or drilled wells

were “rarely”. Consequently, water interventions in An Giang may need to focus on solutions for

using river/canal water for drinking in the short and medium term. In the long term, more

of 340

research is needed to explore how to improve accessibility and affordability of piped water

supply networks for rural farmers.

Water sources used for domestic purposes were similar to those used for drinking water

in both provinces. However, the frequencies of using these sources of water for domestic

purposes were different between the two provinces. For example, drilled well water was the most

frequently used water source for domestic purposes in Thai Binh. Meanwhile, in An Giang,

farmers used river water the most for their domestic purposes. In-depth interviews with farmers

revealed that when the same water source was used for both drinking and domestic purposes,

treatment methods such as boiling water and/or flocculating water were used for drinking water

but not for domestic use water. In addition, given the limited supply of on-farm water, it is

understandable that rain water, piped water, and bottled water were not selected by the

participating farmers as the most frequently used sources of water for domestic purposes. In Thai

Binh, the common and most frequently used source of water for domestic purposes was different

from that used for drinking water (i.e., stored rainwater and drilled well water respectively).

Meanwhile, in An Giang, water from rivers/canals was the most common and frequently used

source of water for both drinking and domestic purposes.

The differences between the two provinces in frequencies of use of all water sources for

drinking and domestic purposes were significant (except for bottle and piped water for drinking),

which confirmed the first hypothesis of this research. The hypothesis indicates that frequencies

of use of water sources for drinking and domestic purposes are different between farmers in Thai

Binh and An Giang. For that reason, such differences need to be considered in designing water

related interventions. For example, the mean frequencies of using water from rain, pipes, and

bottles by farmers of both Thai Binh and An Giang ranged between rarely used and frequently

of 340

used. On the other hand, Thai Binh participating farmers did not use water from ponds, lakes,

rivers and canals for drinking purposes, while participating farmers in An Giang did use these

drinking water sources although rarely (e.g., for pond water) and occasionally (e.g., for river

water). Participating farmers in An Giang did not use water from hand-dug wells and drilled

wells for drinking water. However, participating farmers in Thai Binh rarely and occasionally

used water from hand-dug wells and drilled wells respectively.

4.4.2. Low quality of on-farm water for both drinking and domestic purposes

Compared to the basic indicators (i.e., pH, turbidity, and E. coli) of drinking water quality

set by the WHO and the Vietnamese MOH, the quality of water on participating farms was low.

The water samples from rain, pipe, wells, and rivers (flocculated) had acceptable levels of pH

(i.e., pH within 6.5 to 8.5). Water samples from bottles had a high pH mean level (i.e., mean pH

of 9.4). Bottled water, stored rainwater, and pipe water had turbidity levels that met the national

standard level of turbidity for drinking water (i.e., maximum turbidity of 2 NTUs). However,

well water and flocculated river water exceeded the maximum turbidity level for drinking water

(i.e., 3.7 NTUs and 7.6 NTUs respectively). Levels of pH and turbidity in water have been

shown to influence the microbial quality of water (WHO and OECD, 2003c). The WHO

recommends that E. coli be used as an indicator organism to assess the microbial safety of rural

water and drinking water should be free from E. coli (WHO and OECD, 2003b). E. coli is a well-

defined member of the Enterobacteriaceae family. There are four main reasons to use E. coli as

an indicator organism. First, it is not practical to conduct isolations of all pathogens. Second,

pathogens from human and animal faeces pose the greatest danger to public health and detecting

faecal contamination in drinking water is important for public health safety. Third, methods of

assessing E. coli are simpler, more available, and more affordable when compared to those of

of 340

other microbial agents including potentially pathogenic agents. Fourth, E. coli is a more specific

indicator of faecal contamination compared to other members of the total coliform group. The

presence of E. coli in water indicates contamination with faeces and possible other pathogens in

the water.

Now, let me focus my discussion on the E. coli levels in sources of on-farm drinking

water. Almost 60% of the water samples were greater than the maximum level of E. coli cfu

indicated in the drinking water quality guidelines of Vietnam, Canada, and the WHO (FPTCW,

2013; MoH, 2009b; WHO, 2012). Piped water and bottled water supplied by either local

governments or private companies were authorized and monitored by the local health authorities.

Water from pipes and bottles were expected to meet the national standards of drinking water

quality. However, in this research, it was clear that piped and bottled water used on at least one-

third of the SSI farms did not meet the standards of E. coli (i.e., 30.95% and 41.03% of the

bottled and pipe water samples respectively had E. coli). Another study of piped water in rural

areas of the Mekong Delta also found that E. coli was commonly detected in piped water

(Wilbers et al., 2014). None of the results of the water tests was in conformity with the standards

of drinking water quality set by both the WHO and the Vietnamese MoH. The level of E. coli cfu

in bottled water and piped water used by the participating farmers for drinking can be

classified/scored as an intermediate risk with low action priority (i.e., 1-10 E. coli cfu/100ml)

(WHO, 2012). Meanwhile, the level of E. coli cfu in stored rainwater, well water, and river water

used by the participating farmers can be classified/scored as being a high risk. These

classifications should be combined with corresponding sanitation inspection scores to completely

analyze the risk of water on-farm (WHO, 2012).

of 340

In terms of the quality of on-farm water for domestic purposes, the mean pH level of on-

farm domestic water met the national pH standards for domestic water. However, the mean

turbidity and E. coli levels in the sources of on-farm domestic water were much greater than

those in the national standards. Moreover, the mean level of E. coli cfu in river water used for

domestic purposes in Thai Binh was almost four times greater than that in An Giang. This

difference may be one of the reasons why farmers in Thai Binh use river water frequently for

domestic purposes and almost never for drinking water; meanwhile, farmers in An Giang

appeared to use river water frequently for both drinking water and for domestic purposes. The

microbial qualities of water for both drinking water and domestic purposes were significantly

lower than the national standards of water quality and were significantly different between the

two provinces. This difference and the low microbial quality of water on-farm needs to be

considered when designing interventions and/or further research about on-farm water.

Concerning the microbial quality of on-farm drinking water tested by the CHWs/CVWs

using ColiplateTM tests, the percentage of water samples with the presence of E. coli tested by

CHWs/CVWs were greater than those results done by the laboratories. Although, the ColiplateTM

test results produced by CHWs/CVWs supported the two laboratories’ results of E. coli cfu in

water, only the results of water quality tested using laboratory protocol were used to confirm the

second hypothesis of this research. Some factors that need to be considered when explaining the

differences between the level of E. coli measured by CHWs/CVWs and the laboratories are

discussed in the following paragraphs.

First, ColiplateTM tests were highly sensitive and could indicate the presence of as small

as only one E. coli in water samples. Second, the mean time to testing was much lower for the

ColiplateTM test than for the laboratory tests (i.e., 30 to 45 minutes and 3 to 6 hours respectively).

of 340

Third, CHWs/CVWs incubated water samples using simple tools available on farms (e.g.,

Styrofoam box, light bulbs, and insulating materials such as used clothing) and the incubations

depended on the weather; this explained the actual longer time it took to incubate water samples

in Thai Binh (colder weather) compared to An Giang (warmer weather). Fourth, technical

persons at the laboratories were to dilute a water sample before incubating it if they thought that

there might be a high level of E. coli cfu in the water sample (based on observing the water

sample and reading notes taken during the water sample collection).

However, we did not train CHWs/CVWs to use the dilution technique due to limited time

and resources for the survey as well as possible contamination of the water samples.

Furthermore, the dilution technique was not addressed in the guidelines of the ColiplateTM testing

procedure. Thus, introducing the dilution technique may enhance the limits of ColiplateTM tests

in the future. Finally, although a ColiplateTM test used a simple procedure and CHWs/CVWs

were trained, most of the CHWs/CVWs had never tested water quality before this research.

Meanwhile, technicians from the national laboratories were highly skilled and experienced in

water testing. Given this, the results of water tests by the laboratory were used for further

analysis (i.e., exploring associated factors of on-farm water quality).

The results of water tests using ColiplateTM test kits were also used for confirming the

water tests using Watercheck (i.e., water tests performed by farmers in an extended research, not

reported in this research), and for studying possible solutions (in another research) to address on-

farm water public health (e.g., ownership, participation, awareness, and behaviour change of

farmers, and community stakeholders).

of 340

4.4.3. Associated factors of on-farm water quality

Farm visits during the field study of this research indicated that on-farm water issues

appeared complex and there was a potential correlation between the quality of both drinking and

domestic water within each farm. Farmers used one source of water for multiple purposes and/or

multiple sources of water for one purpose. Furthermore, the dependent variables of interest in

this chapter (e.g., the number of E. coli cfu in on-farm water) have a continuous nature and the

hypothesis of the research in this chapter addressed these associations. Therefore, regression

analysis was an appropriate technique to explore factors that are associated with the level of E.

coli contamination in on farm water for drinking.

Farmers’ income (DIncome)39 was associated with number of E. coli cfu in two out of the

four sources of on-farm water for drinking. Although there are no other studies that look at

exactly the same associations to compare with this interesting finding, some researchers did

indicate the possibly important role economic factors play in rural public health (Duong et al.,

2004). When controlling for other variables in each fitting of the model, two variables that were

negatively associated with the level of E. coli cfu in stored rainwater were income and

satisfaction with the quality of the current sources of drinking water. This indicates that, among

farmers who used stored rainwater for drinking, a poor farmer (i.e., coded as 1 if income is less

than the poverty line) was likely to have a better microbial quality of stored rainwater (i.e., fewer

E. coli cfu counts) than a not poor farmer (i.e., coded as 0 if income is greater than the poverty

line). In addition, if a farmer was satisfied with the quality of stored rainwater (i.e., moving from

39 The variable of income per person per year was dichotomized using the cut-off income level for poverty set by the

Vietnamese government (i.e., less than 4.8 million VND per person per year – equivalent to $244 CAD per person

per year) at the exchange rate of $ CAD = 19,700 on 5th December 2013. An income per person per year below the

cut-off income level (i.e., a poor farm) was coded as 1 and 0 if otherwise (i.e., a not poor farm). I should have named

this variable as DPoverty instead of DIncome to avoid confusion.

of 340

code 0 if not satisfied to code 1 if satisfied), the farmer was likely to have better microbial

quality of stored rainwater. From my farm visits, this indication makes sense as I saw poor

farmers often depended on and cared for stored rainwater more than not poor farmers (who may

have more sources of water for drinking other than stored rainwater) did. Furthermore, farmers

were confident in assessing the quality of their stored rainwater and in indicating their levels of

satisfactions with stored rainwater. Having fish production on farms was also a positively

associated factor for the level of E. coli contamination in rain and well water for drinking. This

association may be explained by the reported inappropriate practice of dumping livestock waste

directly into fishponds.

The number of chickens and ducks on farms were additional positive significantly

associated factors of the level of E. coli contamination in piped water. This means farms with

more chickens and ducks were likely to have worse piped water (i.e., higher E. coli cfu in pipe

water) compared to those farms that have fewer chicken and ducks. Incomes on farm per year

and perceptions of the level of susceptibility to parasites from using untreated drinking water

were also associated with the increased E. coli contamination of well water for drinking

purposes. This means that farmers with a poor yearly income who thought that using untreated

drinking water gave them susceptibility to parasites tended to have higher E. coli contamination

in their well water used for drinking. Farmers who attended more years of school tended to have

better on-farm water quality (i.e., lower E. coli cfu in river water for drinking) compared to

farmers who had attended school for fewer years. On the other hand, female-headed farms with

more numbers of chickens had higher levels of E. coli cfu contamination in river water compared

to male-headed farms with fewer chickens.

of 340

Yet, farms headed by men which had male farmers primarily responsible for livestock

production tended to have more E. coli contamination of river water used for domestic purposes

compared to farms that were headed by women which had female farmers primarily responsible

for livestock production. Furthermore, having more pigs and an occupation other than farming

and perceived increased susceptibility to parasites from using untreated river water for domestic

purposes tended to increase the E. coli contamination of river water used for domestic purposes.

Farmers who had more years of attending school, perceived susceptibility to HPAI from using

untreated domestic water, and who were satisfied with their current sources of domestic water

tended to have higher levels of E. coli contamination in pond water used for domestic purposes.

The SUR model with cross-equation constraints showed an improvement in the analysis

of the relationship between independents variables and E. coli cfu in drinking and domestic

water sources in the two equations. E. coli cfu in drinking water sources with cross-equation

constraints had the same number of IVs as the associated factors as the model fit without cross-

equation constraints. Farmers having other occupations than farming were likely to have more E.

coli cfu in drinking water sources. Meanwhile, farmers who thought there was harm in using

untreated sources of drinking water were likely to have less E. coli contaminated water compared

to those who did not think there is harm in using untreated drinking water. Improvements of the

SUR model with cross-equation constraints in fitting the levels of E. coli cfu in domestic water

support my expectation of the complexity and a potential correlation between the quality of on

farm water for drinking and for domestic uses. Years of attending school, having fish production,

and past experience with AI in poultry were positive significant associated factors. Meanwhile,

the variables of male farmers primarily responsible for livestock production and having heard of

AI were negative significant associated factors to the levels of E. coli cfu in domestic water.

of 340

The results of the SUR model have confirmed the last part of my research hypothesis that

the microbial quality of on-farm water is associated with farmers’ demographics, socioeconomic

status, and their perceptions of risk factors for WRZD transmission. I believe this association

may happen because farmers with different socioeconomic status may have different access to

improved sources of water (e.g. using river water instead of pipe water for drinking).

Furthermore, they may have different levels of resources to protect their sources of water, and

properly treat, distribute and store their water (e.g., insufficient land areas to separate livestock

housing from drinking water sources). Therefore, different demographic and socioeconomic

status may lead to different quality of water on farms. Furthermore, if a farmer does not perceive

that there are risk factors for WRZD transmission, the farmer may not see a need to take actions

to improve the quality of on farm water. The results of the SUR model imply the importance of

considering farmers’ demographics, socioeconomic status, and perceptions to improve the

microbial quality of water on farms.

My research used a cross-sectional design in which on-farm water was assessed at one

point in time. However, water quality may vary over time. For example, in developing countries,

the fecal contamination of drinking water appears to be greater during the wet season (Costyla et

al., 2015). Therefore, in addition to the associated factors of on-farm water quality discussed

above, future studies may be needed to assess contributions of seasonality to water quality on

SSI farms.

4.4.4. Rural water and its regulatory framework on SSI farms in Vietnam

Considering the results and discussion of the multiple sources of low quality on-farm

water and the insights of possible associated factors for on-farm water quality presented in the

previous sections, this section will discuss issues of on-farm water public health and its

of 340

corresponding policy framework in Vietnam. The section will provide a brief overview and

discussion of the regulatory framework of rural water supply and sanitation globally and in

Vietnam, focusing on its impact on humans, animals, and their environment at the SSI farming

level.

Globally, drinking water coverage in 2011 is 1% above the Millennium Development

Goal (MDG)’s target. However, it is estimated that 768 million people still relied on unimproved

drinking water resources and the world does not meet the MDG’s sanitation target by half a

billion people (WHO and UNICEF, 2013). Target 7.C of the MDG aims at reducing by 50% the

proportion of people who do not have access to safe drinking water and basic sanitation (UN,

2013). The proportion of the Vietnamese population using improved sources of drinking water

and improved sanitation is estimated to be 76 - 90% and 50 – 75% respectively (WHO and

UNICEF, 2013). However, rural farmers have limited awareness of the water and environmental

sanitation issue with 70% of them using contaminated water and having no latrines (MoC, 2000).

Only 9% of the rural population uses piped-water on their premises compared to 58% of the

urban population.

Regulatory framework and the National Rural Clean Water Supply and Sanitation Strategy up to

2020 (the strategy)

The evolution of water-related policies in Vietnam began decades ago. The long-term

overall framework for the development of rural water supply and sanitation in Vietnam is driven

by a number of policies in the National Rural Clean Water Supply and Sanitation Strategy up to

2020 (MoC, 2000). However, there has been limited progress and many challenges still exist

with room for improvement, especially relating to engaging multiple stakeholders and

empowering the rural community to ensure a clean water supply and adequate sanitation. Many

of 340

other countries also face similar challenges and it is suggested that future water-related policies

should recognize and support the role of multiple stakeholders (Lockwood and Smits, 2011;

Ravnborg and Jensen, 2012)

The strategy is one of the key national policies that directly influence SSI famers.

Although this strategy involved multiple stakeholders in its design and development stages, the

level of stakeholder consultation for the strategy phase was not comprehensive and the

participation of grassroots stakeholders (i.e., farmers) was limited. Moreover, although livestock

production is part of the strategic goals, veterinary services were not engaged, nor were linkages

between water and livestock production explicitly addressed in the strategy. At the same time,

research studies have shown that SSI farmers and their animals are at high risk of WRZD and

that water-related policies have direct important impacts on human and animal health, and on

sustainable agriculture.

Therefore, the regulatory framework and the strategy need to be reviewed using trans-

disciplinary approaches (e.g., an EcoHealth approach) to define clearer roles and responsibilities

and to be more inclusive of stakeholders. The strategy should address rural water public health

issues more explicitly to not only enhance safety for the rural water supply, but to maximize use

of the limited water resources to improve the health of humans, animals, and their environment.

It has been recommended that the Vietnamese government consider values of incorporating an

EcoHealth approach to revising, formulating, and implementing government policies addressing

SSI farming and water public health (Hall and Dinh, 2009; Hall and Le, 2013; Gleeson et al.,

2011; Le and Hall, 2011b; Le and Hall, 2011a). It is a basic human right for each person to have

access to safe water, sanitation, and hygiene at home. Moreover, water, sanitation, and hygiene

of 340

should be sustainable, and inequalities in access should be eliminated (WHO and UNICEF,

2013).

4.5. Conclusions

This chapter has presented objective assessments of on-farm water quality as well as

associated factors of the quality of on-farm water for drinking and domestic use in the provinces

of Thai Binh and An Giang in Vietnam. The research study revealed that the quality of on-farm

water for both drinking and domestic use was low with most of the water samples contaminated

with E. coli and they did not met the quality standards for drinking and domestic water. Although

testing of E. coli in water is recommended for establishing if a water source is recently

contaminated with faeces, such faecal contamination of water is often intermittent and may not

be revealed in one assessment of a single sample. Therefore, in addition to having simple

affordable tests of E. coli in water, it is important to build capacity for inspecting and improving

household water sanitation, and on-farm water public health management. By doing so, rural

farmers will not only be able to afford and conduct the tests themselves, but also to know when it

is appropriate to take the test and what to do with the test results.

The modeling of the relationships between the objectively assessed quality of on-farm

water and the subjectively assessed on farm water quality by farmers gave some idea of the

factors that were relevant to further explore in the next chapter of this dissertation which

addresses possible associated factors for farmers’ mitigating strategies for WRZD transmission

in the context of SSI farming. The research results have also indicated areas for improving rural

water supply and sanitation and on-farm water and health management through updates, as well

as revising the water regulatory framework/policies on SSI farms in Vietnam. The results of the

research also showed the need to consider rural water testing and training in improving on-farm

of 340

water public health and that the microbial assessment of rural water quality can help improve

public health protection (Allen et al., 2010).

of 340

References

Allen, M. J., Payment, P. R., and Clancy, J. L. (2010). Rapid microbial methods can improve

public health protection. Journal (American Water Works Association), 102, 44-51.

Bach, A. H. (2007). Equivalence in the translation of Vietnamses cultural words in the book

"Wandering through Vietnamese culture" by Huu Ngoc. Vietnam College of Foreign

Languages (VHUN), Hanoi.



Brown, J. D., Swayne, D. E., Cooper, R. J., Burns, R. E., and Stallknecht, D. E. (2007).

Persistence of H5 and H7 avian influenza viruses in water. Avian Dis, 51, 285-289.


In Rome: FAO.

Carr, R. M., Blumenthal, U. J., and Mara, D. D. (2004). Guidelines for the safe use of

wastewater in agriculture: revisiting WHO guidelines. Water Sci Technol, 50, 31-38.

Costa, T. P., Brown, J. D., Howerth, E. W., and Stallknecht, D. E. (2011). Variation in viral

shedding patterns between different wild bird species infected experimentally with low-

pathogenicity avian influenza viruses that originated from wild birds. Avian Pathology,

40, 119-124.

Costyla, C., Bain, R., Cronk, R., and Bartram, J. (2015). Seasonal variation of faecal

contamination in drinking water sources in developing countries: A systematic review.

Science of the Total Environment, 541, 333-343.

of 340




Institute.

Dinh, N. P., Long H.P, Tien, N. T. K., Hien N.T, Phong L.H, and et al (2004). Risk factors for

human infection with avian influenza A H5N1, Vietnam. Emerging Infectious Diseases,

12, 1841-1847.

Domanska, B. K., Minta, Z., Smietanka, K., Sylvie, M., and Berg, A. T. (2010). H5N1 High

Pathogenicity Avian Influenza Virus Survival in Different Types of Water. Avian

Diseases Digest, 5, e180-e181.

Dooho, I. R., Martin, W., and Stryhn, H. (2009). Veterinary Epidemiologic Research. (2nd ed.)

VER, Inc..



Ford, T. E. and Colwell, R. (1996). A Global Decline in Microbiological Safety of Water: A Call

for Action. American Academy of Microbiolog.

FPTCW (2013). Guideline Technical Document: Escherichia coli. Ottawa, Ontario: Federal-

Provincial-Territorial Committee on Drinking Water of the Federal-Provincial-Territorial

Committee on Health and the Environment, Health Canada.

Gilbert, M., Xiao, X., Pfeiffer, D. U., Epprecht, M., Boles, S., Czarnecki, C. et al. (2008a).


the National Academy of Sciences of the United States of America, 105, 4769-4774.

of 340

Gilbert, M., Xiao, X., Pfeiffer, D. U., Epprecht, M., Boles, S., Czarnecki, C. et al. (2008b).


the National Academy of Sciences, 105, 4769-4774.

Gleeson, L. J., Nguyen, D. C., Do, P. H., and Nguyen, T. K. O. (2011). Joint Government -

United Nations Programme to Fight Highly Pathogenic Avian Influenza (HPAI): Final

Evaluation Report.

Guan, T. Y. and Holley, R. A. (2003). Pathogen survival in swine manure environments and

transmission of human enteric illness – a review. Journal of Environmental Quality, 32

(2), 392.

Hall, D. C. and Dinh, X. T. (2009). Perceived constraints on small scale poultry in Vietnam.

Asian Livestock.

Hall, D. C. and Le, Q. B. (2013). Impact on human, animal, and ecosystem health and food

security from restructuring of livestock production farming practices in Vietnam. The

First International One Health Congress, Human Health, Animal Health, the Environment

and Global survival, Melbourne, Australia Feb 2011 [On-line]. Available:

http://download.springer.com/static/pdf/638/art%253A10.1007%252Fs10393-010-0376-

0.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Farticle%2F10.1007%2Fs10393-

010-0376-

0andtoken2=exp=1447890688~acl=%2Fstatic%2Fpdf%2F638%2Fart%25253A10.1007

%25252Fs10393-010-0376-

0.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Farticle%252F1

0.1007%252Fs10393-010-0376-

0*~hmac=80c4fbd920ab0106f79b30de3faab0550896d5a6a1a5f310256c3bf160ef14d9

http://download.springer.com/static/pdf/638/art%253A10.1007%252Fs10393-010-0376-0.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Farticle%2F10.1007%2Fs10393-010-0376-0&token2=exp=1447890688~acl=%2Fstatic%2Fpdf%2F638%2Fart%25253A10.1007%25252Fs10393-010-0376-0.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Farticle%252F10.1007%252Fs10393-010-0376-0*~hmac=80c4fbd920ab0106f79b30de3faab0550896d5a6a1a5f310256c3bf160ef14d9








of 340

Henning, K. A., Henning, J., Morton, J., Long, N. T., Ha, N. T., and Meers, J. (2009). Farm- and

flock-level risk factors associated with Highly Pathogenic Avian Influenza outbreaks on

small holder duck and chicken farms in the Mekong Delta of Viet Nam. Preventive

Veterinary Medicine, 91, 179-188.

Hooda, P. S., Edwards, A. C., Anderson, H. A., and Miller, A. (2000). A review of water quality

concerns in livestock farming areas. Sci Total Environ, 250, 147-67.

Hubbard, R. K., Newton, G. L., and Hill, G. M. (2004). Water quality and the grazing animal.

Journal of Animal Science, 82, E255-E263.

Huynh, T. T. T., Aarnink, A. J. A., Drucker, A., and Verstegen, M. W. A. (2006). Pig production

in Cambodia, Laos, Philippines, and Vietnam: a review. Asian Journal of Agriculture and

Development, 3, 69-90.

IDRC (2012). Ecohealth Research in Practice. (Insight and Innovation in International

Development ed.) (vols. Innovative Applications of an Ecosystem Approache to Health)

International Development Centre (IDRC); Springer.

Ip, H., Flint, P., Franson, J. C., Dusek, R., Derksen, D., Gill, R., Jr. et al. (2008). Prevalence of

Influenza A viruses in wild migratory birds in Alaska: Patterns of variation in detection at

a crossroads of intercontinental flyways. Virol J, 5, 1-10.

JMP (2014). Improved and unimproved water and sanitation facilities. WHO/UNICEF [On-line].

Available: http://www.wssinfo.org/definitions-methods/watsan-categories/

Le, Q. B. and Hall, D. C. (2011a). Stakeholder cooperation and ecohealth to control Highly

Pathogenic Avian Influenza (HPAI) in Vietnam. Canadian Public Health Association

Annual Conference 2011, Montreal Quebec, 2011 [On-line]. Available:


http://www.wssinfo.org/definitions-methods/watsan-categories/


of 340

Le, Q. B. and Hall, D. C. (2011b). Using an ecohealth approach to livestock production:

community surveillance of newly emerging infectious diseases (nEIDs), International

Meeting on Emerging Diseases and Surveillance (IMED 2011.

Lin, S., Fend, M. L., Hu, R. G., Liu, R. Y., Wei, M. Y., and Jang, C. (2010). Characteristics of

Nitrogen Cycling in Farm Systems in a Small Watershed of Three Gorges Reservoir

Area, China. Environmental Science [On-line].

Lockwood, H. and Smits, S. (2011). Supporting Rural Water Supply. Practical Action Publishing

Ltd.

Lye, D. J. (2002). Health Risk Associated with Consumption of Untreaed Water from Household

Roof Catchment Systems. JAWRA Journal of the American Water Resources Association,

38, 1301-1306.

Markwell, D. D. and Shortridge, K. F. (1982). Possible waterborne transmission and

maintenance of influenza viruses in domestic ducks. Applied and Environmental


MoC (2000). Vietnam National Rural Clean Water Supply and Sanitation Strategy up to 2020

Ministry of Agriculture and Rural Development- Minsistry of Construction.

MoC and MARD (2000). National Rural Clean Water Supply and Sanitation Strategy up to the

year 2020 Hanoi, Vietnam: Ministry of Construction and Ministry of Agricultrue and





of 340






UN.

Nazir, J., Haumacher, R., Ike, A., Stumpf, P., B+¦hm, R., and Marschang, R. E. (2010). Long-

Term Study on Tenacity of Avian Influenza Viruses in Water (Distilled Water, Normal

Saline, and Surface Water) at Different Temperatures. Avian Diseases Digest, 5, e174-

e175.

Nguyen, H. N. (2007). Flooding in Mekong River Delta, Vietnam Hanoi: UNDP.

Nguyen, H. V., Vuong, T. A., Pham, D. P., and Vu, V. T. (2011). Safe use of wastewater in

agriculture and aquaculture (Rep. No. 2). The National Centre of Competence in

Research (NCCR) North-South.

Nguyen, T. V., Kitajima, M., Nguyen Vo, M. H., Matsubara, K., Takizawa, S., Katayama, H. et

al. (2008). Bacterial contamination of raw vegetables, vegetable-related water and river

water in Ho Chi Minh City, Vietnam. Water Science and Technology, 58, 2403-2411.

Pfeiffer, J., Pantin, J. M., To, T. L., Nguyen, T., and Suarez, D. L. (2009). Phylogenetic and

biological characterization of highly pathogenic H5N1 avian influenza viruses (Vietnam

2005) in chickens and ducks. Virus Research, 142, 108-120.

Phan, Q. M. (2010). Epidemiological Studies of Highly Pathogenic Avian Influenza in Vietnam.

Massey University.

of 340

Rapport, D. J., Howard, J., Maffi, L., and Mitchell, B. (2006). Avian Influenza and the

Environment: An Ecohealth Perspective.

Ravnborg, H. M. and Jensen, K. M. (2012). The water governance challenge: the discrepancy

between what is and what should be. Water Science and Technology: Water Supply, 12,

799-809.

Roche, B., Lebarbenchon, C., Gauthier-Clerc, M., Chang, C. M., Thomas, F., Renaud, F. et al.

(2009). Water-borne transmission drives avian influenza dynamics in wild birds: The

case of the 2005-2006 epidemics in the Camargue area. Infection, Genetics and

Evolution, 9, 800-805.

Sims, L. D., Domenech, J., Benigno, C., Kahn, S., Kamata, A., Lubroth, J. et al. (2005). Origin

and evolution of highly pathogenic H5N1 avian influenza in Asia. Veterinary Record,

157, 159-164.

Soares, R. J., Ortiz-Pelaez, A., Lai, T. K. L., Dinh, Q., Otte, J., and Pfeiffer, D. (2010).

Associations between attributes of live poultry trade and HPAI H5N1 outbreaks: a

descriptive and network analysis study in northern Vietnam. BMC Veterinary Research,

6, 10.

Stallknecht, D. E., Shane, S. M., Kearney, M. T., and Zwank, P. J. (1990). Persistence of Avian

Influenza Viruses in Water. Avian diseases, 34, 406-411.

Toews, D. W. (2009). Food, Global Environmental Change and Health: EcoHealth to the

Rescue? Mcgill Journal of Medicine, 12, 85-89.

Tran, H. P., Adams, J., Jeffery, J. A. L., Nguyen, Y. T., Vu, N. S., Kutcher, S. C. et al. (2010).

Householder perspectives and preferences on water storage and use, with reference to

dengue, in the Mekong Delta, southern Vietnam. International Health, 2, 136-142.

of 340

Trang, D. T., Hien, B. T. T., Molbak, K., Cam, P. D., and Dalsgaard, A. (2007a). Epidemiology

and aetiology of diarrhoeal diseases in adults engaged in wastewater-fed agriculture and

aquaculture in Hanoi, Vietnam. Tropical Medicine and International Health, 12, 23-33.

Trang, D. T., Molbak, K., Cam, P. D., and Dalsgaard, A. (2007b). Incidence of and risk factors

for skin ailments among farmers working with wastewater-fed agriculture in Hanoi,

Vietnam. Transactions of the Royal Society of Tropical Medicine and Hygiene, 101, 502-

510.

Trinh, T. L. (2001). Wastewater and its Treatment System in the Southern Part of Vietnam. In

International Water Management Institute: Proceedings of a Workshop held in Hanoi,

Vietnam 14 March 2001 Southern Institute of Water Resources Research.




UN (2013). Goal 7: Ensure Environmental Sustainability - Target 7.C. United Nations Goals

[On-line]. Available: http://www.un.org/millenniumgoals/environ.shtml

Vu, T. T. H. (2001). Treatment of Domestic Wastewater and its Reuse in Farm Irrigation in the

Red River Delta. In Wastewater Reuse in Agriculture in Vietnam: Water Management,

Environment and Human Health Aspects Center for Irrigation and Water Supply,

Vietnam Institute for Water Resources Research.

Ward, M. P., Maftei, D. N., Apostu, C. L., and Suru, A. R. (2009). Association Between

Outbreaks of Highly Pathogenic Avian Influenza Subtype H5N1 and Migratory

Waterfowl (Family Anatidae) Populations. Zoonoses and Public Health, 56, 1-9.


http://www.un.org/millenniumgoals/environ.shtml

of 340




WHO (2005). Avian Influenza A (H5N1) Infection in Humans. New England Journal of

Medicine, 353, 1374-1385.

WHO (2007a). Review of latest available evidence on potential transmission of avian influenza

(H5N1) through water and sewage and ways to reduce the risks to human

health (Rep. No. WHO/SDE/WSH/06.1). Geneva: World Health Organisation: Water,Sanitation

and Health Public Health and Environment.

WHO (2007b). Review of latest available evidence on potential transmission of avian influenza




WHO (2011a). Annex 1: Supporting documentation to the Guidelines. In (pp. 443-448). World


WHO (2011b). Guidelines for drinking-water quality. (Fourth ed.) World Health Organisation.

WHO (2011c). Microbial aspects. In Guidelines for drinking-water quality (pp. 117-153). World


WHO (2011d). Water and Public Health. World Health Organization [On-line]. Available:





of 340

WHO and OECD (2003a). Assessing Microbial Safety of Drinking Water: Improving

Approaches and Methods. IWA Publishing.

WHO and OECD (2003b). Assessing Microbial Safety of Drinking Water: Improving


WHO and OECD (2003c). Assessing Microbial Safety of Drinking Water: Improving


WHO and UNICEF (2013). Progress on sanitation and drinking-water - 2013 update France:

World Health Organization and UNICEF 2013.

Wilbers, G. J., Sebesvari, Z., and Renaud, F. G. (2014). Piped-Water Supplies in Rural Areas of

the Mekong Delta, Vietnam: Water Quality and Household Perceptions. Water

(20734441), 6, 2175-2194.

Yajiam, A. and Kurokura, H. (2008). Microbial risk assessment of livestock-integrated


Yajima, A. and Kurokura, H. (2008). Microbial risk assessment of livestock-integrated


of 340

Chapter Five: Factors Associated with Small-Scale Integrated (SSI) Farmers' Engagement

in Mitigating Strategies to Reduce Transmission of Water-Related Zoonotic Diseases

(WRZD) in the Provinces of Thai Binh and An Giang in Vietnam

5.1. Introduction

5.1.1. Rationale

The previous chapters of this dissertation have provided the background for the role of

SSI farms in the provinces of Thai Binh and An Giang in Vietnam. These chapters included the

following: 1) a profile of SSI farms; 2) the perceptions of SSI farmers toward risk factors for the

transmission of water-related zoonotic diseases (WRZD); and 3) the link between water quality

and public health among SSI farms. This background laid the foundation for this chapter in

which I discuss factors that are associated with engaging SSI farmers in implementing mitigating

strategies to reduce transmitting WRZD. In the following paragraphs, I explain some of the key

terms used in this chapter, and the conceptual and theoretical models of the research (see

Appendix H). I also provide a summary of the literature review for this chapter at the end of this

section (i.e., Section 5.1.1). Figure 5.1 additionally presents the structure of this chapter.

There are various definitions for the term mitigation. For example, in the context of a

pandemic influenza, the WHO (2008) has classified mitigation as actions or measures

undertaken beforehand to moderate the effect of an event or to decrease its impact on society and

the environment. Next, strategies and tactics are defined as the “what” and the “how”

components respectively of priority setting in health. Specifically, strategies are defined as action

plans (i.e., the “what”) to achieve health conditions that warrant the most attention and tactics

(i.e., the “how”) are defined as specific actions used to carry out these strategies (Bloom et al.,

2013). The WHO has recommended some key mitigating strategies as useful starting points for

reducing the transmission of WRZD. These mitigating strategies include livestock management,

of 340

source water protection, water storage, water treatment and distribution, and point of

use/household40 as recommendations (WHO, 2004)

In my research, I am interested in the types of mitigating strategies SSI farmers engage in

to reduce the transmission of WRZD and the factors associated with their engaging in such

mitigating strategies. I define a mitigating strategy to reduce the transmission of WRZD as a plan

that includes one or more activities a SSI farmer could choose to reduce transmitting WRZD

(Figure 5.1 and Section 5.2.1). In my research, an example of such a mitigating strategy is good

livestock management, which may include actions or tactics (e.g., treat animal waste at source)

that SSI farmers can choose to reduce the transmission of WRZD (Carr and Bartram, 2004;

WHO, 2004). Thus, in this chapter, I refer to WHO’s recommended key mitigating strategies for

reducing the transmission of WRZD as “the mitigating strategies”. Moreover, an SSI farmer is

considered to be engaging in a mitigating strategy to reduce transmission of WRZD if that

farmer takes certain actions included in that mitigating strategy (Section 5.2.1 and Table 5.1).

The WHO defines WRZD as diseases in humans that are spread from animals and are

related to water such as Highly Pathogenic Avian Influenza (HPAI), some parasites, or diseases

caused by a range of bacteria (e.g., pathogenic strains of Escherichia coli such as Shiga toxin-

producing E. coli) (Carr and Bartram, 2004; WHO, 2004). In Vietnam, 40% of farm animals

have diarrhea (Nguyen, 2004), and 31.3% and 14% of diarrheic calves carried at least one

fimbrial gene and genes for Shiga toxin respectively (Nguyen, Vo, and Vu-Khac, 2011). The

WHO has outlined some criteria for determining whether a potential pathogen is a WRZD. These

criteria include three elements: 1) the pathogen must spend part of its life cycle within one or

40 Point of Use/Household, as a mitigating strategy to reduce the transmission of WRZD, refers to actions relating to

conditions of water, sanitation and hygiene (personal, domestic, and food) which can affect the transmission of

WRZD at point of water use (e.g., handwashing) (WHO, 2004)

of 340

more animal species; 2) it is probable or conceivable that some life stage of the pathogen will

enter water; and 3) transmission of the pathogen from animals to humans must be through a

water- related route (Moe, 2004).

Figure 5.1 below shows a pictorial representation of the structure of Chapter Five.

Figure 5.1. Outline of Chapter Five

Chapter Five:

Presents factors associated with farmers'

engagement in mitigating strategies of

water- related zoonotic disease (WRZD)

transmission in the Thai Binh and An Giang

provinces of Vietnam

Act

ion

1

Clustered actions into mitigating strategies

SSI farmers’

perceptions of

risk factors

for WRZD

transmission

(from Chapter

Three)

Act

ion

4

Act

ion

..

. Act

ion

n

Act

ion

2

Act

ion

3


Do SSI farmers engage

in mitigating strategies

to reduce transmission

of WRZD?


What are factors associated with SSI

farmers’ engagement in mitigating

strategies to reduce transmission of

WRZD?

Livestock

Management

Point of use/

Household

Methods

Water

Storage,

Treatment,

Distribution

Source

Water

Protection

Socioecono

mic factors

of SSI

farms

(from

Chapter

Two)

Water quality

among SSI

farms

(from

Chapter

Four)

of 340

The conceptual model of my research is that SSI farmers make decisions, including

management decisions related to animal and human health, based on their socioeconomic

conditions and their impressions of the water quality around them. The theoretical model for my

research is adapted from the theory of planned behaviour and the health beliefs model (Becker,

2012; Payne et al., 2002; Rosenstock et al., 1988a). This adapted theoretical model for my

research indicates that the belief of SSI farmers in threats to their health and/or wellbeing,

combined with their beliefs in both the benefits and barriers to an action, together with their

socioeconomic status are contributing factors of whether they will engage in mitigating strategies

to reduce the transmission of WRZD (see Appendix A). This model is prefaced on the

observation that an individual’s experiences and actions are reflected in his or her perceptions;

therefore, perceptions can be associated with people’s health seeking behaviours (Brewer et al.,

2007; WHO, 2012). For example, the perceptions of individuals about the threat of diseases,

expectations, cues to actions41, and their socioeconomic status may be associated with their

health behaviours (Becker, 1974; Rosenstock et al., 1988b). Therefore, understanding the

associations between the perceptions of SSI farmers towards WRZD transmission, their

socioeconomic status, and their behaviours could be useful for research into and interventions

needed to improve their engagement in mitigating strategies to reduce WRZD transmission.

A review of the related literature indicated that although SSI farmers have a basic

knowledge about proper hygienic practices, they usually employ ineffective practices for hygiene

measures (e.g., hand washing without soap). Misconceptions about risk factors and/or a lack of

knowledge of the cause-effect relationships, habits, and/or not ranking health risks highly have

41 A cue to action is “the stimulus needed to trigger the decision making process to accept a recommended health

action.” Examples of possible cues to actions are chest pains, mass media campaigns, advice from others, a reminder

postcard from a physician, illness of family member or friend, etc.) (BUMC, 2013).

of 340

been proposed as links to ineffective practices of hygiene measures (Herbst et al., 2009; Keraita

et al., 2010; Rechenburg and Herbst, 2006). Even though SSI farmers are aware of the health

risks and potential health benefits to be gained by implementing mitigating actions, other factors

such as overall economic impact on work efficiency were thought to be associated with their

assessment of mitigating actions (Keraita et al., 2010). Some studies in Africa (Mekoya et al.,

2008; Adesina and Baidu, 1995; Abdul et al., 2014; Negatu and Parikh, 1999) and South East

Asia (Van Mele and Cuc, 2001; Cramb, 2005; Le et al., 2014a) have explored potential factors

that are associated with rural farmers’ choices of adopting a farming technique or an adaptive

measure for climate change. However, these studies did not address SSI farming and the

transmission of WRZD. Yet, these studies provided some ideas for further research of factors

(e.g., incomes) that may be associated with SSI farmers’ choices. In summary, I have not found

any studies that examine factors associated with SSI farmers’ engagement in mitigating

strategies to reduce WRZD transmission. An understanding of these factors may help

inform/bring scientific evidence for policy advocacy and assist with education/knowledge

relating to SSI farming, rural water quality, and public health. Therefore, in my research, I

explored factors that may be associated with SSI farmers’ engagement in mitigating strategies to

reduce transmission of WRZD.

5.1.2. Research questions and hypothesis

5.1.2.1. Research questions

This chapter addresses the sixth and seventh questions of my research, which are: 1) Do

SSI farmers in the provinces of An Giang and Thai Binh engage in mitigating strategies to

reduce the transmission of WRZD? and 2) What factors are associated with SSI farmers’

engagement in mitigating strategies to reduce the transmission of WRZD?”

of 340

5.1.2.2 Research hypothesis

I hypothesized that in the provinces of An Giang and Thai Binh, SSI farmers’

socioeconomic status and their perceptions of risk factors for WRZD transmission are associated

with their engagement in mitigating strategies to reduce transmission of WRZD.

5.2. Methods

This section describes the methods used to identify SSI farmers’ engagement in the

mitigating strategies to reduce transmission of WRZD and the methods used to explore potential

factors that are associated with each of these mitigating strategies. The theoretical model used for

my research guided the selection of these potential factors. The other details of the methods (e.g.,

sample size, studied areas, and questionnaires) used to collect data relating to these potential

factors were presented in Chapters Two, Three, and Four.

5.2.1. Methods used to identify SSI farmers’ engagement in mitigating strategies to reduce

the transmission of WRZD

To identify SSI farmers’ engagement in the mitigating strategies defined in section 5.2.1,

I first collected data pertaining to farmers’ specific choice of actions that were clustered under

the key WHO recommended mitigating strategies (Table 5.1). Then, I created thresholds of

actions to generate variables that reflect farmers’ engagement in each strategy. In these

thresholds, an SSI farmer is considered engaging in a mitigating strategy to reduce transmission

of WRZD if the farmer takes certain actions included in that mitigating strategy (See section

5.3.1.1 for more detailed information). I used questionnaires to collect data from SSI farmers

about their choice of actions to mitigate WRZD transmission. Table 5.1 provides detailed

descriptions of the mitigating strategies to decrease WRZD transmission and the corresponding

actions which were adapted and selected based on the WHO’s recommendations (WHO, 2004).

of 340

The field data collection focused on the actions clustered under the mitigating strategies. I

asked SSI farmers to indicate their choice of actions (Table 5.1) to reduce transmission of

WRZD. Their answers were either yes or no indicating whether they did or did not use each of

the actions that were included in each of the mitigating strategies. Given the collected data on the

corresponding actions, I used the thresholds mentioned above to create variables that reflected

SSI farmers’ engagement (yes or no) in mitigating strategies to decrease WRZD transmission.

These binary variables helped to inform the second part of the hypothesis. An SSI farmer’s

engagement in a mitigating strategy was coded as “1” and “0” otherwise (See section 5.3.1.1 for

more detailed information). I used descriptive statistics to summarize these binary variables.

Section 5.2.2 shows the methods used to explore factors that are associated with these binary

variables (i.e., factors that are associated with SSI farmers’ engagement in mitigating strategies

of WRZD transmission).

of 340

Table 5.1. Small-Scale Integrated (SSI) Farmers’ Mitigating Strategies and Their

Corresponding Actions in Reducing Risk Factors for Water-Related Zoonotic Disease

(WRZD) Transmission (*)

Mitigating strategies Corresponding actions (i.e., tactics)

Livestock management (LM)

strategy

- Vaccinate domestic animals

- Ensure hygienic rearing conditions for

animals

- Treat animal waste at source

- Exclude animals from catchment basin

- Stop using antibiotics as growth promoters

Source water protection (SWP)

strategy

- Treat wastewater

- Treat human and animal waste at source

- Use buffer zones

- Protect wells/springs/groundwater

- Cover drinking water reservoirs

- Repair leaking sanitation systems

- Treat human/animal waste prior to use as

fertilizers

Water storage, treatment, and

distribution (WSTD) strategy

- Change household water treatment

technologies

- Change/use disinfectants

- Repair household water distribution

- Repair leaking household sewers

Point of use/household (POU)

strategy

- Ensure adequate household sanitation

- Invest in improved water supply

- Protect house hold water sources and food

supply from animals

- Wear protective equipment (e.g., mouth

mask) (*) Adapted from the WHO’s recommendations (WHO, 2004).

5.2.2. Methods used to explore factors that are associated with SSI farmers’ engagement in

mitigating strategies

I used a univariate inferential statistical analysis technique to explore factors that are

associated with SSI farmers’ engagement in the mitigating strategies and to test the research

hypothesis. Stata 13 was used to assist in the data analysis (Stata, 2013). More specifically, in my

of 340

research, a probit regression technique (hereafter referred to as probit analysis) was used to

examine specific factors (explanatory variables) that may be associated with SSI farmers’

engaging in mitigating strategies to reduce WRZD transmission (outcome variables). Probit

analysis is commonly used for analyzing binary outcome variables and exploring how each

factor affects the probability of an outcome event occurring (Greene, 2008b; Long and Freese,

2014).

The univariate probit analysis model is specified accordingly as follows: assuming “i” is

an engagement in a mitigating strategy for each farmer, the equation describing a latent variable

that corresponds to SSI farmers’ engagement in mitigating strategies is written as:

Yi* = β0+ βi Xi + ɛi ; with i = 1,2,3,.. n (i)

Yi = {0, 1} (ii)

Yi* is an unobserved variable representing the latent utility or propensity of the Yi under

consideration. Yi is an observed dummy variable reflecting SSI farmers’ engagement in a

mitigating strategy. Yi equals 1 if Yi* >0 (i.e., a farmer used a mitigating strategy) and Yi equals

0 otherwise.

Xi is a vector of observed explanatory variables (e.g., perceptions and water quality) believed to

be relevant to the consideration of the ith equation.

βi is a vector of unknown coefficients to be estimated.

ɛi is an error term, which is normally distributed with mean 0 and variance 1.

Dependent (outcome) variables included SSI farmers’ engagement in mitigating

strategies to reduce transmission of WRZD coded as dichotomous variables. The mitigating

strategies explored in my research (see section 5.1.1 and Table 5.2) were coded as “0” if SSI

farmers did not engage in a mitigating strategy and “1” otherwise. A set of explanatory variables

of 340

of interest was specified based on the conceptual and theoretical models and literature review of

my research. These explanatory variables included characteristics of SSI farmers and their farms

(e.g., age, income, and livestock production) and SSI farmers’ perceptions (e.g., perceived

barriers to actions). Descriptive statistics were used to summarize the explanatory variables used

in the univariate probit analysis.

Probit analysis can show rates of changes in magnitudes and levels of significance of the

associations between the explanatory variables and an outcome variable. In my research,

estimates of individual probit coefficients of the explanatory variables as well as the

corresponding marginal probability effects (i.e., marginal effects) at the means of explanatory

variables were used to report the results of the modeling factors that are associated with SSI

farmers’ engagement in mitigating strategies. A probit coefficient (βi) is a change in the “z-

score” of Yi* given a one unit change in Xi, holding all other explanatory variables constant at

their means. Meanwhile, a marginal effect is the change in Pr (Yi = 1) for a one unit change in Xi,

given that all other explanatory variables are held constant at their means (see the Discussion

section for a more detailed interpretation of probit coefficients and marginal effects). The

significance level was set a priori at p-value of 0.1 or less which is a common significance level

(Hall et al., 2003; JAE, 2014) .

5.3. Results

This section presents the results of my research in this chapter including 1) SSI farmers’

engagement in mitigating strategies for reducing WRZD transmission and their corresponding

actions, and 2) factors that are associated with SSI farmers’ engagement in mitigating strategies

to decrease WRZD transmission. Appendix G provides an additional summary of information

that was presented in Chapter Two and in Chapter Three and was used as part of the data

of 340

analysis in this chapter. The additional information includes summaries of the socio-

demographic characteristics of SSI farmers, SSI farmers’ perceptions of risk factors for WRZD

transmission, and SSI farmers’ livestock production.

5.3.1. SSI farmers’ actions and engagement in mitigating strategies to reduce WRZD

transmission

5.3.1.1 Livestock management (LM) actions

There were five actions used to indicate SSI farmers’ engagement in livestock

management strategies to mitigate WRZD transmission (Table 5.1, Figure 5.2). These actions

included vaccinating domestic animals, treating animal waste at source, discontinuing using

antibiotics as a growth promoter, ensuring hygienic rearing conditions for animals, and excluding

animals from catchment basins. The results showed significant differences between two

provinces in terms of the numbers of SSI farmers employing these actions. The majority of the

SSI farmers in the province of Thai Binh engaged in all five livestock management actions to

mitigate the transmission of WRZD, with vaccination being the most common and excluding

animals from the catchment basin being the least common. In the province of An Giang,

vaccination was the only action used by a majority of the SSI farmers to mitigate the

transmission of WRZD. Each of the remaining actions was employed by a third or less of the SSI

farmers, with exclusion of animals from the catchment basin again being the least common..

5.3.1.2. Source water protection (SWP) actions

SSI farmers in two provinces indicated seven actions that they took to protect sources of

water on their farms (SWP). These actions included 1) treating wastewater, 2) using a buffer

zone, 3) covering the drinking water reservoir, 4) treating animal and human waste prior to its

use as fertilizer, 5) treating animal and human waste at source, 6) protecting well/spring/ground

water, and 7) repairing a leaking sewage system. These actions reflected SSI farmers’

of 340

engagement in a SWP strategy to mitigate WRZD transmission. However, significantly fewer

SSI farmers in An Giang took these actions compared to those in Thai Binh. The greatest number

of SSI farmers in both provinces selected “Covering the drinking water reservoir” as a key risk

mitigating action (Figure 5.3).

5.3.1.3.Water storage treatment and distribution (WSTD) actions

SSI farmers in both provinces indicated that they took four key actions regarding storage,

treatment, and distribution of water to mitigate WRZD transmission (Figure 5.4). These actions

included 1) spending money on household treatment technology, 2) repairing household water

storage, 3) changing/using disinfectants, and 4) repairing household sewers. There was

significantly (p < 0.01) fewer SSI farmers in An Giang compared to those in Thai Binh who took

each of these actions. Among the four actions, the greatest number of SSI farmers in An Giang

chose was repairing household water storage. Meanwhile, repairing the household sewer instead

of the household water storage was the most common action taken by SSI farmers in Thai Binh.

of 340

Figure 5.2. Livestock Management Actions

Figure 5.3. Source Water Protection Actions

0

100

200

300

Thai Binh An Giang

Vaccination of domestic animals (p<0.01) Ensure hygienic rearing conditions for animals (p<0.01)

Treat animal waste at source (p<0.01) Exclude animals from catchment basin (p<0.01)

Stop using antibiotics (p<0.01)

Num

ber

of fa

rme

rs

Graphs by Provinces

Livestock management actions0

100

200

300

Thai Binh An Giang

Treat wastewater (p< 0.01) Treat human/animal waste at source (p<0.01)

Use buffer zones (p<0.01) Protect well/springs/ground water (p<0.01)

Cover drinking water reservoirs (p<0.01) Repair leaking in sewage system (p<0.01)

Treat animal/human wastes prior to use as fertilizers (p=0.01)

Num

ber o

f far

mer

s

Graphs by Provinces

Source water protection actions

of 340

Figure 5.4. Water Storage, Treatment, and Distribution Actions

5.3.1.4. Point of use/household (POU) actions

Four key actions at a POU level used to mitigate transmission of WRZD were the

following: having adequate household sanitation, protecting water sources and food sources for

animals, investing in water supply improvement, and wearing protective equipment. The results

showed that there were significantly greater numbers of SSI farmers in Thai Binh compared to

An Giang who selected each of these four actions. Ensuring adequate household sanitation was

the most common POU action in both provinces. However, the actions that were taken by the

least number of SSI farmers were different between the two provinces – wearing protective

equipment (in the province of An Giang) and protecting water and food sources for animals (in

the province of Thai Binh) (See Figure 5.5).

050

100

150

200

250

Thai Binh An Giang

Spend money in household treatment technology (p<0.01) Change/use disinfectants (p<0.01)

Repair household water storage (p<0.01) Repair household sewer (p<0.01)

Nu

mbe

r o

f fa

rmer

s

Graphs by Provinces

Water storage distribution and treatment actions

of 340

Figure 5.5. Point of Use/Household Actions

5.3.1.5. SSI farmers’ engagement in mitigating strategies to reduce WRZD transmission

Sections 5.3.1.2-4 provided detailed information of the actions that SSI farmers took to

reduce the transmission of WRZD. In this section, these actions were clustered under the four

broad mitigating strategies to reduce the transmission of WRZD using a certain thresholds. The

World Health Organization recommends these mitigating strategies to control and prevent

WRZD in the context of rural areas (WHO, 2004).

Table 5.2 below presents a description of thresholds used to cluster farmers’ actions

under the four mitigating strategies and create variables that reflect SSI farmers’ engagement in

these mitigating strategies to reduce transmission of WRZD. An SSI farmer was considered to

engage in a mitigating strategy if that farmer took certain actions. For example, an SSI farmer

was considered “engaged in LM strategy” if that farmer took three or more of the five LM

actions and “not engaged in LM strategy” otherwise. The thresholds (i.e., number of actions)

were defined based on the goodness-of-fit of the models and the number of observations. The

table also provides summary statistics for SSI farmers’ engagement in the mitigating strategies.

050

100

150

200

250

Thai Binh An Giang

Adequate household sanitation (p<0.01) Invest in improving water supply (p<0.01)

Protect water source and food source for animals (p<0.01) Wear protective and filter water use (p<0.01)

Num

ber o

f far

mer

s

Graphs by Provinces

Point of use/household actions

of 340

Table 5.2. Summary Statistics of Variables Indicating SSI Farmers’ Engagement in

Mitigating Strategies for WRZD Transmission

Variable name Description - SSI farmers' engagement in:

Count ( =

“1”)

% ( =

“1”)

DFE-Livestock

Management

Livestock management as a mitigating

strategy (LM strategy) with “1” if SSI

farmers took three or more of the five

actions included in the LM strategy and

“0” otherwise. 329 55

DFE-Water

Protection

Source water protection as a mitigating

strategy (SWP strategy) with “1” if SSI

farmers took four or more of the seven

actions included in the SWP strategy and

“0” otherwise. 299 50

DFE-Point of

Use

Point of use/household method as a

mitigating strategy (POU strategy) with

“1” if SSI farmers took two or more of the

four actions included in the POU strategy

and “0” otherwise. 312 52

DFE-Water

Distribution

Water storage, treatment, and distribution

as a mitigating strategy (WSTD strategy)

with “1” if SSI farmers took three or more

of the four actions included in the WSTD

strategy and “0” otherwise. 325 54 Notes: D = Dummy, FE = Farmers’ engagement in a mitigating strategy.

5.3.2. Factors associated with SSI farmers’ engagement in the mitigating strategies to reduce


5.3.2.1. Results of probit analyses of SSI farmers’ engagement in mitigating strategies to reduce

WRZD transmission

Table 5.3a-c shows the estimated probit coefficients of the factors that are associated with

each of the four mitigating strategies that SSI farmers engaged in to mitigate WRZD

transmission. Overall, the probit analyses revealed that the models for the four key mitigating

strategies were significant at p<0.1 with McFadden's R2 from 0.4440 to 0.6214, and the

percentages of correct predictions from 83% to 89% (Table 5.5). The models for SSI farmers’

engaging mitigating strategies to reduce WRZD transmission had 14,13, 12, and 14 significant

factors in LM (i.e., DFE-Livestock Management variable), SWP (i.e., DFE-Water Protection

of 340

variable), POU (i.e., DFE-Point of Use variable), and SDT (i.e., DFE-Water Distribution

variable) respectively (Table 5.3a-c). “SSI farmers’ income” (i.e., DIncome variable) appeared to

be negatively significantly associated with SSI farmers’ engagement in all four mitigating

strategies while “Female SSI farmers primarily responsible for family health” (i.e., DFPR-

Health-Woman variable) was positively significantly associated with three mitigating strategies.

5.3.2.2. Marginal probability effects of the factors that are associated with SSI farmers’

engagement in mitigating strategies to reduce WRZD transmission

In this section, more intuitive results of probit analyses are presented (i.e., a marginal

probability effect, or a rate of change of the probability of each of the dependent variables as per

a unit change in an explanatory variable at mean values of other explanatory variables). A

marginal probability effect shows not only levels of significance and signs but also the changes

(in percentage) of the magnitudes of the association between a factor and SSI farmers’

engagement in a mitigating strategy holding other explanatory variables constant at their means.

Table 5.5a shows that the changes in the magnitudes of the associations ranged from as

low as negative 1% to as great as positive 44% for the probability of a per unit change in each of

the factors that are associated with a mitigating strategy, while holding other factors at their

means. Having an additional chicken increases by 1% the probability of a farmer engaging in

good livestock management as a mitigating strategy to decrease WRZD transmission (Table

5.5b). Meanwhile, SSI farmers who considered “worry of WRZD transmission” a trigger for

their actions had an increase of 44%in the probability of engaging in the LM mitigating strategy

(Table 5.5b).

of 340

Table 5.3a. Maximum Likelihood Estimates of Factors that are Associated with Small-Scale Integrated (SSI) Farmers’

Engagement in Mitigating Strategies of Water Related Zoonotic Diseases (WRZD) Transmission Using Probit Analyses

Variables DFE-Livestock Management DFE-Water Protection DFE-Water Distribution DFE-Point of Use

Estimates a Estimates Estimates Estimates

_cons -2.3361* b -2.5403* -3.5636* -1.5002*

(0.6613) c (0.6032) (0.6888) (0.4377)

DProvince - - - 0.9727*

- - - (0.2585)

Schooling 0.0696* 0.0979* 0.0679* -

(0.0362) (0.0334) (0.0350) -

Experience 0.0232 0.0231* 0.0237 0.0100

(0.0175) (0.0145) (0.0154) (0.0137)

DChildren 0.1499 - 0.0016 0.2206*

(0.1141) - (0.1111) (0.0940)

DFPR-Health-Man 0.5871* 0.1698 0.3444 -

(0.3298) (0.2885) (0.3020) -

DFPR-Health-Woman 1.3567* 1.0419* 1.4946* -

(0.3761) (0.3267) (0.3856) -

DFPR-Livestock-Man - - - 0.1190

- - - (0.1938)

DFPR-Livestock-Woman - - - 0.5163*

- - - (0.2963)

DIncome -0.4408* -0.4861* -0.8940* -0.7309*

(0.3121) (0.3087) (0.3080) (0.2653)

Npoultry/yr 0.0252* 0.0032 0.0040 -

(0.0079) (0.0036) (0.0039) -

Notes: D = Dummy, Dom = Domestic, FE = Farmers’ engagement in, FPS = Farmers’ perceived susceptibility to, FPT = Farmers’ perceived trigger of taking

mitigating actions, N = Number, a Estimated values of individual coefficients of explanatory variables – each coefficient represents a change in the “z-score” of

Yi* given a one unit change in Xi; b An asterisk (*) indicates a statistical significance at the level of p<0.10; c Values in brackets are standard errors

of 340

Table 5.3b. Maximum Likelihood Estimates of Factors that are Associated with Small-Scale Integrated (SSI) Farmers’

Engagement in Mitigating Strategies of Water Related Zoonotic Diseases (WRZD) Transmission Using Probit Analyses

(cont’d.)

Variables DFE-Livestock Management DFE-Water Protection DFE-Water Distribution DFE-Point of Use


Npigs/yr 0.0435* b 0.0223* 0.0249* 0.0074

(0.0106) c (0.0068) (0.0087) (0.0063)

Ncattle/yr -0.0378 -0.0004 - -

(0.0419) (0.0074) - -

DFPS-AI-Dom 0.5902* 0.0034 -0.2308 -0.1185

(0.2706) (0.2177) (0.2299) (0.1947)

DFPS-AI-Wastewater 0.4498 0.8853* 0.8076* 0.7799*

(0.3640) (0.3265) (0.3343) (0.2936)

DAI Severe -0.3098 -0.3062 0.7775* -

(0.4752) (0.4256) (0.4296) -

DFPSafe-Drink -1.0939* b - - -0.7434*

(0.4534) c - - (0.3696)

DFPBarr-Cost -1.0431* -0.6013* -0.7807* -0.7700*

(0.2527) (0.2138) (0.2261) (0.1938)

DFPBarr-Not know 0.1992 0.3410 0.7508* -

(0.2547) (0.2226) (0.2397) -

DFPBarr-Peer -0.0301 -0.9428* 0.1689 -

(0.5323) (0.4213) (0.4978) -

DFPBen-Action -0.3946 -0.0880 0.5566* -

(0.3321) (0.3006) (0.3207) -

DFPE-Livestock Management 0.9814* 0.5686* 0.7246* 0.8607*

(0.2707) (0.2119) (0.2361) (0.2971)

Notes: AI = Avian Influenza, D = Dummy, Dom = Domestic, FE = Farmers’ engagement in, FPE = Farmers’ perceived self-efficacy in, FPS = Farmers’

perceived susceptibility to, FPT = Farmers’ perceived trigger of taking mitigating actions, N = Number, a Estimated values of individual coefficients of

explanatory variables – each coefficient represents a change in the “z-score” of Yi* given a one unit change in Xi; b An asterisk (*) indicates a statistical

significance at the level of p<0.10; c Values in brackets are standard errors

of 340

Table 5.3c. Maximum Likelihood Estimates of Factors that are Associated with SSI Farmers’ Engagement in Mitigating

Strategies of Water Related Zoonotic Diseases (WRZD) Transmission using Probit Analyses (cont’d.)

Variables DFE-Livestock Management DFE-Water Protection DFE-Point of Use DFE-Water Distribution


DFPE-Water Protection - - -0.4701* -

- - (0.2788) -

DFPT-Health Advice 0.4292* b 0.4429* 0.6408* 0.4929*

(0.2336) c (0.2138) (0.1857) (0.2251)

DFPT-Loss Income -0.1377 0.1484 - 0.5378*

(0.2957) (0.2475) - (0.2785)

DFPT-Peer 0.8522* 1.0932* 0.3115 0.8266*

(0.3426) (0.2992) (0.2322) (0.3285)

DFPT-Worry 1.3433* 0.6463* 0.6737* 0.2645

(0.3008) (0.2332) (0.2120) (0.2484)

LogE.coli Drinkd - 0.0551 -0.0003 0.0919

- (0.0475) (0.0446) (0.0531)

LogE.coli Dom -0.0874* - -0.0710* -

(0.0477) - (0.0417) -

Turbidity Drink - -0.0016 - 0.0018

- (0.0092) - (0.0090)

Turbidity Domestic - -0.0049* - -0.0048*

- (0.0029) - 0.0030

Model chi-squaree 313.4700* 196.4400* 237.1500* 215.1200*

McFadden's R 2 f 0.6214 0.4440 0.4537 0.4964

% correct predictions 0.8862 0.8349 0.8307 0.8380

Notes: D = Dummy, Dom = Domestic, FE = Farmers’ engagement in, FPE = Farmers’ perceived self-efficacy in, FPT = Farmers’ perceived trigger of taking

mitigating actions, a Estimated values of individual coefficients of explanatory variables– each coefficient represents a change in the “z-score” of Yi* given a

one unit change in Xi; b An asterisk (*) indicates a statistical significance at level of p<0.10; c Values in brackets are standard errors; d E.coli Drink and E.coli

Dom were transformed into LogE.coli Drink and LogE.coli Dom using a natural log transformation; e Test of the overall model fit with null hypothesis indicating

that all individual coefficients of the model equal zero; f This is a pseudo R squared statistic calculated by 1 minus the ratio of the log-likelihood of the full model

(i.e., with all explanatory variables) to the log-likelihood of the intercept model (i.e., without explanatory variables)

of 340

Table 5.4. Predicted Frequency of Mitigating Strategies

DFE-Livestock

Management

DFE-Water

Protection DFE-Point of Use DFE-Water Distribution

Classified D ~D Total D ~D Total D ~D Total D ~D Total

+ 187 19 206 150 28 178 170 34 204 163 26 189

- 23 140 163 25 118 143 30 144 174 28 104 132

Total 210 159 369 175 146 321 200 178 378 191 130 321

Sensitivity 89.05% 85.71% 85.00% 85.34%

Specificity 88.05% 80.82% 80.90% 80.00%

Positive predictive value 90.78% 84.27% 83.33% 86.24%

Negative predictive value 85.89% 82.52% 82.76% 78.79%

False + rate for true ~D 11.95% 19.18% 19.10% 20.00%

False - rate for true D 10.95% 14.29% 15.00% 14.66%

False + rate for classified + 9.22% 15.73% 16.67% 13.76%

False - rate for classified - 14.11% 17.48% 17.24% 21.21%

Correctly classified 88.62% 83.49% 83.07% 83.18% Classified + if predicted Pr (D) >= .5

True D defined as DFE-Livestock Management! = 0; DFE-Water Protection = 0; DFE-Point of Use = 0; DFE-Water Distribution = 0.

Notes: D = Dummy, FE = Farmers’ engagement in

of 340

Table 5.5a. Marginal Effects (at Means) of Factors Associated with SSI Farmers’

Engagement in Mitigating Strategies for WRZD Transmission Using Probit Analyses

Variables

DFE-Livestock

Management

DFE-Water

Protection DFE-Point of Use

DFE-Water

Distribution

Marginal effect a Marginal effect Marginal effect Marginal effect

DProvince - - 0.3860* -

- - 0.1025 -

Schooling 0.0226* 0.0383* - 0.0229*

(0.0119) b (0.0131) - (0.0118)

Experience 0.0075 0.0090 0.0040 0.0080

(0.0058) (0.0057) (0.0054) (0.0052)

DChildren 0.0486 - 0.0875* 0.0005

(0.0371) - (0.0373) (0.0374)

DFPR-Health-

Man 0.1961* 0.0675 - 0.1248

(0.0995) (0.1141) - (0.1043)

DFPR-Health-

Woman 0.3371* 0.3546* - 0.3588*

(0.0668) (0.0887) - (0.0606)

DFPR-Livestock-

Man - - 0.0474 -

- - (0.0771) -

DFPR-Livestock-

Woman - - 0.1984* -

- - (0.1076) -

DIncome -0.1430* -0.1903* -0.2901* -0.3007*

(0.1028) (0.1214) (0.1056) (0.1066)

Npoultry/yr 0.0082* 0.0013 - 0.0013

(0.0024) (0.0014) - (0.0013)

Npigs/yr 0.0141* 0.0087* 0.0029 0.0084*

(0.0031) (0.0026) (0.0025) (0.0027)

Ncattle/yr -0.0123 -0.0001 - -

(0.0138) (0.0029) - -

DFPS-AI-Dom 0.1914* 0.0013 -0.0470 -0.0776

(0.0882) (0.0852) (0.0772) (0.0774)

DFPS-AI-

Wastewater 0.1459 0.3465* 0.3095* 0.2716*

(0.1197) (0.1281) (0.1167) (0.1145)

DAI Severe -0.1005 -0.1198 - 0.2615*

(0.1549) (0.1666) - (0.1431)

Notes: AI = Avian Influenza, D = Dummy, Dom = Domestic, FE = Farmers’ engagement in, FPR = Farmers’

perceived main responsibility, FPS = Farmers’ perceived susceptibility to, FPT = Farmers’ perceived trigger of

taking mitigating actions, N = Number, a Marginal probabilities of changes in dependent variables as per changes in

explanatory variables at mean values of other explanatory variables; b Values in brackets are standard errors; (*)

indicates a statistical significance at the level of p<0.10

of 340

Table 5.5b. Marginal Effects (at Means) of Factors that are Associated with Small-Scale

Integrated (SSI) Farmers’ Engagement in Mitigating Strategies of WRZD Transmission

Using Probit Analyses

Variables

DFE-Livestock

Management

DFE-Water

Protection

DFE-Point of

Use

DFE-Water

Distribution

Marginal effect a Marginal effect Marginal effect Marginal effect

DFPSafe-Drink -0.3547* - -0.2950* -

(0.1543) - (0.1470) -

DFPBarr-Cost -0.3383 -0.2353* -0.3056* -0.2626*

(0.0826) (0.0837) (0.0771) (0.0763)

DFPBarr-Not know 0.0646 0.1335 - 0.2525*

(0.0831) (0.0874) - (0.0806)

DFPBarr-Peer -0.0098 -0.3690* - 0.0568

(0.1727) (0.1650) - (0.1671)

DFPBen-Action -0.1280 -0.0345 - 0.1872*

(0.1097) (0.1177) - (0.1069)

DFPE-Livestock

Management 0.3183* 0.2225* 0.3416* 0.2437*

(0.0867) (0.0830) (0.1176) (0.0787)

DFPE-Water Protection - - -0.1866 -

- - (0.1106) -

DFPT-Health Advice 0.1392* 0.1733* 0.2543* 0.1658*

(0.0767) (0.0838) (0.0737) (0.0770)

DFPT-Loss Income -0.0446 0.0581 - 0.1809*

(0.0956) (0.0969) - (0.0932)

DFPT-Peer 0.2764* 0.4278* 0.1236 0.2780*

(0.1120) (0.1175) (0.0922) (0.1094)

DFPT-Worry 0.4356* 0.2530* 0.2673* 0.0890

(0.1026) (0.0915) (0.0842) (0.0836)

LogE.coli Drink - 0.0216 -0.0001 0.0309*

- (0.0186) (0.0177) (0.0178)

LogE.coli Dom -0.0284* - -0.0282* -

(0.0157) - (0.0166) -

Turbidity Drink - -0.0006 - 0.0006

- (0.0036) - (0.0030)

Turbidity Domestic - -0.0019* - -0.0016*

- (0.0011) - (0.0010)

Notes: D = Dummy, Dom = Domestic, FE = Farmers’ engagement in, FP = Farmers’ perceptions of, FPBarr =

Farmers’ perceived barriers to taking mitigating actions, FPBen = Farmers’ perceived benefit of, FPS = Farmers’

perceived susceptibility to, FPT = Farmers’ perceived trigger of taking mitigating actions, a Marginal probabilities of

changes in dependent variables as per changes in explanatory variables at mean values of other explanatory

variables; b Values in brackets are standard errors; (*) indicates a statistical significance at the level of p<0.10

of 340

5.4. Discussion

In this discussion section, I start with a discussion of the results of SSI farmers’ choices

of actions and their engaging in mitigating strategies to reduce the transmission of WRZD

(section 5.4.1). Then, I unpack the important results of my research – the factors that are

associated with SSI farmers’ engagement in mitigating strategies to reduce the transmission of

WRZD (section 5.4.2). Next, the limitations of my research will be outlined towards the end of

this section (section 5.4.3).

5.4.1. SSI farmers’ choices of actions and their engagement in mitigating strategies to reduce

the transmission of WRZD

On-farm water quality and its link to public health are complex due to the diversity of

WRZDs (Dufour, 2004). This presents a challenge for public health as each WRZD may require

different strategies to mitigate it, depending on the pathogen involved and the source of the

problem (i.e., LM, SWP, WSDT, and POU) (WHO, 2004). Each of these key strategies,

recommended by the WHO, requires a number of actions/tactics to mitigate WRZD transmission

(Carr and Bartram, 2004). Observations of SSI farms during the field survey outlined in my

research indicated that SSI farmers usually took more than one action with various combinations

and frequencies of actions to mitigate the transmission of WRZD. My research uncovered

interesting results concerning SSI farmers’ choices of specific actions together with their broad

engagement in various strategies (outlined in Table 5.1) to mitigate the transmission of WRZD.

5.4.1.1. Wide differences in choices of specific actions to mitigate WRZD transmission

The SSI farmers in the research study selected a wide range of actions (clustered under

the WHO recommended mitigating strategies) to help them reduce the transmission of WRZD.

Specifically, more SSI farmers in Thai Binh than in An Giang chose actions to mitigate the

of 340

transmission of WRZD. These greater numbers are supported by the results of the qualitative

component of my research (e.g., in-depth interviews), which established that in Thai Binh SSI

farmers and the general population were culturally more conservative due to a more difficult

economy as well as colder weather conditions compared to those in An Giang. Economic factors

may also explain why more SSI farmers in Thai Binh compared to An Giang took actions to

mitigate WRZD transmission.

The results in my research from analyzing the actions that SSI farmers choose are similar

to those of other studies and reports. For example, the great number of SSI farmers in my

research who indicated they vaccinate domestic animals is in keeping with local veterinary

reports which indicate that vaccinations of domestic animals against key WRZD (e.g., H5N1) are

maintained (Nguyen, 2014). For some other actions, however, my research results differ from

those of other studies. For example, most of the participating Thai Binh SSI farmers in my

research (81%) indicated they had adequate household sanitation based on their subjective

assessment, while only one-third (35%) of An Giang farmers estimated they did. This differs

from the Ministry of Health’s (MoH) statistics which showed that few households (18.5%) had

adequate household sanitation (MoH, 2006). Some possible explanations for this difference in

results include that 1) SSI farmers in my research subjectively assessed their household

sanitation conditions while the MOH’s experts objectively assessed the household sanitation

conditions of farms, and 2) my research target groups (i.e., SSI farmers) were more specific than

those who were part of the MoH research (i.e., general rural farmers). However, this difference

gives us some ideas about possible gaps between the perceptions and knowledge SSI farmers

of 340

have relating to adequate household sanitation – i.e., SSI farmers may think they have adequate

household sanitation but this may not be the case.

5.4.1.2. SSI farmers’ broad engagement in mitigating strategies to reduce WRZD transmission

The results of SSI farmers’ engagement in implementing mitigating strategies to reduce

WRZD transmission are interesting for two reasons. First, they show that more than half of SSI

farmers do engage in the mitigating strategies to reduce WRZD transmission (Table 5.2), each of

which combines some specific actions to reduce the transmission of WRZD. Again, this reflects

the complexity of the interface between livestock production, on-farm water quality, and public

health issues. Second, although it is necessary to look at each mitigating strategy as a

combination of various actions to reduce the transmission of WRZD (Carr and Bartram, 2004;

WHO, 2004), I was unable to find any studies that explored the ways SSI farmers engage in

mitigating WRZD transmission as a collection of actions. Thus, the results of my research could

be the basis for future studies regarding further measurements of SSI farmers’ mitigating

strategies to reduce WRZD transmission (e.g., an objective assessment of farmers’ engagement

and/or adoption of mitigating strategies to reduce transmission of WRZD).

5.4.2. Factors associated with SSI farmers’ engagement in mitigating strategies to reduce


As indicated earlier in this chapter, the main objective of my research was to explore

factors that are associated with SSI farmers’ engagement in mitigating strategies to reduce the

transmission of WRZD (Table 5.1). Therefore, in this section I focus on discussing the factors

that are associated with SSI farmers’ engagement in the mitigating strategies. However, prior to

discussing associations, it is necessary to note that, compared to other regressions (e.g., linear

of 340

regression or logit regression), the interpretation of probit coefficients and marginal effects is

less straightforward.

A negative coefficient indicated that an increase in an explanatory variable is associated

with a decrease in the predicted probability of the dependent variable. On the other hand, a

positive value coefficient indicated that an increase in an explanatory variable is associated with

an increase in the predicted probability of the dependent variable. A marginal probability effect

at the means presented the rate of change in the probability of a farmer engaging in a mitigating

strategy for a per unit change in an explanatory variable, holding all other explanatory variables

at their means. More specifically, for a continuous explanatory variable (Xi), a marginal

probability effect was a rate of change in the probability of a binary dependent variable (i.e.,

engagement in a mitigating strategy), given a unit change in that continuous explanatory variable

at mean values of other explanatory variables. This rate of change was expressed in a percentage.

For a binary explanatory variable (Xi) with two values (i.e., “0” or “1”), a marginal probability

effect was the difference between the probability of a dependent variable given Xi = 0 and the

probability of that dependent variable given Xi = 1, holding all other explanatory variables

constant at their means. In addition to indicating significant levels and signs of the associations

between the explanatory variables on SSI farmers’ engagement in mitigating strategies, marginal

probability effects show the change in magnitudes of the associations.

A standard maximum likelihood estimating procedure was used to estimate the individual

coefficients of the explanatory variables (Greene, 2008a). With respect to how well the probit

model fits the data, a direct assessment of the effectiveness of the probit model in modeling the

mitigating strategies was used. In the direct assessment, the probit model fits were measured by

of 340

computing predicted probabilities of SSI farmers’ engagement in mitigating strategies, and then

comparing it with the actual engagement in mitigating strategies. Given the non-linear feature of

the probit model, marginal probability effect analysis is normally used to have a more direct

interpretation of the coefficients of the explanatory variables (Greene, 2008b). The goodness-of-

fit measured by model chi-square, McFadden's R2, and percentages of correct predictions of all

probit models (concordance) for the four key mitigating strategies indicated that selected

explanatory variables in the probit models contributed to explaining the variation in SSI farmers’

engagement in the mitigating strategies.

Overall, the results in this chapter supported the hypothesis that SSI farmers’

socioeconomic status and their perceptions of risk factors for WRZD transmission are associated

with their engagement in mitigating strategies to reduce the transmission of WRZD. The

following section will explore in detail how the results of each mitigating strategy supported the

hypothesis.

5.4.2.1. Factors associated with SSI farmers’ engagement in LM as a mitigating strategy to

reduce WRZD transmission

The factors significantly associated with SSI farmers’ engagement in a LM mitigating

strategy for reducing WRZD transmission were related to the demographics of farm households,

livestock production, SSI farmers’ perceptions, and the quality of water on SSI farms. In the

following paragraphs, some of the most important factors that are significantly associated with

SSI farmers’ engagement in a LM mitigating strategy will be discussed.

SSI farmers’ years of attending school, as expected, have a positive effect on their

engagement in a LM mitigating strategy. For those SSI farmers who have an additional year of

attending school, the marginal probability effects on their engagement in a LM mitigating

of 340

strategy increases to reduce transmission of WRZD. Although I did not find studies that

specifically explored associations between education and SSI farmers’ engagement in a LM

mitigating strategy, some studies of SSI farmers in Vietnam found education to be an important

factor in influencing their choices for adaptive climate change measures (Le et al., 2014b). The

number of livestock (i.e., chickens and pigs) on farms also has a positive effect on SSI farmers’

engagement in a LM mitigating strategy. This means that as a SSI farmer gains one more

chicken or pig, the probability of taking on a LM mitigating strategy increases, assuming all

other explanatory variables are held constant at their means. Therefore, enhancing rural

education and encouraging larger scale production of chickens and pigs may contribute to

improving SSI farmers’ engagement in LM to reduce WRZD transmission.

The negative sign of the association between the log42 number of E.coli cfu in domestic

water and SSI farmers’ engagement in a LM mitigating strategy to reduce WRZD transmission

means that an SSI farmer has less probability of taking on a LM mitigating strategy for one

greater log number of E.coli cfu in domestic water assuming all other explanatory variables are

held constant at their means. Research in the U.S. also reported an association with a negative

sign between the log E. coli cfu number in farm water and farm management (Park et al., 2014).

However, caution is needed when comparing my research and the U.S. research due to the

differences in the context of the two studies (e.g., different operational definitions of farm

management). It appears that a high E. coli cfu count in domestic water may provide a warning

sign of a decrease in farmers’ engagement in LM strategy.

42 Natural logarithm of the number of E.coli cfu in domestic water variable

of 340

5.4.2.2. Factors associated with SSI farmers’ engagement in SWP as a mitigating strategy to


As indicated in the results section, 11 explanatory variables/factors were significantly

associated with SSI farmers’ engagement in a SWP mitigating strategy. These factors were again

related to the demographics of farm households, livestock production (e.g., number of pigs), and

SSI farmers’ perceptions. The results indicate that SSI farmers who perceived costs and peer-

pressure as barriers to their engagement in risk mitigating strategies were less likely to engage in

a SWP mitigating strategy compared to those who did not. These results agree with other studies

that have shown the associations between economic factors and SSI farmers’ behaviour (Duong

et al., 2004; Keraita et al., 2010). These results also confirmed my expectation, prior to the

research, of associations with negative signs between economic factors such as costs and SSI

farmers’ behaviours. This mean that a high cost related to a mitigating strategy may prevent SSI

farmers from engaging in that strategy. Conversely, SSI farmers with more years of attending

school, more years of farming, or more pigs were more likely to engage in a SWP mitigating

strategy. Thus, economic development policies and programs targeting SSI farmers would

contribute to more fully engaging them in protecting sources of water to mitigate WRZD

transmission. Moreover, SSI farmers with less formal education and fewer years of farming may

need more support than other SSI farmers in order to engage them more in source water

protection to mitigate WRZD transmission.

Discrete factors significantly associated with SSI farmers’ engagement in a SWP

mitigating strategy include the following: men playing a major role in family health on farms;

the perceived susceptibility to AI from contacting untreated wastewater; one’s perceived self-

efficacy in taking on source water protection actions; advice from health workers as a trigger;

of 340

perceived pressure from peer SSI farmers as a trigger; and perceived worry about being affected

by HPAI as a trigger. As SSI farmers move from “No” to “Yes” in these discrete factors, the

corresponding probabilities of taking on SWP as a mitigating strategy increases.

No other studies were found that specifically explored factors that are associated with SSI

farmers’ engagement in SWP. However, social factors were reported to influence a farmer’s

commitment to pro-environment action in general (Michel-Guillou and Moser, 2006), and

suggestions were made with respect to the need to study triggers of farmers’ adoption of

mitigating strategies (Keraita et al., 2010). I suggest that rural policies and programs in Vietnam

addressing gender equality, farmers’ perceptions of health risk factors, and community health

workers’ capacity should be prioritized to help improve source water protection among SSI

farmers in mitigating WRZD transmission.

5.4.2.3. Factors associated with SSI farmers’ engagement in POU as a mitigating strategy to


The factors significantly associated with SSI farmers’ engagement in a POU mitigating

strategy to reduce WRZD transmission were also related to geographic regions, demographics of

farm households, SSI farmers’ perceptions, and the quality of water on SSI farms. Some of these

factors will be discussed in the following paragraph.

It was not surprising that factors positively and significantly associated with SSI farmers’

engagement in a POU mitigating strategy were related to province, having children in a

household, gender roles in livestock production, perceptions, and advice from health workers.

The probabilities for taking on a POU mitigating strategy increase among SSI farmers with these

of 340

factors compared to those who did not. Meanwhile, income per person per year43, the perceived

safety of untreated drinking water, the perceived cost as a barrier to taking actions, and the log44

number of E.coli cfu in domestic water, as expected, were significantly and negatively associated

with SSI farmers’ engagement in POU mitigating strategies to reduce transmitting WRZD. A one

unit increase in each of these factors, holding other factors constant at their means, decreases the

probability of taking on a POU mitigating strategy. This means, for example, that when a farm

becomes poor (i.e., move up from code 0 to code 1), the probability of that farm taking on a POU

mitigating strategy decreases. Policies and programs that address POU strategies in mitigating

WRZD transmission among SSI farmers should, for example, tailor their programs by province

or combine them with poverty reduction programs.

5.4.2.4. Factors associated with SSI farmers’ engagement in WSTD as a mitigating strategy to


A number of factors relating to the socioeconomic status of SSI farm households, number

of livestock produced, SSI farmers’ perceptions, and the quality of water on SSI farms were

associated with SSI farmers’ engagement in a WSTD mitigating strategy. Some of these factors

are discussed in detail in the following paragraphs.

Economic factors such as income or the perceived cost of taking mitigating actions as a

barrier to taking action once again are associated with SSI farmers’ engagement (or not) in a

WSTD mitigating strategy to lessen WRZD transmission. This means SSI farmers who were

43 The variable of income per person per year was dichotomized using the cut-off income level for poverty set by the

Vietnamese government (i.e., less than 4.8 million VND per person per year – equivalent to $244 CAD per person

per year) at the exchange rate of $ CAD = 19,700 on 5th December 2013. An income per person per year below the

cut-off income level (i.e., a poor farm) was coded as 1 and 0 if otherwise (i.e., a not poor farm). I should have named

this variable as DPoverty instead of DIncome to avoid confusion. 44 The variable E.coliDom was transformed into LogE.coliDom (i.e., number of E.coli cfu in domestic water) using

a natural log transformation

of 340

living under the poverty line or who perceived costs as barriers to their mitigating strategies have

lower probabilities of engaging in a WSTD mitigating strategy compared to those who were not

poor or did not perceive cost to be a barrier. These results, as expected, align with other studies

that have shown the associations between economic factors and SSI farmers’ behaviours (Duong

et al., 2004; Keraita et al., 2010).

On the other hand, an additional year of attending school, one more pig, or one greater

log number of E. coli cfu in drinking water are positively significantly associated with SSI

farmers’ engagement in a WSTD strategy or for increasing the corresponding probabilities of

SSI farmers engaging in a WSTD mitigating strategy. Moreover, compared to a SSI farm that

has shared responsibility for family health between men and women, a farm with women

primarily responsible for family health increases the probability of engaging in a WSTD strategy.

Therefore, policies and programs addressing SSI farmers’ WSTD strategies to mitigate WRZD

transmission should prioritize issues such as poverty reduction and gender equality.

5.4.2.5. Common factors associated with SSI farmers’ engagement across four mitigating

strategies to reduce WRZD transmission

It is interesting to notice some factors that were consistently associated with the four

mitigating strategies. For example, perceived self-efficacy in engaging in a LM mitigating

strategy and advice from health workers were positively significantly associated with all four

mitigating strategies. Six factors were significantly associated with three mitigating strategies:

years of attending school, female farmers playing the main roles in family health, the perceived

susceptibility to HPAI from using untreated wastewater, the number of pigs, cost as a barrier to

taking action, and peer pressure. This consistency suggests that in order to improve SSI farmers’

engagement in mitigating strategies to reduce WRZD transmission, it may be important to

of 340

consider the above significant factors in developing water-related programs and policies

targeting SSI farmers.

Other studies also suggested that when designing programs and developing policies with

respect to the behavioural actions taken by rural farmers, it may be important to consider rural

farmers’ engagement in strategies to mitigate infectious diseases (OECD, 2012). In addition to

consistency, there were mixtures of factors that were positively and negatively associated with

SSI farmers’ mitigating strategies. These mixtures, as expected, are supported by previous

studies which indicated that demographic factors and traditional practices may facilitate or

hinder their actions (OECD, 2012), and only focusing on health-related factors is insufficient for

SSI farmers to engage in a mitigating strategy for disease transmission (Keraita et al., 2010).

5.5. Conclusions

In this chapter, I have provided the background, the methods, and the corresponding

results for my research addressing SSI farmers’ choices of actions as well as their broad

engagement in mitigating strategies to reduce WRZD transmission. My research results from the

previous chapters were used in this chapter to model factors that are associated with SSI farmers’

engagement in individual mitigating strategies to decrease WRZD transmission.

SSI farmers in Thai Binh and An Giang indicated that they have a wide range of actions

to choose from to reduce WRZD transmission. These actions were used to provide a more in-

depth understanding of SSI farmers’ engagement in four key mitigating strategies to decrease

WRZD transmission as recommended by the WHO. The following strategies were included:

livestock management; source water protection; water storage, treatment, and distribution; and a

point of use/household strategy. The research in this chapter has modeled factors associated with

of 340

SSI farmers’ engagement in these mitigating strategies. The results, showing a combination of

various factors that are associated with SSI farmers’ engagement in each of the mitigating

strategies, have confirmed my study’s hypothesis that SSI farmers’ socioeconomic status and

their perceptions of risk factors for WRZD transmission are associated with their engagement in

mitigating strategies to reduce the transmission of WRZD.

In conclusion, policies and programs relating to rural water and public health and the

mitigating strategies to reduce WRZD transmission should focus on a combination of factors

such as SSI farmers’ perceived self-ability to take livestock management actions as well as to

receive advice from health workers. It is also important to pay attention to other factors that

significantly and positively contribute to three out of the four mitigating strategies to lessen

WRZD transmission including education, women’s roles in family health, perceived

susceptibility to HPAI from using untreated wastewater, pig production, and peer pressure on

SSI farmers. Additionally, attention must be paid to economic factors such as perceived cost as a

barrier, which was the only factor that was significantly and negatively associated with SSI

farmers’ engagement across the four mitigating strategies of reducing the transmission of

WRZD.

A solid understanding SSI farmers’ choice of actions and factors that are associated with

their engaging in (or not engaging in) mitigating strategies to decrease WRZD transmission has

important implications for future research, interventions, and policy advocacy. Such

understanding will help in changing farmers’ behaviours, reducing rural poverty, implementing

effective water policies and mixed agriculture policies, as well as improving animal and

environmental health services in Vietnam.

of 340

of 340

References

Abdul, H. A., Ayamga, M., and Donkoh, S. A. (2014). Smallholder adoption of soil and water

conservation techniques in Ghana. African Journal of Agricultural Research, 9, 539-546.

Adesina, A. A. and Baidu, F. J. (1995). Farmers' perceptions and adoption of new agricultural

technology: evidence from analysis in Burkina Faso and Guinea, West Africa.

Agricultural Economics, 13, 1-9.

Becker, M. H. (1974). The Health belief model and personal health behavior. Thorofare, N.J. :

Slack, c1974.


Bloom, D. E., Cafiero, E. T., and Chu, M. (2013). Setting health priority: strategy versus tactics.

PSI Impact Managzine.

Brewer, N. T., Chapman, G. B., Gibbons, F. X., Gerrard, M., McCaul, K. D., and Weinstein, N.

D. (2007). Meta-Analysis of the Relationship Between Risk Perception and Health

Behavior: The Example of Vaccination. Health Psychology, 26, 136-145.




Publishing.

Cramb, R. A. (2005). Farmers' strategies for managing acid upland soils in Southeast Asia: An

evolutionary perspective. Agriculture, Ecosystems and Environment, 106, 69-87.


of 340

Dufour, A. (2004). Water-related zoonotic disease impacts - geographical prevalence. In

Waterborne Zoonoses: Identification, Causes, and Control (pp. 91-150). London: WHO

and IWA.



Greene, W. H. (2008a). Maximum Likelihood Estimation. In Econometric Analysis (6th ed.,

Pearson Prentice Hall.

Greene, W. H. (2008b). Models for Discrete Choice. In Econometric Analysis (6th Edition ed.,

Pearson Prentice Hall.

Hall, D. C., Knight, T. O., Coble, K. H., Baquet, A. E., and Patrick, G. F. (2003). Analysis of

Beef Producers' Risk Management Perceptions and Desire for Further Risk Management

Education. Review of Agricultural Economics, 25, 430-448.




JAE (2014). Author Guidelines: Online Manuscript Submission. Wiley Online Library [On-line].

Available: http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1477-

9552/homepage/ForAuthors.html





http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1477-9552/homepage/ForAuthors.html

http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1477-9552/homepage/ForAuthors.html

of 340


Ltd.

Le, D. H., Li, E., Nuberg, I., and Bruwer, J. (2014a). Farmers's assessments of private adaptive



Le, D. H., Li, E., Nuberg, I., and Bruwer, J. (2014b). Farmers' assessments of private adaptive





Marteau, T. M., Hollands, G. J., and Fletcher, P. C. (2012). Changing Human Behavior to

Prevent Disease: The Importance of Targeting Automatic Processes. Science, 337, 1492-

1495.

Mekoya, A., Oosting, S. J., Fernandez, R. S., and Akke, V. J. (2008). Farmers' perceptions about

exotic multipurpose fodder trees and constraints to their adoption. Agroforestry Systems,

73, 141-153.

Michel-Guillou, E. and Moser, G. (2006). Commitment of farmers to environmental protection:

From social pressure to environmental conscience. Journal of Environmental Psychology,

26, 227-235.




of 340

MoH (2006). The survey on environmental sanitation in rural Vietnam Hanoi: Ministry of

Health.

Neal, D., Vujcic, J., Hernandez, O., and Wood, W. (2014). Handwashing and the Science of

Habit. Sanitation Updates [On-line]. Available:

https://sanitationupdates.files.wordpress.com/2014/12/neal-handwashing2014.pdf

Negatu, W. and Parikh, A. (1999). The impact of perception and other factors on the adoption of

agricultural technology in the Moret and Jiru Woreda (district) of Ethiopia. Agricultural

Economics, 21, 205-216.

Nguyen, V. D. (2014). Results of the Control and Prevention of Animal Diseases in early 2014

Thai Binh: Thai Binh Department of Veterinary.

Nguyen, T. D., Vo, T. T., and Vu-Khac, H. (2011). Virulence factors in Escherichia coli isolated

from calves with diarrhea in Vietnam. Journal of Veterinary Science, 12, 159-164.

Nguyen, V. Q. (2004). The role of Salmonella and E. coli in diarrhea syndrome of cattle and

calves in provinces of Central Vietnam and the production of antibodies for treatment of

diarrhea disease caused by Salmonella and E. coli. National Institute of Veterinary

Research, Hanoi.

OECD (2012). Farmer Behaviour, Agricultural Management and Climate Change. OECD

Publishing.

Park, S., Navratil, S., Gregory, A., Bauer, A., Srinath, I., Szonyi, B. et al. (2014). Farm

Management, Environment, and Weather Factors Jointly Affect the Probability of

Spinach Contamination by Generic Escherichia coli at the Preharvest Stage. Appl Environ

Microbiol, 80, 2504-2515.

of 340

Payne, N., Jones, F., and Harris, P. (2002). The impact of working life on health behavior: The

effect of job strain on the cognitive predictors of exercise. Journal of Occupational

Health Psychology, 7, 342-353.

Rabbi, S. and Dey, N. (2013). Exploring the gap between hand washing knowledge and practices

in Bangladesh: a cross-sectional comparative study. BMC Public Health, 13, 89.

Rechenburg, A. and Herbst, S. (2006). Hygienic Aspects of Sanitation and Water in the Rural

Areas of the Mekong Delta, Vietnam (Rep. No. "Prosperity and Poverty in a Globalised

World Challenges for Agricultural Research"). University of Bonn, Institute for Hygiene

and Public Health, Germany.

Rosenstock, I. M., Strecher, V. J., and Becker, M. H. (1988a). Social Learning Theory and the


Rosenstock, I. M., Strecher, V. J., and Becker, M. H. (1988b). Social Learning Theory and the


Stata (2013). Stata Users' Guide Release 13. Stata: Data Analysis amd Statistical Software [On-

line]. Available: http://www.stata.com/manuals13/u.pdf




Van Mele, P. and Cuc, N. T. T. (2001). Farmers' Perceptions and Practices in Use of

Dolichoderus thoracicus (Smith) (Hymenoptera: Formicidae) for Biological Control of

Pests of Sapodilla. Biological Control, 20, 23-29.

http://www.stata.com/manuals13/u.pdf


of 340

Vindigni, S. M., Riley, P. L., and Jhung, M. (2011). Systematic review: handwashing behaviour

in low- to middle-income countries: outcome measures and behaviour maintenance.

Tropical Medicine and International Health, 16, 466-477.


WHO (2012). Health Risk Perception World Health Organisation Western Pacific Region.

of 340

Chapter Six: Conclusion

6.1. Overview

This dissertation has presented the original results of my Ph.D research, which presents

the research story of a commonly used mixed agricultural model in the provinces of Thai Binh

and An Giang in Vietnam. This research story began with a developed socioeconomic profile of

the small-scale integrated (SSI)45 farming model in these two provinces followed by an

assessment of SSI farmers’ perceptions of risk factors for transmitting WRZD46 and their choice

of actions to mitigate the transmission of WRZD. The research story also included the results of

determining the sources of water used on farms and an objective assessment of the quality of

water used on farms. Furthermore, the research story presented factors that are associated with

the quality of water on SSI farms and factors that are connected to SSI farmers’ engagement in

mitigating strategies to reduce the transmission of WRZD.

The dissertation is structured using a manuscript-based format. Besides the introduction

and conclusion chapters, I presented the research results as a collection of four manuscript-based

chapters that were prepared as a basis for submission to peer-review journals preferably in the

field of agricultural economics, water quality, and public health. Although each manuscript-

based chapter is a stand-alone manuscript, I arranged the chapters of this dissertation in a logical

and continuous order that presented the results in answer to my overall research questions. This

45Small-scale integrated (SSI) farming is defined as farming practices in which farmers combine some forms of

livestock and are directly or indirectly dependent on agriculture (Hall et al., 2006; Irini and Rapsomanikis, 2005;

Pica-Ciarmarra et al., 2011). 46 In this research, I used the WHO’s definition of WRZD as diseases in humans that come from animals and are

related to water such as Highly Pathogenic Avian Influenza, some parasites, or diseases caused by a range of

bacteria (e.g., pathogenic strains of Escherichia coli) (WHO, 2004). Criteria for determining WRZD include: 1) the

pathogen must spend part of its life cycle within one or more animal species; 2) it is probable or conceivable that

some life stage of the pathogen will enter water; and 3) transmission of the pathogen from animals to humans must

be through a water-related route (Moe, 2004)

of 340

presentation began with an understanding of the profile of the SSI farming model followed by

SSI farmers’ perceptions of risk factors for the transmission of WRZD and their on-farm water

quality, and ended with an understanding of factors that are associated with both on-farm water

quality and farmers’ mitigating strategies to reduce the transmission of WRZD.

In this chapter, I integrate the research results presented in the four manuscript-based

chapters with an emphasis on how the chapters build upon each other. The integration begins

with a summary of the literature review concerning the SSI farming model (i.e., Section 6.2)

followed by a summary of the research questions, hypotheses, and methods (Section 6.3). In

section 6.4, I summarize the contributions of my research to scientific knowledge with respect to

the SSI farming model, policies, and practices in the context of rural water quality and public

health. Section 6.5 provides a review of challenges and limitations of the research. The last two

sections (Section 6.6-7) provide overall recommendations for future research and policy

advocacy with respect to the SSI farming model, rural water quality and public health, and my

final thoughts respectively.

6.2. Summary of the Literature Review

The SSI farming model is a commonly used model for mixed agricultural practice in

Vietnam and has been promoted as a model in rural areas in Vietnam for a couple of decades.

This model followed a policy shift from a centrally planned and subsidized system to a market

oriented system in Vietnam, and it will continue to be an important part of the agricultural

system in Vietnam in the future. In this model, farmers use an integrated farm management

system to improve their farming practices and to achieve high economic efficiency. The model

has proven to be successful in helping farmers make efforts to optimize their on-farm resources

and to generate food and income. In this model, there are interactions among farmers, their

of 340

animals, and the water environment on SSI farms. However, these interactions have the potential

for adverse health impacts such as the transmission of WRZD. My review of the literature found

no studies that explore factors that are associated with on-farm water quality and farmers’

engagement in mitigating strategies to reduce the transmission of WRZD. However, I have

studied these factors in my research using the following research questions, hypothesis, and

methods.

6.3. Research Questions, Hypotheses, and Methods

The results of my research have addressed the following overall research questions:

1. What is the socioeconomic profile of SSI farmers in the Thai Binh and An Giang

provinces of Vietnam?

2. Is the socioeconomic profile of SSI farmers in the provinces of Thai Binh and An Giang

consistent with an EcoHealth approach in the context of SSI farming?

3. What are the perceptions of SSI farmers concerning the risk factors for transmitting

WRZD in the provinces of Thai Binh and An Giang?

4. What sources of water do SSI farmers use for drinking and domestic purposes47 and what

are the corresponding basic indicators48 of water quality for those sources in the

provinces of Thai Binh and An Giang?

5. How does the quality of on-farm water that SSI farmers in the provinces of Thai Binh and

An Giang use for drinking and domestic purposes relate to their perceptions about both

water quality and risk factors for transmitting WRZD?

47 In my research, drinking water is defined as “water used for direct drinking or food processing”. Domestic water

is defined as “water use for domestic purposes but nor for direct drinking or processing food” (MoH, 2009b; MoH,

2009a) 48 The WHO recommends Escherichia coli (E. coli) colony forming unit (cfu), turbidity, and pH as basic microbial

and related indicators for assessing rural water quality (WHO, 2012)

of 340

6. Do SSI farmers in the provinces of Thai Binh and An Giang engage in mitigating

strategies to reduce transmitting WRZD?

7. What factors are associated with SSI farmers’ engagement in mitigating strategies to

reduce the transmission of WRZD in the provinces of Thai Binh and An Giang?

Further to the description and exploration of the profile of SSI farms and farmers’

perceptions of risk factors for transmission of WRZD in Chapter Two and Three respectively,

four key hypotheses were formulated and tested in Chapter Four and Five. These hypotheses

were:

1. The frequencies of use of water sources for drinking and domestic purposes among

farmers in Thai Binh are different from those in An Giang.

2. The sources of water use for drinking and domestic purposes49 in both Thai Binh and An

Giang farms are of low quality.

3. SSI farmers’ socioeconomic status and perceptions of risk factors for transmission of

WRZD are associated with the microbial content of water on farms.

4. SSI farmers’ socioeconomic status and their perceptions of risk factors for transmission

of WRZD are associated with their engagement in mitigating strategies to reduce the

transmission of WRZD.

In my research, I have used a combination of qualitative and quantitative methods to

describe the profile of the SSI farmers and to explore their perceptions about risk factors for and

actions needed to reduce the transmission of WRZD. Furthermore, I used various modelling

techniques to explore associating factors of on-farm water quality (i.e., multiple linear regression

49 Drinking water is defined as “water used for direct drinking or processing food” (MoH, 2009b). Domestic water is

defined as “water used for domestic use but not for direct drinking or processing food” (MoH, 2009a).

of 340

and seeming unrelated regression technique) and associating factors for farmers’ engagement in

mitigating the transmission of WRZD (i.e., Probit regression).

6.4. Contribution to Knowledge

6.4.1. An understanding of SSI farming model in the context of rural water and public health

This dissertation has described the profile of the SSI farming model in the provinces of

Thai Binh and An Giang in the context of rural water and public health. This description focused

on the socioeconomic profile of the SSI farms, SSI farmers’ perceptions of water quality and

their perceptions about risk factors for transmitting WRZD, SSI farmers’ choices of actions and

their engagement in mitigating strategies to reduce the transmission of WRZD, and the quality of

drinking water used on SSI farms.

The developed socioeconomic profile of SSI farmers in Thai Binh and An Giang in

Chapter Two helped to present a picture of the SSI farming model in the two provinces by

focusing on demographic characteristics, livestock production, and farm economics. Chapter

Two also provided insight into the SSI farming model with respect to understanding the

ecological and health implications of the EcoHealth approach to SSI farming and into reducing

the unintended of consequences of SSI farming. Chapter Three provided a more specific

understanding of the SSI farming model concerning SSI farmers’ perceptions of risk factors for

transmission of WRZD. Chapter Three also demonstrated the possible important roles these

perceptions can play in improving the health of farmers and their animals, and in providing

guidance for policy formulations. Chapter Four (the first part)50 presented an understanding of

the sources and the quality of water used on SSI farms. The low microbial and related quality of

50 The second part of Chapter Four, factors that are associated with the microbial quality of water used on SSI farms,

is summarized in the next section - Section 6.4.2

of 340

water for drinking compared to the guidelines of clean water supply, sanitation, and water quality

assessment of the WHO and the Government of Vietnam suggested possible implications for

rural water testing, and on-farm water public health training and management. The first part51 of

Chapter Five presented an understanding of SSI farmers’ choices of actions to reduce risk factors

for the transmission of WRZD. The following sub-sections (i.e., 6.4.1.1-4) provide a summary of

key contribution to knowledge this research study makes with respect to an understanding of the

profile of the SSI farming model in the context of rural water and public health.

6.4.1.1. A socioeconomic profile of the SSI farming practice in Thai Binh and An Giang

Participating SSI farmers in the study had limited education, but most of them had

attended primary school. Almost 80% of SSI farmers in both provinces had attended secondary

or lower schools. The average SSI family size was four people with fewer female farmers mainly

responsible for livestock production, fish production, and family health compared to male

famers. The proportion of the SSI farmers who lived in poverty is one in five. This proportion is

five times greater in An Giang compared to Thai Binh, and is two times greater than the general

poverty proportion in rural areas of Vietnam (GSO Vietnam, 2013) The mean annual number of

livestock produced per farm was as follows: 17 chickens, 87 ducks, 13 pigs, and two cattle.

Moreover, although SSI farmers were still practicing integrated livestock and aquaculture

production, they did not usually refer to this as a model of farming practice.

The SSI farming model demonstrated a truly complex environment in which interactions

take place between SSI farmers, their livestock, and the on-farm water environment. This

complexity was reflected in the following key results: SSI farmers used animals’ manure as

51 The second part of Chapter Five which discussed, factors that are associated with SSI farmers’ engagement in

mitigating strategies to reduce transmission of WRZD, are summarized in the next section - Section 6.4.2

of 340

fertilizer and food for plants and fish; livestock production can bring both income and food for

SSI farmers; and products from gardens can be used as food for both fish and animals, as well as

provide an income source for SSI farmers. There were also geographic variations to SSI farming.

Yet, the interactions that take place on SSI farms have potential linkages to increasing the risk

factors for transmitting WRZD. However, SSI farmers have also had only limited participation in

the design, implementation, and monitoring of policies relating to livestock production, rural

water supplies and sanitation although they are the owners and implementers of the SSI farming

model.

Gender and social equity among SSI farming appeared to be a concern where poverty

existed among the participating SSI farmers in the two provinces. SSI household farms had a

greater mean family size than the Vietnamese 2009 national average. Consequently, it is

understandable that farmers are finding additional income from off-farm seasonal work.

However, there are potential social challenges for SSI farmers when they look for off-farm

seasonal work given their limited education level, younger age, poor working conditions, and far

distance from home with limited social benefits.

The SSI farming model was initially introduced to enhance the sustainability of farmers’

livelihoods, but several issues may influence the sustainability of the model and the health of

farmers and their livestock. Examples of these issues include low income, off-farm work among

farmers, low water quality on farms, lack of farmers’ participation in influencing related public

policies, and lack of capacity to face challenges and competitiveness when Vietnam joins the

WTO. In reality, these issues may worsen the undesirable consequences of SSI farming instead

of improving them.

of 340

The description of the SSI farming model can be used as a basis for future studies of this

model. An EcoHealth approach is good for community development and there are examples of

applying the EcoHealth approach in academic research, policy advocacy, and community

development (Nguyen, et al.; Hall, Dinh, Vu, Tran, and Le, 2012). When using an EcoHealth

perspective and taking into consideration the increased transmission of WRZD, it is suggested

that the SSI farming model needs to be reviewed to ensure that it will not be a simple agricultural

economic model in which food and income are outputs. Although this SSI farming model

reflects some of the principles of ecosystem approaches to health (EcoHealth), the model should

be a more holistic EcoHealth-based model in which the health of both farmers and their livestock

are interwoven with on-farm production and the protection of the water environment, and where

health is considered an equally important output of the model. Understanding and improving the

SSI farming model using an EcoHealth approach may help with managing the risk factors

associated with the transmission of WRZD.

6.4.1.2. SSI farmers’ limited perceptions of risk factors for transmission of WRZD

Participating SSI farmers perceived the overall on-farm water conditions in their villages

to be adequate. They were also aware of the potential harm to health when using untreated

sources of water, and assessed water quality subjectively and primarily based on impressions and

previous experiences. Most of the participating SSI farmers considered drinking untreated water

and handling untreated domestic water as being a moderate to high risk factor for the

transmission of Highly Pathogenic Avian Influenza (HPAI), diarrhea, coliform bacteria, and

parasites. If infected, they believed that HPAI, diarrhea, and coliform bacteria would be severe or

very severe. Economic factors were perceived by most of the participating SSI farmers as both

barriers and benefits to taking mitigating actions against the transmission of WRZD. With

of 340

respect to perceived self-efficacy, most of the participating SSI farmers believed that they have a

moderate or high ability to conduct mitigating actions against the transmission of WRZD. The

top three triggers/cues for taking mitigating actions were media, worries about being infected,

and advice from health workers.

6.4.1.3. SSI farmers’ engagement in mitigating strategies to reduce transmission of WRZD

SSI farmers in Thai Binh and An Giang indicated they have wide differences in their

choice of actions to reduce the risk factors for transmitting WRZD. Farmers’ choices of actions

provided an in-depth understanding of their engagement in the four key mitigating strategies52 of

transmission of WRZD as recommended by the WHO. These strategies included livestock

management; source water protection; water storage, treatment and distribution; and point of

use/household strategies. More than half of the SSI farmers in the study did engage in the four

mitigating strategies to reduce the risk of transmitting WRZD that was associated with on-farm

water use.

All results of SSI farms in the context of rural water and public health provided ideas for

formulating hypotheses with respect to modeling factors that are associated with on-farm water

quality and farmers’ engagement in mitigating the risks of transmitting WRZD (i.e., the second

parts of Chapter Four and Chapter Five).

6.4.1.4.Low quality of water used on SSI farms for both drinking and domestic purposes

The microbial quality was low for on-farm water for both drinking and domestic use, and

most of the water samples contaminated with E. coli did not met the quality standards for

52 In my research, I defined a mitigating strategy to reduce transmission of WRZD as a plan that includes one or

more activities a SSI farmer could choose to reduce transmission of WRZD (see figure 5.1 and section 5.2.1 in

Chapter Five). An example of a mitigating strategy to reduce transmission of WRZD in my research is good

livestock management which may include actions or tactics (e.g., treat animal waste at source) that SSI farmers

could choose to reduce transmission of WRZD (Carr and Bartram, 2004; WHO, 2004).

of 340

drinking and domestic water. Although testing of E.coli in water is recommended to determine if

a water source has recently been contaminated with faeces, such faecal contamination of water is

often intermittent and might not be revealed in one assessment of a single sample. Therefore, in

addition to simple affordable tests for SSI farmers of E. coli in water, it is important to build

farmers’ capacity for inspecting and improving their household water sanitation and on-farm

water public health management. By doing so, SSI farmers will not only be better able to afford

to conduct the tests, but also to know the appropriate time and circumstances for taking

corresponding measures following water testing results.

6.4.2. An understanding of factors associated with on-farm water quality and SSI farmers’

engagement in mitigating strategies to reduce the transmission of WRZD

The description of SSI farms in the context of rural water and public health were used to

model factors that are associated with the objectively assessed on-farm water quality and factors

that are associated with SSI farmers’ engagement in mitigating strategies to reduce transmission

of WRZD. The modelling results have possible implications for rural water supply and

sanitation, on-farm public health management, and for policy advocacy in improving the health

of farmers, their livestock, and their water environment.

6.4.2.1. Factors that are associated with on-farm water quality

With respect to drinking water quality, farmers who had greater income per person per

year and were not satisfied with the quality of their current sources of drinking water were likely

to have better stored rainwater (i.e., lower E. coli cfu in water) compared to those who had lower

income and were happy with the current quality of their drinking water. In terms of piped water

for drinking, farmers who thought that using untreated water for drinking brings susceptibly to

parasite infection were likely to have better piped water (i.e., lower E. coli cfu in piped water)

of 340

compared to those who thought that using piped water does not bring susceptibly to parasite

infection. With regard to well water for drinking, it is likely that farms that were headed by

females, involved in fish production, and perceived susceptibility to parasites when using

untreated water for drinking had better piped water (i.e., lower E. coli cfu in pipe water). This

was in comparison to farms headed by men and farmers who thought that using well water did

not bring susceptibly to parasitic infection. The quality of river water for drinking was associated

with gender, years of attending school, and number of chickens produced.

In terms of domestic water, the microbial (E. coli) quality of river water used for

domestic purposes was associated with gender, occupations other than farming, number of pigs,

men primarily responsible for livestock production, and perceived susceptibility to parasites from

untreated domestic water. Years of attending school, numbers of chickens, perceptions of

susceptibility to HPAI from untreated domestic water, and satisfaction with current sources of

domestic water were associated with E. coli in pond water for domestic purposes.

The modeling of factors that are associated with the quality of on-farm water gave some

ideas of factors that may be relevant to explore further possible associating factors for farmers’

mitigating strategies to reduce the transmission of WRZD in the context of SSI farming. The

results of my research have indicated there is room for improving rural water supply and

sanitation, and on-farm water and health management through updates and revision of the water

regulatory framework/policies for SSI farms in Vietnam. The results of my research also showed

a need to consider rural water testing and training in improving on-farm water public health.

of 340

6.4.2.2. Factors that are associated with SSI farmers’ engagement in mitigating strategies to

reduce the transmission of WRZD

The four mitigating strategies included the following: livestock management; source

water protection; water storage, treatment and distribution; and point of use/household. More

than half of the SSI farmers in the study did engage in these four mitigating strategies to reduce

risk factors for the transmission of WRZD associated with on-farm water use. There were

combinations of various factors that contributed to SSI farmers’ engagement in each of the four

mitigating strategies to reduce transmission of WRZD. The significant factors that positively

contribute to SSI farmers engaging in the four mitigating strategies were their self-ability to take

livestock management actions and to receive advice from health workers. It is also important to

pay attention to other factors that contributed significantly and positively to three out of four of

the mitigating strategies to reduce the transmission of WRZD including years of attending

school, women’s role in family health, perceived susceptibility to HPAI from using untreated

wastewater, pig production, and peer pressure on SSI farmers. On the other hand, it is important

to recognize perceived cost as a barrier as it was the only one significant negative contributing

factor to SSI farmers engaging across the four mitigating strategies to reduce the transmission of

WRZD.

6.5. Limitations

Some limitations must be taken into consideration when interpreting my research results.

First, although my research had a relatively large sample size, my research only focused

on SSI farming in the context of the targeted river deltas. VAC is also a dynamic model with

much variation, even within a geographical region (Dang et al., 2005; Devendra and Thomas,

2002b). For example, seven VAC farming systems were commonly practiced in the Mekong

of 340

Delta (Dang et al., 2005). Furthermore, although some frameworks exist that identify SSI

farming (Dixon et al., 2004; Nguyen, 2014a), no universal criteria are in place to precisely define

SSI farming (Heidhues and Bruntrup, 2003). Thus, there may be considerable variation in

aspects such as the number of livestock, the volume of fish, and the volume of crops that

individuals produce who are identified as SSI farmers. These variations thus make my research

results more difficult to generalize for all regions. Therefore, for SSI farming in other regions of

Vietnam (e.g., irrigated lowlands or rain fed uplands), the more specific context of these regions

must be considered and/or further research may be needed before generalizing the results from

my research. Multiple criteria should be used to examine and interpret information relating to the

profile and on-farm production of SSI farming.

Second, in terms of study design, my research used a cross-sectional study design.

Therefore, factors relating to farmers’ perceptions, socioeconomic status, and farmers’ choices of

actions were assessed at a particular point-in-time. Thus, these factors as well as choices of

actions may change over time (Marteau et al., 2012; Vindigni et al., 2011). Further, the

association between these factors and farmers’ mitigating strategies was identified by giving the

existing farmers’ perceptions and mitigating strategies, which were based on their subjective

self-assessments, which may not meet the standard(s) for proper actions. Moreover, my research

did not objectively assess whether farmers have the correct knowledge about rural water quality

and transmission of WRZD or whether or not they undertake proper actions that meet a certain

standard. At the same time, gaps may exist between not only perceptions but also knowledge

(Rabbi and Dey, 2013), habit (Neal et al., 2014) and actions. Therefore, factors used in modeling

in my research as well as the corresponding directions and magnitude of associations may

change considerably when other factors such as standards, quality of mitigating strategies

of 340

applied, timeline, knowledge, and habits are applied or considered. Future studies using study

designs that take into account these other factors are recommended to understand better the

complexity of farmers’ engagement in strategies to mitigate the transmission of WRZD.

Finally, the four mitigating strategies used in this study were modelled separately with a

focus on perceptions of risk factors for the transmission of WRZD and farmers’ socioeconomic

status as possible explanatory variables of interest. However, farmers’ engagement in these

mitigating strategies may be more complex and may relate to each other because all mitigating

strategies take place on farms by farmers. Therefore, farmers’ engagement in one mitigating

strategy may be related to their engagement in other mitigating strategies. As a result, it may be

valuable to estimate farmers’ engagement in all mitigating strategies jointly by using a

multivariate regression analysis (e.g., multivariate probit modelling technique). The literature

review undertaken for this study found that no previous study explored factors that are associated

with SSI farmers’ mitigating strategies to reduce the transmission of WRZD with respect to on-

farm water issues. Consequently, the approach in my research was to focus on modeling

individual mitigating strategies separately by using a univariate probit regression technique.

Beyond the context of the research results reported in this dissertation, I plan to use more

advanced techniques (e.g., multivariate probit regression) to explore further the complexity of

water quality, public health, and farmers’ behaviours/decision making process in mitigating

strategies to reduce the transmission of WRZD.

6.6. Overall Recommendations

Use an Ecosystem approach to health in Vietnam in the context of SSI farming:

An EcoHealth approach is considered to be an excellent approach for community

development around the world and good examples of the application of the EcoHealth approach

of 340

exist in Vietnam. An EcoHealth approach considers both human and animal health as well as

environmental health through equitable and sustainable methods (IDRC, 2012). An EcoHealth

approach to the SSI farming model for both farmers and policy makers is recommended for

consideration in Vietnam. Revising the SSI farming model and related policies with respect to an

EcoHealth approach may ensure that the model will not be merely a simple agricultural

economic model in which food and income are outputs. Instead, it should be a more holistic

EcoHealth-based model in which the health of farmers and animals is interwoven with on-farm

livestock production and the protection of the environment, and these are considered to be an

equally important output of the model.

Implementing a transdisciplinary approach: The complexity of SSI farming requires

the knowledge and involvement of various disciplines such as veterinary health, public health,

water security, the private sector, and so on, to not only improve farming practice and generate

income but also to improve the health of farmers and their animals and to avoid polluting the

environment.

More active and engaged community participation: To optimize the limited water

resources available for farmers and to reduce the negative socio-economic and health impacts on

the rural population, it is important to enhance community participation to a greater extent not

only from policy makers but also from farmers themselves in the policy making process. This

enhancement could help to more explicitly address the issues relating to small-scale integrated

farming such as water, health, and small-scale livestock production.

Improve rural water quality: In designing rural water interventions and developing

rural water policies, it is important to consider the fact that SSI farmers depended on multiple

sources of water on farms for drinking and domestic purposes, and that there are differences in

of 340

frequencies of use between provinces (i.e., Thai Binh and An Giang). The microbial quality of

water used on SSI farms is also low and this challenge needs to be addressed. In addition to

having simple affordable tests of E. coli in water, it is also important to build the capacity of SSI

farmers and the government for inspecting and improving household water sanitation, and for

on-farm water public health management. By doing so, rural farmers will not only be able to

afford and conduct the tests, but also know when it is appropriate to take the test and what to do

with the test results. Therefore, any efforts to improve the microbial quality of water used on SSI

farms should focus on providing better education for SSI farmers, improving farmers’ income,

increasing awareness among SSI farmers of the potential harm caused by using untreated water,

and sharing the experiences of farms with fish production and farms that have had past

experiences with HPAI in poultry production.

Increase farmers ‘engagement in mitigating strategies to reduce WRZD

transmission: Multiple factors are associated with SSI farmers’ engagement in mitigating

strategies to reduce the transmission of WRZD. Policies and programs to increase farmers’

engagement in the four mitigating strategies to reduce transmission of WRZD should focus on

key factors such as improving SSI farmers’ perceived self-ability in taking livestock

management actions and increasing their access to advice from health/veterinary workers. It is

also important to improve farmers’ education, promote women’s role in family health, increase

awareness of susceptibility to HPAI from using untreated wastewater, encourage larger scale pig

production, and increase peer pressure among SSI farmers. Furthermore, it is important to

improve farmers’ income while also offering affordable services relating to rural water quality

and livestock management for SSI farmers.

of 340

6.7. Conclusions/Final Thoughts

This dissertation has provided insights into the SSI farming model in Vietnam with

respect to livestock production, on-farm water quality, and public health, and also into an

EcoHealth approach. Results of my research have potential implications for future research,

interventions, and policy advocacy with respect to rural water policies, mixed agriculture

policies, and delivery of animal and environmental health services. The EcoHealth approach has

been considered a valuable and sustainable approach for community development. In Vietnam,

there are some examples of the EcoHealth approach in academic research, policy advocacy, and

community development. The SSI farming model in Vietnam needs to be reviewed and studied

further using an EcoHealth approach and considering risks for increased transmission of WRZD.

An EcoHealth approach may help SSI farmers to enhance their public health water management

and turn the SSI farming model into a holistic integrated farming model, which not only

maximizes beneficial impacts but also reduces any adverse health impacts of the model. In such

a model, the health of farmers, animals and the environment should be interwoven with livestock

production and be considered an equally important output of the SSI farming model.

of 340

References


Publishing.

Dang, K. N., Le, T. D., Nguyen, V. S., and Verdengem, M. (2005). Development of "VAC"

Integrated Farming Systems in the Mekong Delta, Vietnam - A View of a System and a

Participatory Approach.

Devendra, C. and Thomas, D. (2002). Smallholder farming systems in Asia. Agricultural

Systems, 71, 17-25.

Dixon, J., Tanyeri-Abur, A., and Wattenbach, H. (2004). Framework for analyzing impacts of

globalization on smallholders. Food and Agriculture Organization of the United Nations

(FAO) [On-line]. Available: http://www.fao.org/docrep/007/y5784e/y5784e02.htm

GSO Vietnam (2013). General poverty rate by residence and by region and Year - Rural. GSO

Vietnam [On-line]. Available: http://www.gso.gov.vn/default_en.aspx?tabid=783



Heidhues, F. and Bruntrup, M. (2003). Subsistence agriculture in Central and Eastern Europe:

how to break the vicious circle? (vols. 22) Halle (Saale): Leibniz Institute of Agricultural

Development in Central and Eastern Europe.



http://www.fao.org/docrep/007/y5784e/y5784e02.htm

http://www.gso.gov.vn/default_en.aspx?tabid=783

of 340




Marteau, T. M., Hollands, G. J., and Fletcher, P. C. (2012). Changing Human Behavior to

Prevent Disease: The Importance of Targeting Automatic Processes. Science, 337, 1492-

1495.




MoH (2009). National technical regulation on domestic water quality (QCVN 02: 2009/BYT).



MoH (2009). National technical regulation on drinking water quality (QCVN 01: 2009/BYT).



Neal, D., Vujcic, J., Hernandez, O., and Wood, W. (2014). Handwashing and the Science of

Habit. Sanitation Updates [On-line]. Available:

https://sanitationupdates.files.wordpress.com/2014/12/neal-handwashing2014.pdf

Nguyen, M. T. N. (2014). Translocal Householding: Care and Migrant Livelihoods in a Waste-

trading Community of Vietnam's Red River Delta. Development and Change, 45, 1385-

1408.

of 340




Rabbi, S. and Dey, N. (2013). Exploring the gap between hand washing knowledge and practices

in Bangladesh: a cross-sectional comparative study. BMC Public Health, 13, 89.

Vindigni, S. M., Riley, P. L., and Jhung, M. (2011). Systematic review: handwashing behaviour

in low- to middle-income countries: outcome measures and behaviour maintenance.

Tropical Medicine and International Health, 16, 466-477.




of 340

APPENDIX A: THEORETICAL MODEL

of 340

APPENDIX B: QUESTIONNAIRES USED DURING THE FIELD STUDY

B.1. Questionnaires in Vietnamsese:

1. Mã số hộ nông dân trả lời bảng hỏi

2. Giới tính?

1. Nam

2. Nữ

3. Khác

3. Ngày tháng năm sinh của Quý vị là gì?

__________

4. Quý vị dân tộc gì?

1. Kinh

2. Khác: ghi rõ

5. Trình độ học vấn cao nhất của Quý vị là gì?

6. Gia đình Quý vị có bao nhiêu nhân khẩu?

of 340

7. Số nhân khẩu dưới 18 tuổi?

8. Quý vị đã làm nghề vừa chăn nuôi vừa thả cá, trồng trọt được bao nhiêu năm rồi?

9. Ngoài nghề nông, Quý vị còn làm nghề gì khác?

10. Nhìn chung, ai là người chịu trách nhiệm CHÍNH về chăn nuôi gia súc gia cầm, thả cá, và

vấn đề sức khỏe gia đình?

Nam Nữ Cả hai Không

chắc chắn

Từ chối

không trả

lời

Chăn nuôi gia súc gia cẩm ❏ ❏ ❏ ❏ ❏

of 340

Thả cá ❏ ❏ ❏ ❏ ❏

Sức khỏe gia đình ❏ ❏ ❏ ❏ ❏

11. Vị trí của Quý vị với chủ hộ là gì?

1. Chủ hộ

2. Vai trò khác

12. Quý vị nhìn nhận về điều kiện môi trường nước nói chung của thôn mình như thế nào?

1. Rất tốt

2. Tốt

3. Bình thường

4. Kém

5. Rất kém

6. Không có ý kiến

13. Quý vị có cho rằng nguồn nước chưa qua xử lý của quý vị (chưa lọc hoăc chưa đun sôi)

có an toàn để uống không?

1. Có

2. Không

14. Quý vị có xử lý nước uống và nước sinh hoạt của mình không?

Lọc Đun sôi Cả lọc

và đun

sôi

Cách

khác

Không

xử lý

Không

có ý

kiến

Nước uống ❏ ❏ ❏ ❏ ❏ ❏

Nước sinh hoạt ❏ ❏ ❏ ❏ ❏ ❏

15. Có vấn đề gì đã xảy ra trong quá khứ khiến Quý vị lọc, đun sôi (hoăc cả hai) nước uống?

1. Không

2. Có: hãy ghi rõ vấn đề gì_____________________________

of 340

16. Nếu có (lọc, đun sôi, cả hai) thì theo quý vị nước đã lọc và/hoặc đã đun sôi có an toàn để

uống không?

Có phải cả lọc và đun sôi Chỉ lọc Chỉ đun sôi Không

17. Hãy điền vào mức độ nguy hại cho sức khỏe của những nguồn nước sau (chưa lọc hoặc

chưa đun sôi).

Cực kỳ

hại

Nguy hại

cao

Nguy hại

vừa

Nguy hại

thấp

Không

nguy hại

Không

có ý kiến

Nước uống (chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

Nước sinh hoạt (chưa xứ lý) ❏ ❏ ❏ ❏ ❏ ❏

Nước ao hồ (chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

Nước đóng chai ❏ ❏ ❏ ❏ ❏ ❏

18. Bây giờ tôi muốn hỏi quý vị về cúm chim/cúm gia cầm: Quý vị đã bao giờ nghe đến cúm

chim/cúm gia cầm chưa?

1. Có, đã nghe đến cúm chim/cúm gia cầm

2. Không, chưa bao giờ nghe đến cúm chim/cúm gia cầm

19. Hãy điền vào khả năng nhiễm cúm chim từ việc sử dụng những nguồn nước sau (chưa

đun sôi hoặc chưa lọc)

Cực cao Cao Vừa Thấp Không

có khả

năng

Không

có ý kiến




Nước thải chăn nuôi, gia cầm

(chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

20. Nếu nhiễm cúm chim, Quý vị nghĩ mức độ nặng của bệnh sẽ như thế nào?

1. Rất nặng

2. Nặng

3. Vừa

of 340

4. Nhẹ

5. Không sao


21. Gia cầm của Quý vị đã từng bị tiêu hủy do dịch bệnh cúm gia cầm tại địa phương chưa?

1. Chưa từng bị hủy

2. Đã từng bị hủy

3. Không nhớ


22. Quý vị sẽ làm gì nếu phát hiện gia súc, gia cầm, chim bị ốm hoặc chết?

1.Bỏ qua

2.Tham khảo ý kiến các nông dân khác

3.Thông báo cho trưởng thôn

4.Chôn tiêu hủy

5.Báo cho cán bộ phụ trách thú y địa phương

6.Sử dụng như thức ăn (cho người hoặc gia súc gia cầm)

7.Không có ý kiến

8.Khác: đề nghị nêu rõ

23. Hiện tại gia cầm của Quý vị đã được tiêm phòng dịch cúm gia cầm chưa?

1. Tiêm rồi

2. Chưa tiêm


24. Quý vị đã bao giờ nghe đến vi trùng tồn tại trong nước như vi rút, vi khuẩn, giun sán

chưa?

1. Đã nghe qua

2. Chưa bao giờ

25. Hãy điền vào khả năng mắc bệnh tiêu chảy từ việc sử dụng những nguồn nước sau (chưa

lọc hoặc chưa đun sôi).

Rất cao Cao Vừa Thấp Không Không có ý kiến




Nước thải chăn nuôi, gia cầm

(chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

of 340

26. Nếu mắc tiêu chảy, Quý vị nghĩ mức độ nặng của bệnh sẽ như thế nào?

1. Rất nặng

2. Nặng

3. Vừa

4. Nhẹ

5. Không sao


27. Một trong những loại vi trùng có thể tìm thấy trong nước ở nông trại và đôi khi gây nguy

hại là E.coli. Đây là loại vi khuẩn nhỏQuý vị đã bao giờ nghe đến E.coli hay bệnh do

nhóm coliform gây ra chưa?

1. Có, chuyển sang câu kế tiếp và tiếp tục

2. Không, bỏ qua câu kế tiếp và tiếp tục

28. Điền khả năng nhiễn bệnh nhóm coliform từ nguồn nước sau (chưa đun sôi hay lọc).

Rất cao Cao Vừa Thấp Không Không có ý

kiến




Tiếp xúc nước thải chăn nuôi,

gia cầm (chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

29. Nếu Quý vị mắc bện từ nhóm coliform, mức độ nặng của bệnh sẽ thế nào?

1. Rất nặng

2. Nặng

3. Vừa

4. Nhẹ

5. Không sao


30. Một nguồn khác có thể gây bệnh từ nước của nông trại và đôi khi gây nguy hại là giun

sán.Qúy vị đã bao giờ nghe đến giun sán chưa?

1. Có, trả lời câu kế tiếp và tiếp tục

2. Chưa, bỏ qua 2 câu kế tiếp và tiếp tục

of 340

31. Điền khả năng nhiễm giun sán từ những nguồn nước sau (chưa đun sôi hoặc lọc).

Rất cao Cao Vừa Thấp Không Không có ý

kiến




Tiếp xúc nước thải chăn nuôi,

gia cầm (chưa xử lý) ❏ ❏ ❏ ❏ ❏ ❏

32. Nếu quý vị mắc bệnh giun sán, mức độ nặng nhẹ sẽ như thế nào?

1. Rất nặng

2. Nặng

3. Vừa

4. Nhẹ

5. Không sao


33. Mức độ quan trọng/nghiêm trọng của các vấn đề liên quan đến việc sử dụng nước trong

việc giảm nguy cơ mắc bệnh truyền nhiễm là gì?

Cực kỳ

quan trọng

Quan

trọng cao

Vừa Ít quan

trọng

Không quan

trọng

Không có ý

kiến

Nguồn nước ❏ ❏ ❏ ❏ ❏ ❏

Không dùng chung

nguồn nước người và

vật nuôi

❏ ❏ ❏ ❏ ❏ ❏

Khoảng cách nguồn

nước và vật nuôi ❏ ❏ ❏ ❏ ❏ ❏

Thiếu nước ❏ ❏ ❏ ❏ ❏ ❏

Lượng mưa ❏ ❏ ❏ ❏ ❏ ❏

Chất lượng ❏ ❏ ❏ ❏ ❏ ❏

Vấn đề khác ❏ ❏ ❏ ❏ ❏ ❏

34. Quan điểm hiện tại của Quý vị về tầm quan trọng của việc sử dụng nước đúng cách trong

việc giảm nguy cơ các bệnh truyền nhiễm cũng như cúm gia cầm như thế nào?

1. Rất quan trọng

2. Quan trọng

of 340

3. Quan trọng vừa

4. Quan trọng ít

5. Không quan trọng


35. Về tổng thể, mức độ hài lòng với nguồn nước sau mà Quý vị đang sử dụng là gì?

Hoàn toàn hài

lòng

Hài lòng

cao

Hài lòng

vừa

Ít hài

lòng

Không

hài lòng

Không có ý

kiến

Để uống ❏ ❏ ❏ ❏ ❏ ❏

Cho SXNN ❏ ❏ ❏ ❏ ❏ ❏

36. Xin hãy cho biết suy nghĩ của Quý vị về mức độ nguy cơ mắc các bệnh truyền nhiễm từ

những nguồn nước sau đây?

Nguy cơ

rất cao

Nguy cơ

cao

Nguy cơ

vừa

Nguy cơ

ít/thấp

Không

có nguy

cơ

Không

có ý kiến

Mưa ❏ ❏ ❏ ❏ ❏ ❏

Hồ ❏ ❏ ❏ ❏ ❏ ❏

Sông ❏ ❏ ❏ ❏ ❏ ❏

Kênh ❏ ❏ ❏ ❏ ❏ ❏

Giếng đào ❏ ❏ ❏ ❏ ❏ ❏

Giếng khoan ❏ ❏ ❏ ❏ ❏ ❏

Nguồn khác ❏ ❏ ❏ ❏ ❏ ❏

37. Những yếu tố nào dưới đây có tác động đến cách xác định chất lượng nước của Quý vị?

1. Dựa vào ấn tượng tạo ra bởi bất cứ một chất nào trong nước và nước

SXNN, mùi, vị, mầu

2. Theo kinh nghiệm

3. Theo quan niệm về rủi ro

4. Thái độ đối với việc ô nhiễm hóa chất đối với nước

5. Những biểu hiện liên quan đến nguồn nước

6. Quen với nguồn nước cụ thể

7. Sự tin tưởng vào nhà cung cấp vào nguồn nước, nước thải

8. Truyền thông

of 340


10. Các yếu tố khác: cụ thể

38. Quý vị có hoặc lấy thông tin về các bệnh truyền nhiễm từ nguồn nào dưới đây và mức độ

tin cậy của những nguồn nay ra sao?

Rất tin

cậy

Tin cậy

cao

Khá tin

cậy

It tin cậy Không

tin cậy

Không

có ý kiến

Ti vi ❏ ❏ ❏ ❏ ❏ ❏

Báo chí/ tờ rơi ❏ ❏ ❏ ❏ ❏ ❏

Loa đài ❏ ❏ ❏ ❏ ❏ ❏

Gia đình/bạn bè/nông dân khác ❏ ❏ ❏ ❏ ❏ ❏

Cán bộ/ trung tâm y tế ❏ ❏ ❏ ❏ ❏ ❏

Cán bộ trung tâm thú y ❏ ❏ ❏ ❏ ❏ ❏


39. Theo Quý vị, việc áp dụng biện pháp dự phòng có giúp chống lại các bệnh truyền nhiễm

không?

1. Không

2. Có, ví dụ như rửa tay

3. Có, ví dụ như cô lập nguồn nhiễm (người, động vật, nước)

4. Có, ví dụ như đeo các thiết bị bảo hộ (khẩu trang)

5. Có, ví dụ như tiêm phòng vắc- xin

6. Có, ví dụ như dùng thuốc

7. Có, ví dụ như nấu chín/không ăn thức ăn tươi sống


9. Biện pháp khác

40. Co lợi ích nào khác (ngoài sức khỏe) đã khuyến khích Quý vị hành động phòng chống

các bệnh truyền nhiễm? Đề nghị nêu rõ

of 340

41. Co lợi ích nào đã khuyến khích Quý vị KHÔNG hành động phòng chống các bệnh truyền

nhiễm lây qua đường nước có nguồn gốc từ động vật? Đề nghị nêu rõ.

42. Quý vị KHÔNG hành động phòng chống bệnh truyền nhiễm do các yếu tố nào sau đây?

1. Không đủ điều kiện kinh tế để thực hiện

2. Không biết cách (kỹ năng)

3. Không biết (hiểu biết)

4. Hàng xóm không hành động

5. Quá bận với những việc khác


7. Yếu tố khác

43. Yếu tố nào sau đây tác động/ảnh hưởng đến việc Quý vị ra quyết định thực hiện hành

động phòng chống các bệnh truyền nhiễm?

1. Hàng xóm đã làm tương tự

2. Khuyến cáo từ cán bộ sức khỏe cộng đồng, cán bộ thú y chuyên trách

3. Báo đài

4. không bận với những việc khác

5. Lo lắng về việc bị lây nhiễm hoặc bị ảnh hưởng bởi bệnh truyền nhiễm

6. Mất thu nhập


8. Yếu tố khác

44. Phần này, tôi sẽ phỏng vấn Quý vị về đặc điểm hoạt động chăn nuôi gia súc gia cầm, thủy

sản, và trồng trọt của Quý vị.Quý vị có những hoạt động sản xuất nông nghiệp nào?

1. Trồng lúa

2. Sắn

of 340

3. Tre

4. Ngô

5. Lạc/đậu

6. Rau

7. Hoa quả

8. Fish production

9. Livestock production

45. Ba nguồn thông tin quan trọng nhất quý vị nhận được về quản lý gia trại là gì? (điền số 1,

2, 3 of 9):

Nông dân khác __________

Thành viên gia đình __________

Báo in, tài liệu tuyên truyền __________

Đài, tivi __________

Cán bộ chuyên trách xã __________

Tập huấn __________

Công ty bán hàng nông nghiệp __________

Tổ chức phi chính phủ __________

Nguồn khác __________

46. Quý vị có loại hình chăn nuôi (gia súc, gia cầm, thả cá) nào

Ước tính số con

(bé/con) thương có một

vụ

Ước tính số con

(lớn/trưởng thành)

thường có một vụ

Ước tính số kg

thường có một vụ

Ghi chú

Thả cá

Gà

Vịt

Lợn

Dê

Gia cầm

khác

Gia súc

khác

Vật nuôi

khác

47. Đối với NƯỚC UỐNG, Quý vị dùng từ những nguồn nước nào và tần suất sử dụng

of 340

những nguồn này như thế nào?

Rất thường

xuyên

Thường

xuyên

Thỉnh

thoảng

Ít khi Không

bao giờ

Không có ý kiến

Ao/hồ ❏ ❏ ❏ ❏ ❏ ❏

Sông/kênh ❏ ❏ ❏ ❏ ❏ ❏

Nước mưa ❏ ❏ ❏ ❏ ❏ ❏

Nước máy ❏ ❏ ❏ ❏ ❏ ❏

Giếng đào ❏ ❏ ❏ ❏ ❏ ❏


Nước mua (đóng

chai/bình) ❏ ❏ ❏ ❏ ❏ ❏


48. Đối với nước SXNN, Quý vị dùng từ những nguồn nước nào và tần suất sử dụng những

nguồn này như thế nào?

Rất thường

xuyên

Thường

xuyên

Thỉnh

thoảng

Ít khi Không

bao giờ

Không có ý

kiến

Ao/hồ ❏ ❏ ❏ ❏ ❏ ❏

Sông/kênh ❏ ❏ ❏ ❏ ❏ ❏

Nước mưa ❏ ❏ ❏ ❏ ❏ ❏

Nước máy ❏ ❏ ❏ ❏ ❏ ❏

Giếng đào ❏ ❏ ❏ ❏ ❏ ❏


Nước mua ❏ ❏ ❏ ❏ ❏ ❏


49. Gia cầm của Quý vị có bị nhiễm cúm gia cầm trong đợt dịch năm 2004?

1. Có

2. Không

3. Không nhớ


50. Gia cầm của Quý vị có bị nhiễm cúm gia cầm trong vòng 12 tháng qua không?

1. Có

2. Không: Bỏ qua 2 câu kế tiếp

of 340

3. Không nhớ


51. Chất thải chăn nuôi của Quý vị có được xả trực tiếp (không qua xử lý) vào nguồn nước

dùng trong sản xuất nông nghiệp không?

1. Không

2. Có xả trực tiếp vào ao, hồ, đầm thả cá

3. Có xả trực tiếp vào kênh, rạch, sông, ngòi


5. Có xả trực tiếp vào nơi khác: ghi rõ____________________

52. Thu nhập đầu người hàng năm của quý vị là khoảng bao nhiêu?

1. < 4.8 triệu đồng

2. 4.8-10 triệu đồng

3. 10-20 triệu đồng

4. 20-40 triệu đồng

5. >40 triệu đồng


53. Quý vị bảo vệ bản thân khỏi nguy cơ mắc bệnh truyền nhiễm khi tiếp xúc với nước

SXNN như thế nào?

1. Mặc đồ bảo hộ

2. Đeo găng tay bảo vệ tay

3. Rửa chân tay sau khi làm việc ngoài đồng ruộng

4. Không bảo vệ


6. Dùng biện pháp khác

54. Quý vị có áp dụng những biện pháp/hành động dưới đây không?

1. Tiêm phòng cho vật nuôi

2. Đảm bảo điều kiện chăn nuôi sạch sẽ đạt tiêu chuẩn cho vật nuôi

3. Xử lý chất thải vật nuôi tại nguồn

4. Cách ly vật nuôi khỏi các vùng đánh bắt

5. Không sử dụng kháng sinh làm chất tăng trọng

6. Không áp dụng biện pháp nào nêu trên


of 340

55. Tầm quan trọng của những biện pháp sau trong việc tránh nguy cơ mắc các bệnh truyền

nhiễm là gì?QUẢN LÝ GIA SÚC GIA CẦM

Rất quan

trọng

Quan

trọng

Quan

trọng

vừa

Ít quan

trọng

Không

quan trọng

Không

có ý kiến

Tiêm phòng cho vật nuôi ❏ ❏ ❏ ❏ ❏ ❏

Đảm bảo điều kiện chăn nuôi

sạch sẽ đạt tiêu chuẩn cho vật

nuôi

❏ ❏ ❏ ❏ ❏ ❏

Xử lý chất thải vật nuôi tại

nguồn ❏ ❏ ❏ ❏ ❏ ❏

Cách ly vật nuôi khỏi các vùng

đánh bắt ❏ ❏ ❏ ❏ ❏ ❏

Không sử dụng kháng sinh làm

chất tăng trọng ❏ ❏ ❏ ❏ ❏ ❏


1. Xử lý nước thải

2. Xử lý chất thải động vật/con người tại nguồn

3. Sử dụng vùng đệm (ví dụ: bèo)

4. Bảo vệ, che chắn nguồn nước từ giếng đào/giếng khoan/suối

5. Che/chắn/đậy nguồn hoặc bể đựng nước uống

6. Sửa chữa các chỗ thất thoát trong hệ thống nước thải

7. Xử lý chất thải từ động vật và người trước khi sử dụng làm phân bón

8. Không áp dụng biện pháp nào nêu trên



nhiễm là gì?BẢO VỆ NGUỒN NƯỚC

Rất quan

trọng

Quan

trọng

Quan

trọng vừa

Ít quan

trọng

Không

quan trọng

Không có ý

kiến

Xử lý nước thải ❏ ❏ ❏ ❏ ❏ ❏

Xử lý chất thải động vật/con

người tại nguồn ❏ ❏ ❏ ❏ ❏ ❏

Sử dụng vùng đệm (ví dụ:

bèo) ❏ ❏ ❏ ❏ ❏ ❏

of 340

Bảo vệ, che chắn nguồn

nước từ giếng đào/giếng

khoan/ suối

❏ ❏ ❏ ❏ ❏ ❏

Che/chắn/đậy nguồn hoặc bể

đựng nước uống ❏ ❏ ❏ ❏ ❏ ❏

Sửa chữa các chỗ thất thoát

trong hệ thống nước thải ❏ ❏ ❏ ❏ ❏ ❏

Xử lý chất thải từ động vật

và người trước khi sử dụng

làm phân bón

❏ ❏ ❏ ❏ ❏ ❏


1. Che bể nươc hoặc chứa nước vào thùng, bể

2. Đầu tư vào hệ thống xử lý nước sinh hoạt gia đình

3. Thay/sử dụng chất tẩy rửa nước

4. Sửa chữa đồ chứa nước trong gia đình

5. Sửa chữa hệ thống cống thoát của gia đình


nhiễm là gì?LƯU TRỮ, XỬ LÝ, VÀ PHÂN PHỐI NƯỚC

Rất quan

trọng

Quan

trọng

Quan

trọng

vừa

Ít quan

trọng

Không

quan trọng

Không

có ý

kiến

Chứa nước vào thùng, bể ❏ ❏ ❏ ❏ ❏ ❏

Đầu tư vào hệ thống xử lý nước

sinh hoạt gia đình ❏ ❏ ❏ ❏ ❏ ❏

Thay/sử dụng chất tẩy rửa nước ❏ ❏ ❏ ❏ ❏ ❏

Sửa chữa đồ chứa nước trong

gia đình ❏ ❏ ❏ ❏ ❏ ❏

Sửa chữa hệ thống cống thoát

của gia đình ❏ ❏ ❏ ❏ ❏ ❏


1. Nâng cao vệ sinh cá nhân và trong gia đình (ví dụ rửa tay xà

phòng)

2. Vệ sinh gia trại hợp lý/xử lý chất thải gia dinh tốt

3. Đầu tư nâng cấp nguồn nước

4. Phòng các lây nhiễm chéo

5. Nâng cao vệ sinh thực phẩm (ví dụ rửa rau thịt)

of 340

6. Nấu chín/không ăn đồ ăn sống

7. Bảo vệ nước và nguồn cung cấp thức ăn cho động vật

8. Trao đổi thông tin về nguy cơ đối với các thành viên khác trong gia

đình, cộng đồng

9. Đeo thiết bị bảo hộ (như khẩu trang) hay dùng nước lọc

10. Dùng thuốc kháng sinh

61. Tầm quan trọng của những biện pháp này trong việc tránh nguy cơ mắc các bệnh truyền

nhiễm là gì?TẠI ĐIỂM SỬ DỤNG (HỘ GIA ĐÌNH)

Rất quan

trọng

Quan

trọng

Quan trọng

vừa

Ít quan

trọng

Không

quan trọng

Không có

ý kiến

Nâng cao vệ sinh cá nhân và

trong gia đình (ví dụ rửa tay

xà phòng)

❏ ❏ ❏ ❏ ❏ ❏

Xử lý chất thải gia đình tốt ❏ ❏ ❏ ❏ ❏ ❏

Đầu tư nâng cấp nguồn nước ❏ ❏ ❏ ❏ ❏ ❏

Phòng các lây nhiễm chéo ❏ ❏ ❏ ❏ ❏ ❏

Nâng cao vệ sinh thực phẩm

(ví dụ rửa rau thịt) ❏ ❏ ❏ ❏ ❏ ❏

Nấu chín/không ăn đồ ăn

sống ❏ ❏ ❏ ❏ ❏ ❏

Bảo vệ nước và nguồn cung

cấp thức ăn cho động vật ❏ ❏ ❏ ❏ ❏ ❏

Trao đổi thông tin về nguy

cơ đối với các thành viên

khác trong gia đình, cộng

đồng

❏ ❏ ❏ ❏ ❏ ❏

Đeo thiết bị bảo hộ (như

khẩu trang) hay dùng nước

lọc

❏ ❏ ❏ ❏ ❏ ❏

Dùng thuốc kháng sinh ❏ ❏ ❏ ❏ ❏ ❏

62. Quý vị thường xuyên xử lý chất thải chăn nuôi như thế nào?

1. Thu lại

2. Thải ra sông/hồ/cánh đồng

3. Đổ ở bãi thải gần nhà

4. Chôn

5. Đốt

6. Khác: Đề nghị cho biết cụ thể

of 340


63. Quý vị có sử dụng chất thải của người, nước thải/chất thải chăn nuôi cho việc thả cá

không?

1. Có

2. Không: Bỏ qua 2 câu kế tiếp


Water testing results questions:

pH levels:

64. What are pH levels in on-farm stored rainwater for drinking?

65. What are pH levels in on-farm pipe water for drinking?

66. What are pH levels in on-farm well water for drinking?

67. What are pH levels in on-farm bottled water for drinking?

68. What are pH levels in on-farm river water (fl.) for drinking?

69. What are pH levels in on-farm stored rainwater for domestic purposes?

70. What are pH levels in on-farm well water for domestic purposes?

71. What are pH levels in on-farm river water (fl.) for domestic purposes?

72. What are pH levels in on-farm pond water for domestic purposes?

Turbidity levels:

73. What are turbidity levels in on-farm stored rainwater for drinking?

74. What are turbidity levels in on-farm pipe water for drinking?

75. What are turbidity levels in on-farm well water for drinking?

76. What are turbidity levels in on-farm bottled water for drinking?

77. What are turbidity levels in on-farm river water (fl.) for drinking?

78. What are turbidity levels in on-farm stored rainwater for domestic purposes?

79. What are turbidity levels in on-farm well water for domestic purposes?

80. What are turbidity levels in on-farm river water (fl.) for domestic purposes?

81. What are turbidity levels in on-farm pond water for domestic purposes?

E. coli cfu levels:

82. What are E. coli cfu levels in on-farm stored rainwater for drinking?

83. What are E. coli cfu levels in on-farm pipe water for drinking?

84. What are E. coli cfu levels in on-farm well water for drinking?

85. What are E. coli cfu levels in on-farm bottled water for drinking?

86. What are E. coli cfu levels in on-farm river water (fl.) for drinking?

87. What are E. coli cfu levels in on-farm stored rainwater for domestic purposes?

88. What are E. coli cfu levels in on-farm well water for domestic purposes?

89. What are E. coli cfu levels in on-farm river water (fl.) for domestic purposes?

90. What are E. coli cfu levels in on-farm pond water for domestic purposes?

of 340

B.2. Questionnaires in English:

1. Code of the participating farmer

2. What is your gender?

1. Male

2. Female

3. Other

3. When were you born? ___________(dd/mm/yy)

4. What is your ethnicity?

1. Kinh

2. Other: specify

5. What was your highest school attended?

6. How many people are there in your house?

7. How many kids (at the age under 18) are there in your house?

of 340

8. How many years have you been an integrated farmer?

9. What is your occupation other than farming?

10. Overall, who is primarily responsible for livestock production and health of the family?

Issue Main responsibility

Man Woman Both Not sure Do not want to

answer

Livestock production

Fish production

Family health

11. What is your relationship with other household members?

1. Head of the household

2. Other roles

12. How do you see the general environmental water condition of your village?

1. Excellent

2. Good

3. Adequate

4. Poor

5. Very poor

6. Do not want to answer

of 340

13. Do you think that your source of untreated water (not boiled or filtered) for drinking is safe?

1. Yes

2. No

14. Do you treat your drinking or household water?

Filter Boil Both boil and

filter

Other

(state)

No

treatment

Do not

want to

answer

Drinking

water ❏ ❏ ❏ ❏ ❏ ❏

Household

water ❏ ❏ ❏ ❏ ❏ ❏

15. If yes (both, or filter, or boil) then do you treat because of a past event?

1. No

2. Yes (please state what event) _________________________________

16. If yes (both, filter, boil) then do you think your filtered and/or boiled drinking water is safe to

drink?

Yes, but must do both Filter only Boil only No

17. Please check the level of harmfulness to your health of the following sources of water (not

boiled or filtered).

Extreme High Moderate Low No effect No

opinion

Consuming drinking water

(not treated)

Handling household water

(not treated)

Handling pond water (not

treated)

Bottled water

18. Now I want to ask you about bird flu/avian flu. Have you heard of bird flu/avian flu?

1. Yes, I have heard of bird flu/avian flu

2. No, I have not heard of bird flu/avian flu

of 340

19. Please check the level of susceptibility of bird flu from the following sources of water (not


Extreme High Moderate Low No

effect

No

opinion

Drinking water (not treated)

Handling household water (not

treated)


treated)

Handling livestock or poultry

waste water (not treated)

20. If you were to get bird flu, how severe a disease do you think it would be?

1. Very severe

2. Severe

3. Moderate

4. Mild

5. Trivial

6. No opinion

21. Was your poultry destroyed because of an outbreak of bird flu (HPAI) in your province?

1. No, our poultry was never destroyed

2. Yes, our poultry was destroyed

3. Do not remember

4. No opinion

22. What do you do when you observe unusual number of sick/death your poultry/livestock?

1. Ignore

2. Consult other farmers

3. Notify the village leader

4. Burn/Bury

5. Notify the local para-veterinarians

6. Use as food (for livestock or human)

7. Nothing

8. Other, please specify

23. Currently, is your poultry being vaccinated because of the bird flu (HPAI) outbreak?

1. Yes vaccinated

2. No, not vaccinated yet

3. No opinion

of 340

24. Have you ever heard about the microbes that may exist in water such as viruses, bacteria, and

helminths?

1. Yes, I have

2. No, I have not

25. Please check the level of susceptibility to diarrhea from the following sources of water (not


Extrem

e

Hig

h

Moderat

e

Low No

effect

No

opinion

Drinking water (not

treated)


(not treated)


treated)

Handling livestock or

poultry waste water (not

treated)

26. If you were to get diarrhea, how severe a disease do you think it would be?

1. Very severe

2. Severe

3. Moderate

4. Mild

5. Trivial

6. No opinion

27. One kind of microbes that can be found in farm water and sometime dangerous is E. coli.

Have you heard of E. coli or coliform disease?

1. Yes, move on to the next question and continue

2. No, skips the next question and continue

28. Please check the level of susceptibility to coliform disease from the following sources of

water (not boiled or filtered).

Extrem

e

Hig

h

Modera

te

Lo

w

No

effect

No opinion

Drinking water (not treated)


(not treated)


treated)

of 340



treated)

29. If you were to get coliform disease, how severe a disease do you think it would be?

1. Very severe

2. Severe

3. Moderate

4. Mild

5. Trivial

6. No opinion

30. Another source of disease that can be found in water on a farm and sometimes can be

harmful is parasites such as ascarids or helminths. Have you heard of the parasites called

ascarids or helminths?

1. Yes, move on to the next question and continue

2. No, skips 2 next questions and continue

31. Please check the level of susceptibility of disease from parasites from the following sources

of water (not boiled or filtered).

Extreme High Moderate Low No

effect

No opinion

Drinking water (not

treated)


(not treated)


treated)



treated)

32. If you were to get disease from parasites, how severe a disease do you think it would be?

1. Very severe

2. Severe

3. Moderate

4. Mild

5. Trivial

6. No idea

of 340

33. How important/serious are the issues of water usage to the reduction of risk factors for

WRZDs (e.g., AI)?

Extremel

y

important

High

important

Moderate

important

Low

important

Not

importa

nt

No

opinion

Source of water

Not share between

human and livestock

Closeness of water

source and livestock

Shortage of water

Rainfall

Water quality

Other

issues________

34. Here is the scale representing the important of proper water usage in reducing risk factors for

WRZDs as well as AI. On which step of the scale do you feel you personally stand now?

1. Extremely important

2. High important

3. Moderate important

4. Low important

5. Not important

6. No opinion

35. Overall, are you satisfied or dissatisfied with the current source of water and wastewater that

you are using for yourself, your livestock, and your crop?

Totally

satisfied

Highly

satisfied

Moderate

satisfied

Low

satisfied

Dissatis

fied

No

opinion

Yourself

(drinking)

livestock, crop

36. Please indicate your thought about the level of risk of WRZDs of the following sources of

water

Extremely

risk

High

risk

Moderate

risk

Low

risk

No risk No

opinion

From the rain

Lake

River

of 340

Canal

Dug well

Drilled well

Other (please specify)

37. What factors drive your thoughts of the quality of water and wastewater?

1. Based on impression produced by any substance on the water and

wastewater, taste, or smell, and also on the organism as a whole

(organoleptic)

2. Previous experiences

3. Risk perceptions

4. Attitudes towards water chemical contamination

5. Contextual cues relating to the source of water supply

6. Familiarities with specific water properties

7. Trust in suppliers/source

8. Media

9. No opinion

10. Other factors _please specify

38. What are the sources of information relating to WRZDs as well as AI that you get and how

reliable are these sources?

Extr

eme

High Moder

ate

Low No No

opinion

TV

Newspapers

Radio

Family/friends/other farmers

Government (CHCs)

Vet/Doctor

Other source_ please

specify__________________

_______

39. Do you think applying preventive actions will protect against the WRZDs?

1. No

2. Yes, such as hand washing

3. Yes, Isolation

4. Yes, Wearing protective (e.g., mouth mask, filter water use)

5. Yes, Vaccination

6. Yes, Taking antiviral

of 340

7. Yes, Cook food/not eating raw food

8. No opinion

9. Other (please name it) ______________

40. Are there benefits (other than health) that ENCOURAGE you to take actions will protect

against the WRZDs?

Please indicate_____________________________________________________________

41. Are there benefits that encourage you NOT to take actions will protect against the WRZDs?

Please indicate_____________________________________________________________

42. What are the barriers that prevent you from undertaking action that can protect against the

WRZDs?

1. Cannot afford the cost of actions

2. Do not know how to take actions

3. Do not understand the need to take action

4. Influence by neighbours who do not take actions

5. Do not have time for taking action, too busy with other things

6. No opinion

7. Other reason (please explain it)

______________________________________________

43. What are the triggers/cues to that influence your decision to undertake action that can protect

against the WRZDs?

1. Neighbours do the same

2. Advice from health officer, veterinarians

3. Radio/newspaper

4. Not busy with other tasks

5. Worry about being infected or affected by EID

6. Lost income

7. No opinion

8. Other reason (please explain it)

______________________________________________

44. What are the agricultural production types that you have?

1. Rice

2. Cassava

3. Bamboo

of 340

4. Corn

5. Peanuts

6. Vegetables (state)

7. fruit (state)

8. Fish production

9. Livestock production

45. What are your top three sources of farm management information (put number 1, 2, 3 of 9 ):

Fellow farmers

Family members

Newspaper or other printed media

Radio or TV

Commune expert

Training program

Feed supplier

NGO

Other (state)

46. What types of livestock do you normally have in your farm and its respective number?

Estimated number per

season (YOUNG)

Estimated number per

season (ADULT)

Estimated production

volume (kg) per season

Note

Fish

Chicken

Duck

Pig

Goat

Other poultry

Other cattle

Other pet

47. What type of water sources does your family have access to and uses the most for drinking?

Most of the

time

Frequently Occasionally Rarely Never No

opini

on

Pond/ lake

River/ stream/canal

Rain water collection

of 340

Government

supply/piped

Hand dug wells

Drilled wells

Buying water

Other sources

_________

48. What type of water sources does your family have access to and uses the most for your

poultry and other livestock?

Most of

the time

Frequentl

y

Occasiona

lly

Rarely Never No

opini

on

Pond/ lake

River/ stream/canal

Rain water collection

Government

supply/piped

Hand dug wells

Drilled wells

Buying water

Other sources

_________

49. Was your poultry infected with bird flu (HPAI) during the HPAI outbreak of 2004?

1. Yes

2. No

3. Do not remember

4. No opinion

50. Was your poultry infected with bird flu (HPAI) during the last 12 months?

1. Yes

2. No, skip next 2 questions

3. Do not remember

4. No opinion

51. Do you dump waste from livestock production directly to the source of water use for

agriculture production?

1. No

2. Yes, directly to pond/lake

of 340

3. Yes, directly to canal/river/creek

4. No opinion

5. Yes, directly to other places. Please

specify__________________________________

52. What is your total on-farm income per person per year?

1. < 4.8 million VND

2. 4.8-10 million VND

3. 10-20 million VND

4. 20-40 million VND

5. >40 million VND

6. No opinion

53. How do you protect yourself from the risk of WRZDSs?

1. Wear protection cloth

2. Wear glove to protect hands

3. Wash hands and feet after field work

4. No protection

5. No opinion

6. Other, please explain

54. Please indicate if you have been employing one or more of the following livestock

management methods

1. Vaccination of domestic animals

2. Ensure hygienic rearing conditions for animals

3. Treatment of animal waste at source

4. Exclusion of animals from catchment basin

5. Stop using antibiotics as growth promoters

6. None of the above

7. No opinion

55. How important are these methods to avoid being transmitted with WRZDSs from water and

wastewater source? LIVESTOCK MANAGEMENT

Very

importan

t

Highly

Important

Moderat

e

importan

t

Low

importan

t

Not

important

No

opinio

n

Vaccination of

domestic animals

of 340

Ensure hygienic

rearing conditions

for animals

Treatment of animal

waste at source

Exclusion of

animals from

catchment basin

Stop using

antibiotics as

growth promoters

56. Please indicate if you have been employing one or more of the following source water

protection methods

1. Wastewater treatment

2. Human/animal waste treatment at source

3. Using buffer zones (e.g., vegetable)

4. Protect of well/springs/ground water

5. Protect Cover drinking water reservoir

6. Repair leaking in sewage system

7. Treatment of animal/human wastes prior to use as fertilizers

8. None of the above

9. No opinion

57. How important the methods below are to avoid being transmitted with WRZDSs from water

and wastewater source? SOURCE WATER PROTECTION

Very

important

Highly

Important

Moderat

e

importan

t

Low

importan

t

Not

important

No

opinion

Wastewater treatment

Human/animal waste

treatment at source

Using buffer zones

(e.g., vegetable)

Protect of

well/springs/ground

water

Cover drinking water

reservoir

Repair leaking in

sewage system

of 340

Treatment of

animal/human wastes

prior to use as

fertilizers

58. Please indicate if you have been employing one or more of the following mitigation practices

in relating to water storage treatment and distribution

1. Cover water storage reservoir or store water in confined tanks

2. Spend money in household water treatment technology

3. Change/use disinfectants

4. Repair household water storage

5. Repair household sewer

59. How important the methods below are to avoid being transmitted with WRZDSs from water

and wastewater source? WATER STORAGE TREATMENT ANF DISTRIBUTION

Very

importa

nt

Highly

Importa

nt

Moderat

e

importa

nt

Low

importa

nt

Not

importa

nt

No

opinion

Cover water storage

reservoir or store water

in confined tanks

Spend money in

household water

treatment technology

Change/use

disinfectants

Repair household water

storage

Repair household sewer

60. Please indicate if you have been employing one or more of the following mitigation

strategies/practices/behaviours

1.Improve personal, domestic and food hygiene (e.g. hand washing)

2.Adequate household sanitation

3.Invest in improving water supply

4.Prevent cross-contamination

5.Improve food hygiene

6.Cook foods/ now eating raw food

7.Protect water source and food source for animals

of 340

8.Communicate risk to other members of household, community

9.Wear protective (e.g., mouth mask), filter water use

10. Taking antiviral

61. How important it is to avoid being transmitted with WRZDSs from water and wastewater

source? POINT OF USE/HOUSEHOLD

Very

importa

nt

Highly

Importa

nt

Moderat

e

importan

t

Low

importa

nt

Not

important

No

opinio

n

Improve personal,

domestic and food

hygiene (e.g. hand

washing)

Adequate household

sanitation

Invest in improving

water supply

Prevent cross-

contamination

Improve food hygiene

Cook foods/ now eating

raw food

Protect water source

and food source for

animals

Communicate risk to

other members of

household, community

Wear protective (e.g.,

mouth mask), filter

water use

Taking antiviral

62. How does your household usually handle livestock waste?

1.Collect

2.Dump in river/lake/rice field

3.Dump in a site near by

4.Bury

5.Burn

6.Other Please specify

7.No opinion

of 340

63. Do you use waste from livestock production for fish production?

1. Yes

2. No, skip the next 2 questions

3. No opinion

Water testing results questions:

pH levels:

64. What are pH levels in on-farm stored rainwater for drinking?

65. What are pH levels in on-farm pipe water for drinking?

66. What are pH levels in on-farm well water for drinking?

67. What are pH levels in on-farm bottled water for drinking?

68. What are pH levels in on-farm river water (fl.) for drinking?

69. What are pH levels in on-farm stored rainwater for domestic purposes?

70. What are pH levels in on-farm well water for domestic purposes?

71. What are pH levels in on-farm river water (fl.) for domestic purposes?

72. What are pH levels in on-farm pond water for domestic purposes?

Turbidity levels:

73. What are turbidity levels in on-farm stored rainwater for drinking?

74. What are turbidity levels in on-farm pipe water for drinking?

75. What are turbidity levels in on-farm well water for drinking?

76. What are turbidity levels in on-farm bottled water for drinking?

77. What are turbidity levels in on-farm river water (fl.) for drinking?

78. What are turbidity levels in on-farm stored rainwater for domestic purposes?

79. What are turbidity levels in on-farm well water for domestic purposes?

80. What are turbidity levels in on-farm river water (fl.) for domestic purposes?

81. What are turbidity levels in on-farm pond water for domestic purposes?

E. coli cfu levels:

82. What are E. coli cfu levels in on-farm stored rainwater for drinking?

83. What are E. coli cfu levels in on-farm pipe water for drinking?

84. What are E. coli cfu levels in on-farm well water for drinking?

85. What are E. coli cfu levels in on-farm bottled water for drinking?

86. What are E. coli cfu levels in on-farm river water (fl.) for drinking?

87. What are E. coli cfu levels in on-farm stored rainwater for domestic purposes?

88. What are E. coli cfu levels in on-farm well water for domestic purposes?

89. What are E. coli cfu levels in on-farm river water (fl.) for domestic purposes?

90. What are E. coli cfu levels in on-farm pond water for domestic purposes?

of 340

APPENDIX C: CONSENT SCRIPTS

C.1. Consent Scripts in Vietnamese

PHIÊU TINH NGUYÊN THAM GIA ĐIÊU TRA

A. Phiếu tình nguyện tham gia điều tra _Dành cho Bảng hỏi và lấy mẫu nước xét nghiệm:

Phiếu này phải được phỏng vấn viên đọc cho người tham gia

Xin chào và xin cảm ơn Quý vị đã đến dự cuộc thảo luận nhóm hôm nay! Tên tôi là….. tôi là

thành viên của nhóm điều tra thí điểm. Thay mặt nhóm điều tra, tôi xin kính mời các cô/bác

anh/chi tham gia buổi thảo luận nhóm hôm nay. Đây là một phần của điều tra thí điểm mang tên

“Động vật, nước, và y tế công cộng tại Việt Nam”. Việc tham gia điều tra thí điểm này là hoàn

toàn mang tính tình nguyện, Quý vị có quyền quyết định có tham gia buổi thảo luận nhóm này

hay không. Điều tra thí điểm nhằm kiểm tra các công cụ được thiết kế cho điều tra nhằm xác

định những chỉ số cơ bản của chất lượng nước sử dụng tại nông hộ (ví dụ như độ đục và vi khuẩn

chỉ định); tìm hiểu loại hình chăn nuôi, thả cá trong nông hộ quy mô nhỏ; ghi nhận những suy

nghĩ và hành động của nông dân liên quan đến việc sử dụng nước, nước thải cũng như nguy cơ

lây truyền bệnh qua nước; xác định mối liên hệ giữa những chỉ số nêu trên với suy nghĩ và

phương cách của nông dân.

Bảng hỏi và lấy mẫu nước xét nghiệm là một trong các hoạt động của điều tra thí điểm tại xã.

Trong bảng hỏi này, tôi sẽ hỏi các nhóm câu hỏi liên quan đến cá nhân hộ gia đình và nông trại

của Quý vị như 1) thông tin về cá nhân và hộ gia đình; 2) quan niệm và thái độ liên quan đến các

yếu tố ảnh hưởng đến sức khỏe của việc sử dụng nước và nước thải; 3) các loại hình chăn nuôi,

thả cá và trông trọt tại nông trại; 4) thu nhập có được từ chăn nuôi, thả cá và trông trọt tại nông

trại; và 5) các hành động nhằm loại bỏ, giảm thiểu tần suất, mức độ tiếp xúc với nguy cơ mắc các

bênh truyên nhiễm cũng như cúm gia cầm liên quan đến việc sử dụng nước và nước thải. Tôi sẽ

ghi lai câu trả lời của Quý vị. Dự kiến mất khoảng 45 phút để hoàn thành bảng hỏi này. Liên

quan đến phần lấy mẫu xét nghiệm, đại diện của TT Y tế dự phòng Thái Bình sẽ xin phép lấy

mẫu nước uống và nước sinh hoạt của nông trại để xét nghiệm về các chỉ sô cơ bản của chất

lượng nước. Tôi đề nghị Quý vị chỉ dẫn chúng tôi đến nguồn nước uống và nguồn nước sinh hoạt

được dùng nhiều nhất tại nông hộ cho người, vật nuôi, và trồng trọt. Tất cả các câu trả lời và

thông tin thu thập đươc sẽ được giữ bí mật riêng tư. Tuy nhiên những hoạt động vi phạm pháp

luật nếu có sẽ phải báo cáo với chính quyền địa phương. Nếu các Quý vị có bất cứ câu hỏi nào

liên quan đến điều tra thí điểm này tôi sẽ vui lòng trả lời trong khả nằng có thể.

Dự kiến, điều tra này sẽ giúp chỉnh sửa công cụ điều tra cho phù hợp trước khi đươc sử dụng

trong một điều tra quy mô lớn hơn nhằm giúp hiểu biết tốt hơn về mối liên hệ giữa quan niệm về

chất lượng nước sử dụng tại nông trại và việc giảm thiểu nguy cơ mắc bệnh truyền nhiễm trong

nông hộ chăn nuôi quy mô nhỏ ỏ Thái Bình, từ đó cung cấp thông tin và khuyến nghị cho việc

thiết kể và thực hiện chương trình, chính sách phù hợp mới mối quan tâm, nhu cầu cũng như

điều kiện kinh tế của hộ nông dân do đó giúp cải thiện sức khỏe nông dân, vật nuôi và môi

trường một cách bền vững. Các thông tin cá nhân thu thập được sẽ được đảm bảo giữ bí mật

riêng tư. Mã số sẽ được dùng cho thảo luận. Chỉ có tôi mới biêt đó là mã số gì. Quý vị có thể

of 340

quyết định không tham trả lời bất kỳ câu hỏi nào hoặc ngừng tham gia điều tra tại bất cứ thời

điểm nào nếu muốn. Nếu ai có câu hỏi gì liên quan đến điều tra thí điểm này thì có thê liên hệ

BS. Lê Bá Quỳnh theo sổ điện thoại 0165 247 1058.

Xin trân trọng cảm ơn!

of 340

B. Phiếu tình nguyện tham gia điều tra _Dành cho Phỏng vấn sâu, Quan sát:


Xin chào và xin cảm ơn các Quý vị đã đến dự cuộc thảo luận nhóm hôm nay! Tên tôi là….. tôi là

thành viên của nhóm điều tra thí. Thay mặt nhóm điều tra, tôi xin kính mời các cô/bác anh/chi

tham gia buổi thảo luận nhóm hôm nay. Đây là một phần của điều tra thí điểm mang tên “Động

vật, nước, và y tế công cộng tại Việt Nam”. Việc tham gia điều tra thí điểm này là hoàn toàn

mang tính tình nguyện, cô/bác anh/chi có quyền quyết định có tham gia buổi thảo luận nhóm này

hay không. Điều tra thí điểm nhằm kiểm tra các công cụ được thiết kê cho điều tra nhằm xác

định những chỉ số cơ bản của chất lượng nước sử dụng tại nông hộ (ví dụ như độ đục và vi khuẩn

chỉ định); tìm hiểu loại hình chăn nuôi, thả cá trong nông hộ quy mô nhỏ; ghi nhận những suy

nghĩ và hoạt động của nông dân liên quan đến việc sử dụng nước, nước thải cũng như nguy cơ

lây truyền bệnh qua nước; xác định mối liên hệ giữa những chỉ số nêu trên với suy nghĩ và

phương cách của nông dân.

Phỏng vấn sâu, quan sát nông trại là một trong các hoạt động của điều tra thí điểm tại xã. Trong

phần phỏng vấn sâu tôi sẽ hỏivà thảo luận với Quý vị vấn đề đến 1) quan niệm và thái độ liên

quan đến các yếu tố ảnh hưởng đến sức khỏe của việc sử dụng nước và nước thải; 2) các loại

hình chăn nuôi, thả cá và trông trọt tại nông trại; 4) thu nhập có được từ chăn nuôi, thả cá và

trồng trọt tại nông trại; và 5) các hành động nhằm loại bỏ, giảm thiểu tần suất, mức độ tiếp xúc

với nguy cơ mắc các bênh truyên nhiễm cũng như cúm gia cầm liên quan đến việc sử dụng nước

và nước thải. Tôi sẽ ghi lai câu trả lời và phần chia sẻ ý kiến của Quý vị. Dự kiến mất khoảng 30

phút để hoàn thành phần phỏng vấn sâu. Sau khi hoặc trong khi phỏng vấn sâu tôi đề nghị được

quan sát nông trại của Quý vị, ghi chép lại và chụp ảnh nếu cần. Tất cả các câu trả lời của các

Quý vị và thông tin thu thập sẽ được giữ bí mật riêng tư. Tuy nhiên những hoạt động vi phạm

pháp luật nếu có sẽ phải báo cáo với chính quyền địa phương. Nếu các Quý vị có bất cứ câu hỏi

nào liên quan đến điều tra thí điểm này tôi sẽ vui lòng trả lời trong khả nằng có thể. Tổng thời

gian dành cho Phỏng vấn sâu, quan sát nông trại dự kiến khoảng 1 giờ đồng hồ.

Dự kiến, kết quả của điều tra này sẽ giúp chỉnh sửa công cụ điều tra cho phù hợp trước khi đươc

sử dụng trong một điều tra quy mô lớn hơn nhằm giúp hiểu biết tốt hơn về mối liên hệ giữa quan

niệm về chất lượng nước sử dụng tại nông trại và việc giảm thiểu nguy cơ mắc bệnh truyền

nhiễm trong nông hộ chăn nuôi quy mô nhỏ ỏ Thái Bình, từ đó cung cấp thông tin và khuyến

nghị cho việc thiết kể và thực hiện chương trình, chính sách phù hợp mới mối quan tâm, nhu cầu

cũng như điều kiện kinh tế của hộ nông dân do đó giúp cải thiện sức khỏe nông dân, vật nuôi và

môi trường một cách bền vững

Các thông tin cá nhân thu thập được sẽ được đảm bảo giữ bí mật riêng tư. Mã số sẽ được dùng

cho thảo luận. Chỉ có tôi mới biêt đó là mã số gì. Các Quý vị có thể quyết định không tham trả

lời bất kỳ câu hỏi nào hoặc ngừng tham gia điều tra tại bất cứ thời điểm nào nếu muốn. Nếu ai có

câu hỏi gì liên quan đến điều tra thí điểm này thì có thê liên hệ BS. Lê Bá Quỳnh theo sổ điện

thoại 0165 247 1058.


of 340

C. Phiếu tình nguyện tham gia điều tra _Dành cho Thảo luận nhóm:


Xin chào và xin cảm ơn Quý vị đã đến dự cuộc thảo luận nhóm hôm nay! Tên tôi là….. tôi là

thành viên của nhóm điều tra thí điểm. Thay mặt nhóm điều tra, tôi xin kính mời các Quý vị

tham gia buổi thảo luận nhóm hôm nay. Đây là một phần của điều tra thí điểm mang tên “Động

vật, nước, và y tế công cộng tại Việt Nam”. Việc tham gia điều tra thí điểm này là hoàn toàn

mang tính tình nguyện, Quý vị có quyền quyết định có tham gia buổi thảo luận nhóm này hay

không. Điều tra thí điểm nhằm kiểm tra các công cụ được thiết kê cho điều tra nhằm xác định

những chỉ số cơ bản của chất lượng nước sử dụng tại nông hộ (ví dụ như độ đục và vi khuẩn chỉ

định); tìm hiểu loại hình chăn nuôi, thả cá trong nông hộ quy mô nhỏ; ghi nhận những suy nghĩ

và hoạt động của nông dân liên quan đến việc sử dụng nước, nước thải cũng như nguy cơ lây

truyền bệnh qua nước; xác định mối liên hệ giữa những chỉ số nêu trên với suy nghĩ và phương

cách của nông dân.

Thảo luận nhóm là một trong các hoạt động của điều tra thí điểm tại xã. Trong phần thảo luận

nhóm này, tôi sẽ đề nghị Quý vị thảo luận cùng nhau về những chủ đề chính dưới đây:

1. Những vấn đề gì liên quan đến nước và chất thải ảnh hưởng nhiều nhất đến sinh kế, chăn

nuôi, thả cá của các Quý vị, và tại sao đó là vấn đê đáng quan tâm?

2. Những vấn đề đó xảy ra thường xuyên như thế nào trong quá khứ?

3. Nguyên nhân của những vấn đề đó là gì?

4. Hậu quả của các vấn đề đó như thế nào đối với bản thân, gia đình, và chăn nuôi thả cá

của các Quý vị?

5. Quý vị đã đối phó/ giải quyết những vấn đề đó như thế nào?

6. Ai có thể giúp Quý vị giải quyết vấn đề này?

7. Các Quý vị cần những sự giúp đỡ gì, như thế nào?

8. Nếu các vấn đề nêu trên xảy ra trong tương lai, Các Quý vị sẽ giải quyết thế nào?

9. Có những chính sách, chương trình, quy định gì của các tổ chức khác giúp giải quyết vấn

đề nêu trên? Nếu có, đó là gì và theo Quý vị có hiệu quả không?

Tôi đề nghị các Quý vị thảo luận với nhau thay vì với tôi. Tôi sẽ chỉ điều khiển cuộc thảo luận,

ghi chép. Dự kiến thảo buổi thảo luận nhóm không kéo dài quá 2 giờ đồng hồ và sẽ được ghi âm.

Đề ghị các Quý vị lưu ý là không có ý kiến nào là sai hay đúng, và khuyến khích các Quý vị chia

sẻ ý kiến khác nhau nhưng cần tôn trọng ý kiến người khác mặc dù có đồng ý hay không. Tất cả

những thông tin trong cuộc thảo luận sẽ được giữ bí mật. Trừ trường hơp những hoạt động vi

phạm pháp luật nếu có sẽ phải báo cáo với chính quyền địa phương. Nếu các Quý vị có bất cứ

câu hỏi nào liên quan đến điều tra thí điểm này tôi sẽ vui lòng trả lời trong khả nằng có thể. Dự

kiến, điều tra này sẽ giúp chỉnh sửa công cụ điều tra cho phù hợp trước khi đươc sử dụng trong

một điều tra quy mô lớn hơn nhằm giúp hiểu biết tốt hơn về mối liên hệ giữa quan niệm về chất

lượng nước sử dụng tại nông trại và việc giảm thiểu nguy cơ mắc bệnh truyền nhiễm trong nông

hộ chăn nuôi quy mô nhỏ ỏ Thái Bình, từ đó cung cấp thông tin và khuyến nghị cho việc thiết kể

và thực hiện chương trình, chính sách phù hợp mới mối quan tâm, nhu cầu cũng như điều kiện

kinh tế của hộ nông dân do đó giúp cải thiện sức khỏe nông dân, vật nuôi và môi trường một

cách bền vững

of 340

Các thông tin cá nhân thu thập được sẽ được đảm bảo giữ bí mật riêng tư. Mã số sẽ được dùng

cho thảo luận. Chỉ có tôi mới biêt đó là mã số gì. Quý vị có thể quyết định không tham trả lời bất

kỳ câu hỏi nào hoặc ngừng tham gia điều tra tại bất cứ thời điểm nào nếu muốn. Nếu ai có câu

hỏi gì liên quan đến điều tra thí điểm này thì có thê liên hệ BS. Lê Bá Quỳnh theo sổ điện thoại

0165 247 1058.


of 340

C.2. Consent Scripts in English

A. Consent script for Questionnaire and water sample collection

This consent script is required to be read to participants by enumerators/field investigators.

Greetings everybody and thank you for agreeing to meet with me! My name is … I am a member

of the research team. On behalf of the research team, I am inviting you to participate in the research

titled Ecosystem approaches to improved water and farm health in Vietnam with assistant from

the commune and district collaborators in Thai Binh and An Giang provinces. Your participation

in this research is highly appreciated and voluntary, so it is your decision to choose to participate

or not. The research is interested in determining the basic level of on-farm water quality (i.e.,

turbidity and total coliforms/fecal coliform indicators); examining the types of fish and livestock

raised on small scale farms in Thai Binh and An Giang; recording farmers’ perceptions and actions

with respect to water and wastewater usage and risk of disease to animals and humans from water;

and determining the associations between these indicators and the presence of fish and livestock,

and the recorded farmers’ perceptions and actions. As part of the research, I will ask you some

questions about 1) the general description of yourself and your household; 2) your attitudes,

perception, and your actions relating to risk factors of WRZD transmission; 3) your types of

livestock, fish, and crop; 4) your income from different livestock, fish, and crop; and 5) your

actions to eliminate or reduce the frequency, magnitude, or severity of exposure to risks, or

minimization of the potential impact of a risk of emerging infectious diseases as well as avian

influenza transmission relating to water and wastewater usage. I will write down your answers to

these questions in the questionnaire. It will take approximately 45 minutes of your time to answer

these questions. All information in your answers to the questionnaire will be kept confidential.

However, please note that I will have to report to local authority if any illegal or abuse observed.

Please feel free to ask questions anytime if you have any. I will be happy to explain the research

in detail if you wish.

It is envisaged that this research will bring about better understanding of associations between the

perceptions of water and wastewater usage and the reduction of risk factor of infectious diseases

amongst small scale farmers in Thai Binh and An Giang province and from that a larger scale

study is expected to follow to obtain further information and provide appropriate recommendations

for better design and implementation of programs/policies that are more suitable to the concerns

of small scale livestock production farmers, reflect the need, perceptions of farmers and their

socioeconomic status, their environment, and therefore can help improving health of farmers,

livestock, and environment sustainably (in the long run).

There is a procedure in place to make sure that all personal information will be kept private and

confidential. A code will be assigned to each questionnaire and only I will be able to identify which

code is for whom. You may choose not to answer any of the questions and/or decide to discontinue

your participation any time if you are no longer wish to participate in the research without any

penalty. If you have any questions, concerns about the research, please contact Dr. Quynh Ba Le

@ 0165 247 1058. Thank you for your participation!

of 340

B. Consent script for in-depth interview and observation

This consent script is required to be read to participants by enumerators/field investigators.

Greetings everybody and thank you agreeing to meet with me! My name is …I am a member of

the research team. On behalf of the research team, I am inviting you to participate in the research

titled Ecosystem approaches to improved water and farm health in Vietnam. Your participation

in this research is highly appreciated and voluntary, so it is your decision to choose to participate

or not.

The research is interested in determining the basic level of on-farm water quality (i.e., turbidity

and total coliforms/fecal coliform indicators); examining the types of fish and livestock raised on

small scale farms in Vietnam; recording farmers’ perceptions and actions with respect to water

and wastewater usage and risk of disease to animals and humans from water; and determining the

associations of these indicators with the presence of fish and livestock, and the recorded farmers’

perceptions and actions. As part of the research, I will ask you some questions, observe your farm

and undertake in-depth interview, and take water sample from your farm for testing. The questions

will be about 1) the general description of yourself and your household); 2) your attitudes,

perception, and your actions relating to risk factors of WRZD transmission; 3) your type of

livestock, fish, and crop; 4) your income from different livestock, fish, and crop; and 5) your

actions to eliminate or reduce the frequency, magnitude, or severity of exposure to risks, or

minimization of the potential impact of a risk of emerging infectious diseases as well as avian

influenza transmission relating to water and wastewater usage. It will take about two and 30

minutes for the in-depth interview. The in-depth interview will be recorded. The total time for an

in-depth interview and observation will be about 1 hour.

It is envisaged that this research will bring about better understanding of associations between the

perception of water and wastewater usage and the reduction of risk factors for infectious diseases

amongst small scale farmers in Thai Binh and An Giang provinces and from that a larger scale

study is expected to follow to obtain further information and provide appropriate recommendations

for better design and implementation of programs/policies that are more suitable to the concerns

of small scale farmers, reflect the need, perceptions of farmers and their socioeconomic status,

their environment, and therefore can help improving health of farmers, livestock, and environment

sustainably (in the long run). There is a procedure in place to make sure that all personal

information will be kept private and confidential. A code will be assigned to each interview and

water sample and only I will be able to identify which code is for whom. You may choose not to

answer any of the questions and/or decide to discontinue your participation any time if you no

longer wish to participate in the research without any penalty. If you have any questions, concerns

about the research, please contact Dr. Quynh Ba Le @ 0165 247 1058.

Thank you for your participation!

of 340

C. Consent script for focus group discussion

This consent will be used. This oral consent script is required to be read to participants by

enumerators/field investigators.

Greetings everybody and thank you for being here today for the group discussion! My name is …I

am a member of the research team. On behalf of the research team, I am inviting you to participate

in the research titled Ecosystem approaches to improved water and farm health in Vietnam. Your

participation in this research is highly appreciated and voluntary, so it is your decision to choose

to participate or not. The research is interested in determining the basic level of on-farm water

quality (i.e., turbidity and total coliforms/fecal coliform indicators); examining the types of fish

and livestock raised on small scale farms in Vietnam; recording farmers’ perceptions and actions

with respect to water and wastewater usage and risk of disease to animals and humans from water;

and determining the associations of these indicators with the presence of fish and livestock, and

the recorded farmers’ perceptions and actions. As part of the research, I will ask you to discuss the

following questions as a group:

1. Which water and wastewater related problems have been the most urgent for your

livelihood, livestock/fish production and why?

2. How often is the problem in comparison with the past?

3. What are the causes of the problem?

4. What are the consequence of the problem to you, your family, and your livestock?

5. How did you deal with the problem?

6. Who do you think should help you in this situation?

7. What kind of help did/do you need?

8. How will you deal with the problem if it happens again?

9. Are there any policy/regulation/ program from other organization to address this

problem?

I want you to talk to each other rather than to me. I will be moderating and taking notes of the

discussion, which will take about two hours and will be recorded. There is no right or wrong

comments, please respect but do not hesitate to disagree with what others have said or to share

different opinion. All information relating to the group discussion will be kept confidential.

However, please note that I will have to report to local authority if any illegal or abuse observed.

Please feel free to ask questions anytime if you have any. I will be happy to explain the research

in detail if you wish. It is envisaged that this research will bring about better understanding of

associations between the perception of water and wastewater usage and the reduction of risk factor

of infectious diseases amongst small scale farmers in Thai Binh and An Giang province and from

that a larger scale study is expected to follow to obtain further information and provide appropriate

recommendations for better design and implementation of programs/policies that are more suitable

to the concerns of small scale farmers, reflect the need, perceptions of farmers and their

socioeconomic status, their environment, and therefore can help improving health of farmers,

livestock, and environment sustainably (in the long run).

of 340

There is a procedure in place to make sure that all personal information will be kept private and

confidential. A code will be assigned to each interview and water sample and only I will be able

to identify which code is for whom. You may choose not to answer any of the questions and/or

decide to discontinue your participation any time if you no longer wish to participate in the research

without any penalty. If you have any questions, concerns about the research, please contact Dr.

Quynh Ba Le @ 0165 247 1058.

Thank you for your participation!

of 340

APPENDIX D: ETHICS APPROVALS

D.1. University of Calgary Ethics ID: REB13-0470

- By the Conjoint Faculties Research Ethics Boards

D.2. Hanoi School of Public Health Ethics Approval: No. 151/2013/YTCC_HD3

- By the Chair of the Ethical Review Board for Biomedical Research, Hanoi School of

Public Health

D.3. Hanoi School of Public Health Ethics Approval: No. 008/2014/YTCC_HD3

- By the Chair of the Ethical Review Board for Biomedical Research, Hanoi School of

Public Health

of 340

APPENDIX E: CHAPTER TWO’S APPENDIX

Table 2.5 (Chapter Two) shows the final results of young fish production without outliers

and stratified over the different levels of education and different levels of income among the

participating farmers. There were changes in the mean numbers of young fish production by

strata of education and income. More specifically, in Thai Binh, the participating farmers who

attended either secondary school or post-high school and had incomes of 10 to 40 million VND

per person per year were the farmers who produced the greatest mean number of young fish per

year. The stratification confirmed that the data collected from the wealthy group of farmers

relating to young fish production were outliers in the young fish production variable.

Tables E.1-3 provided additional detailed analysis of data on fish production in Thai Binh

and An Giang.

In Table E.1, I present a brief description of the fish production among those participating

farmers who produced fish. This table shows the great amounts of variation (Std. Dev) in young

fish production. The corresponding coefficients of variation of fish production in Thai Binh and

An Giang were 5.7 and 9.5 respectively. Therefore, in Table E.2, I have used median, 25

percentile, and 75 percentile to assist the presentation of data on annual fish production. Table

E.2 also shows the results without the outliers identified in Section 2.2.4. However, the

coefficients of variation of young fish production were still high 5.1 and 4.0 in Thai Binh and An

Giang respectively. Therefore, to describe the numbers of young fish produced per year in the

participating farms more clearly, I have presented these numbers of young fish produced per year

over different strata of income, education, and both income and education in Table E.3 (with

outliers) and Table 2.5 (in the main text of Chapter Two, without outliers).

of 340

Table E.1. Annual Fish Production on Small-Scale (SSI) Farms in the Provinces of Thai

Binh and An Giang in Vietnam

Provinces Mean Std. Dev.

Thai Binh

Young fish a 24,238 138,592

Adult fish b 2,060 3,585

Fish c 1,773 2,391

An Giang

Young fish 43,400,000 411,000,000

Adult fish 1,117 17,333

Fish (kg) 1,087 12,232

Total

Young fish 28,700,000 334,000,000

Adult fish 1,454 14,056

Fish (kg) 1,393 9,246

a = Young fish produced (in number); b = Adult fish produced (in number); and c = Fish produced (in kg)

Among the participating farmers who produced fish at the time of the survey, farmers in

Thai Binh and An Giang estimated the number of young fish, the number of adult fish, and the

volume of fish in kg they normally had per season (Table E.2). I did not ask type of fish in the

questionnaires; however, it appeared from our observations, in-depth interviews, and group

discussion that the majority of the participating farms raised various “trash fish and weed fish”53.

53 This term was used by the participating farmers to indicate the various types of local fish in small mixed-fish

production not for commercial purpose.

of 340

A few farms had fattening ponds, either finishing ponds, or hatching ponds mostly for Pangasius

bocourti fish in An Giang province. The number of fish farms at the time of the survey in Thai

Binh was about four times greater than that in An Giang (240 and 58 respectively) (Table E.2)

Table E.2. Quantity and Kilograms of Fish Produced per Year on Small-Scale Integrated

(SSI) Farms in the Provinces of Thai Binh and An Giang, Vietnam (without outliers)

Variable Obs % Mean Median

Std.

Dev. Min Max

25th

percentile

75th

percentile

Thai Binh

Young

fish 151 50% 4,852 1,000 13,095 20 100,000 500 3,000

Adult fish 166 55% 2,041 700 3,589 30 30,000 300 2,500

Fish (kg) 240 80% 1,779 700 2,395 5 10,000 1,700 2,000

An Giang

Young

fish 58 19% 8,159 800 68,011 3 100,000 100 3,600

Adult fish 36 12% 9,311 450 49,210 3 300,000 35 1,000

Fish (kg) 51 17% 6,296 30 28,797 2 150,000 10 250

Total

Young

fish 209 35% 5,769 1,000 15,857 3 1,000,000 500 3,000

Adult fish 202 34% 3,337 600 21,252 3 300,000 300 2,000

Fish (kg) 291 49% 2,570 625 12,495 2 150,000 200 2,000

Given the identified outliers in the variable of young fish production (Table E.2), more

detailed results of young fish production was stratified over the different levels of education and

different levels of income among the participating farmers (Table E.3). The outliers were

included in the results presented in this table. Most of the participating farmers who produced

young fish were those who had levels of education from primary school to high school. In An

Giang, it appeared that those participating farmers who had more years of attending school and

greater income produced the greatest mean numbers of young fish per year. In Thai Binh, the

participating farmers who attended up to primary school and had incomes of 10 to 20 million

of 340

VND per person per year were the farmers who produced the greatest mean number of young

fish per year. However, the standard deviations of the number of young fish produced in these

farms were still several time greater than the corresponding means. Therefore, data in table 2.5 in

the main text of Chapter Two is presented in another way (i.e., without outliers).

of 340

Table E.3. Quantity of Young Fish Produced per Year on Small-Scale Integrated (SSI)

Farms in the Provinces of Thai Binh and An Giang in Vietnam Stratified over Education

and Income (with outliers)

Provinces/

income b

Year of school attended a

1 2 3 4 5

Thai Binh

< 4.8 - 3,800.0 2,500.0 - -

- (5,377.7) (4,203.6) - -

4.8-10 - 2,469.2 35,335.9 933.6 520.0

- (2,593.1) (171,343.9) (896.5) (678.8)

10-20 - 108,600.0 5,146.1 86,516.7 14,190.0

(313,441.2) (17,010.5) (287,694.7) (21,759.3)

20-40 - 1,000.0 10,000.0 1,000.0 4,000.0

- n.ac (11,742.0) n.a n.a

>40 - 10,000.0 1,508.3 800.0 200.0

- n.a (985.1) (244.9) n.a

An Giang

< 4.8 3 1,307.4 75.0 129.5 -

- (2,001.0) (35.4) (247.0) -

4.8-10 - 1,357.6 3,100.0 1,850.0 -

- (1,544.8) (3,489.3) (2,333.4)

10-20 - 18,006.1 6,000.0 7,006.7 -

- (27,239.7) n.a (11,263.2) -

20-40 - 14,000.0 1,667Md 2.9M -

- (8,485.3) (2,080M) (4.5M) -

>40 - 3,000M - 2,500M -

n.a - (3,530M) - (.) Numbers in brackets are standard deviations; a 1 = no school; 2 = primary school; 3 = secondary school; 4 = high

school; and 5 = post-high school; b Income was reported in millions of Vietnamese dong (VND); c No s.d. is reported

where only one observation exists; d Means of young fish production greater than one million are reported in

millions denoted by M.

of 340

APPENDIX F: CHAPTER FOUR’S APPENDICES

F.1. Brief Ecology of Escherichia coli and WHO’s Recommendations for Assessing Rural

Water Quality

E. coli, a member of the coliform bacteria, is part of the Enterobacteriaceae family and is

a facultative anaerobic, Gram-negative, non-spore forming, rod-shape bacterium. E. coli is

naturally found in the intestinal tract of humans and warm-blooded animals at a concentration of

approximately 109 cells per gram of faeces (Edberg et al., 2000). E. coli is not naturally found in

soil and water or on plants (Winfield and Groisman, 2003). Surface water and ground water

(under direct influence of the surface water) are commonly contaminated with human and animal

faeces. A presence of E. coli in a source of water indicates that the source is contaminated with

faecal materials from humans and/or animals and is microbiologically not safe for drinking. E.

coli is generally sensitive to environmental stressors and does not survive in the environment for

as long as protozoans and some viruses do. E. coli can survive for 4–12 weeks in water with

moderate micro flora and a temperature of 15–18°C (Edberg et al., 2000). Most waterborne E.

coli isolates produce enzyme β-glucuronidaze (Fricker et al., 2008) which is used to facilitate the

detection of E. coli in water (FPTCW, 2013).

The Ministry of Health (MoH) of Vietnam developed its national technical regulations on

drinking and domestic water quality (MoH, 2009b; MoH, 2009a) based on the World Health

Organization guidelines for assessing rural water quality (WHO, 2011a; WHO, 2011b; WHO

and OECD, 2003a). The technical regulations and guidelines from the MoH and WHO were used

to guide the assessments of on-farm water quality in this research and are referred to hereafter as

the recommendations for assessing rural water quality. These recommendations include

of 340

microbiological and physicochemical indicators of rural water quality. Escherichia coli or E. coli

was used as an indicator organism of faecal contamination of water and possible presence of

other harmful bacteria and viruses. Levels of pH and turbidity were used as these indicators are

positively correlated with pathogen removal in water.

Figures F.1 summarizes the main characteristics of the coliform group and the

corresponding recommendations in assessing microbial water quality based upon information

obtained from the World Health Organization and the Organization for Economic Co-operation

and Development (WHO and OECD, 2003a). There are several reasons why E. coli is

recommended as an indicator organism for assessing microbial safety of rural water. First of all,

it is not practical to conduct isolations of all pathogens in water due to the large number existing

pathogens, its uneven distribution in water, and required resources for the isolation (Payment and

Pintar, 2006). Secondly, pathogens from human and animal faeces pose the greatest danger to

public health compared to other contaminants that are regularly found in water. This makes the

ability to detect faecal contamination in drinking water important for public health safety

(FPTCW, 2013). Thirdly, current methods of assessing E. coli have been simpler, more

available, and more affordable than in late 1980s (Edberg et al., 2000). Fourthly, E. coli is

considered a more specific indicator of faecal contamination compared to other members of the

total coliform bacteria group, and can be rapidly identified and easily quantified in water. E. coli

has a similar lifespan, comparable to that of other enteric bacterial pathogens, does not usually

multiply in the environment, and can be detected even if it is greatly diluted due to the great

number of E. coli that are excreted in faeces (FPTCW, 2013; WHO, 2011b; WHO and OECD,

2003b).

of 340

In addition to assessing E. coli level in on-farm water, other WHO recommended basic

physical indicators of on-farm water quality (i.e., level of pH and turbidity) were also assessed.

Levels of pH and turbidity in water have been shown to influence microbial quality of water

(WHO and OECD, 2003b).

Figure F.1. The Coliform Group

Source: This figure is created using information from WHO and OECD, 2003.

WHO recommends pH and turbidity levels in water to be included as parts

Total coliforms

Thermotol--erant

coliforms

E. coli

Main characteristics Recommendations

- Include Escherichia, Citrobacter, Enterobacter,

and Klebsiella genera

- Include lactose fermenting bacteria (e.g.,

Enterobacteria cloacae and Citrobacter freundii) which can be found in both faeces,

environment, and drinking water.

- Include members of genera (e.g., Budvicia, and

Rahnella) that are never found in mammalian

faeces - Oxidase negative, ferment lactose at 35-37oC

with the production of acid, gas, and aldehyde within 24-48 hours

- Possess β-galactosidase

- Not an index of

faecal

contamination or of health risk

- Basic information

on source water quality

- Easy to detect and

enumerate in water

- Account for 60% to 90% of the total coliforms

group

- Include primarily the genus of Escherichia (Hall, 1999), and to lesser extent Klebsiella,,

Enterobacter, and Citrobacter

- May originate from organically enriched water

and does not necessarily suggest faecal contamination except E. coli

- Ferment lactose at 44-45oC

- Less reliable index

of faecal contamination than

E. coli but

acceptable for

water-quality

examination when

on other method is available

- A well-defined member of the family

Enterobacteriaceae

- Possession of β-galactosidase β-glucuronidase

- Most strains of E. coli grow at 44-45 oC,

ferments lactose and mannitol with production of

acid and gas and produces indole from

tryptophan

- Rapid and reliable tests for E. coli are available.

- Found in water and environment subject to

recent faecal contamination

- Abundant in human and animal faeces, fresh

faeces may attain concentration of 109 of E. coli per gram.

- Widely preferred

as an index of

faecal contamination

of 340

of the core set of microbial and related indicators in assessing and monitoring water quality

(WHO, 2012).

F.2. Protocols Used for Assessing On-Farm Water Quality

Two protocols were used in this research to assess the quality of on-farm water. These

protocols included a protocol for laboratories to assess on-farm water quality (i.e., pH, turbidity,

and E. coli) and a protocol for community health/veterinary workers (CHWS/CVWs) to quantify

E. coli. This appendix provides detailed information of these protocols.

F.2.1. Protocol for Laboratories to Assess Basic Microbial and Related Water Quality

This section elaborates the first protocol used for collecting and testing water quality in

the participating SSI farms. Both drinking water54 and domestic water55 were assessed using the

basic indicators and methods recommended by WHO. This protocol was used by two national

microbiological laboratories of the Vietnam National Institute of Veterinary Research and the Ho

Chi Minh City Institute of Public Health - to assess the basic indicators of on-farm water quality.

The basic indicators included: 1) number of E. coli colony forming units (cfu) in on-farm water

(using membrane filtration method); 2) level of pH of on-farm water; and 3) level of turbidity of

on-farm water. The research team collaborated with two national microbiological laboratories -

one in the north and one in the south of Vietnam - to collect and test water samples in Thai Binh

and An Giang respectively. Both laboratories followed the same national laboratory protocol for

assessing water quality which conforms to WHO/UNICEF water testing standards (MoH, 2009b;

MoH, 2009a).

54 Drinking water is defined as “water used for direct drinking or processing food” (MoH, 2009a) 55 Domestic water is defined as “water used for domestic use but not for direct drinking or processing food” (MoH

and MARD, 2011)

of 340

Basic indicators of water quality assessed included: 1) E. coli colony forming unit

(cfu)/100ml in on-farm water (using membrane filtration method); 2) pH of on-farm water; and

3) turbidity of on-farm water. Membrane filtration (MF) method is one of the two common

methods used to detect and quantify coliform bacteria in water (i.e., “multiple fermentation

tube/most probable number method” and “membrane filter/membrane filtration method”. The

MF method was selected because of its advantages over the multiple fermentation tube in

assessing large number of water samples. Examples of the advantages include more rapid

yielding results given relatively large numbers of samples, less labor intensive, results obtained

directly by colony count (Bartram and Pedley, 1996; TCVN 6187, 2009; WHO, 2011a).

With respect to collection of water samples, each national microbiological laboratory sent

a technical team to the participating farms to collect two water samples. In each province, 34

(11%) and 63 (21%) farms in Thai Binh and An Giang respectively had duplicated water

samples used for testing of E. coli, pH, and turbidity. For duplicated samples, mean results of

two samples were reported. At each of the participating farm, two water samples of 200ml each

were collected for testing of the three basic indicators - one sample was for drinking water

testing and the other sample was for domestic water testing. Collected samples of water were

treated within 6-8 hours of sampling. The collection, storage, and transportation of the water

samples followed the national guidance on water quality and sampling (TCVN 6663, 2011).

With regards to the MF method, the participating laboratories followed the technical

guidelines for water quality testing of Vietnam MoH (TCVN 6187, 2009). These guidelines

conformed to standards for detection and enumeration of faecal indicator organisms in water by

WHO and ISO (i.e., ISO 9301 - 12:1990). Four key steps of the MF method conducted include

of 340

filtration, incubation, confirmatory tests, and calculation of results. The number of E. coli colony

forming units (cfu) in each 100ml of sample water was calculated using the following formula:

𝐶(𝑐𝑓𝑢 𝑖𝑛 𝑉 𝑚𝑙) =𝐴 × 𝑁

𝐵

C: Number of E. coli cfu confirmed in 100ml of sample water

A: Number of presumptive E. coli cfu that were positive with Indole test

B: Number of presumptive E. coli cfu transferred for further incubation in Trypone Soy Agar

(TSA) medium

N: Number of presumptive E. coli cfu on filtered incubated membranes with Lactose TTC agar

medium

V: Volume filtered sample water

The national laboratories followed the national technical standards and guidelines for

determination of pH level using pH meter (TCVN 6492, 2011; IHPH, 2013a) and followed the

EPA method 180.1 (Powell, 2007) for determination of turbidity by Nephelometry unit using a

portable turbidity meter (i.e., Hatch Model 2100P) (EPA, 1993; IHPH, 2013b). The field

investigator tested the protocol in a pilot study of 30 farms before using it in the full research.

F.2.2. Protocol for Community Health/Veterinary Workers to Quantify E. coli

The second protocol was for community health/veterinary workers CHW/CVWs to

collect and test drinking water samples of the participating SSI farms. The protocol was color-

changed quantification of total coliforms and E. coli bacteria in water using ColiplateTM tests

(Bluewater Biosciences, 2013). ColiplateTM is an immunofluorescence plate test of water

produced by Bluewater Biosciences Inc. Canada. The use of ColiplateTM tests were supported

financially by a doctoral research award (IDRA) provided by the International Development

of 340

Research Centre (IDRC, 2012). This protocol conforms to EPA based test regulatory guidelines

(Bluewater Biosciences, 2013). I selected and trained twenty CHW/CVWs in 1) water sample

collection procedure following the Vietnam Ministry of Health guidelines for collecting water

samples for microbial (MoH, 2009b) and 2) using the ColiplateTM testing procedure (e.g., testing

steps and handling storage, cleanup and disposal of the test kits) . For each farm, a 96-well

immunofluorescence ColiplateTM plate test was used by CVW (in the province of Thai Binh) and

CHW (in the province of An Giang) to quantify coliform bacteria and E. coli in on-farm water.

The field investigator tested ColiplateTM tests before being used by the CHWs/CVWs. Water

quality tests were carried out on fresh water samples collected on site in the communities.

The procedure used in this survey for testing on-farm water is described as follows. The

CHWs/CVWs collected water samples on farms, stored it in cool Styrofoam boxes, and

transported the collected water samples back to commune health/veterinary stations. Within six

hours since the water samples were collected, CHWs/CVWs poured the water samples into the

provided 96-well ColiplateTM plates and started incubating the water samples in Styrofoam boxes

in four days. Each CHW/CVW was instructed to check and take notes in a given paper form if

any of the 96 wells in each of the ColiplateTM plates with the sample water turns blue during the

four-day incubation. One water sample was used in one 96-well ColiplateTM plate. ColiplateTM

tests have varied incubation time depending on incubation temperatures - from 24 hours (at 35-

37oC) to 60 hours (at 20-25oC). After incubation, the CHWs/CVWs reported the numbers of

wells with blue water in each plate. These blue wells indicate the presence of coliform bacteria.

To quantify E. coli, CHWs/CVWs used ultraviolet (UV) lamps, provided by the research, to light

on the blue colored wells in a dark room. All wells that reflected the UV light with fluorescent

of 340

light were the wells that have E. coli. CHWs/CVWs counted the numbers of blue wells with

fluorescent light reflection. The numbers of blue wells and blue wells with fluorescent light

reflection were used to demine the number of colony-forming units (cfu) of coliform bacteria and

E. coli respectively using a Most Probable Number (MPN) table.

F.3. Detailed Model Building Strategies for MLR and SUR

Regression analyses (Greene, 2008a; Greene, 2008b) were used to model the association

between the objectively assessed on-farm water quality and its potential farm related associated

factors. Sections below present detailed strategies used for building MLR and SUR models and

detailed model fitting/selecting procedures used in each model for analysing research data.

F.3.1. Modeling building strategies

The overall strategy for building models for this chapter was to start with a multiple

linear regression (MLR) model, then use a ‘seemingly unrelated regression’ (SUR) model

(Greene, 2008a) to analyze associated factors of the quality of on-farm drinking water and on-

farm domestic water. Within each of the MLR and SUR models, four specific model building

strategies (Dohoo et al., 2003; Greene, 2008a) were used: specifying a full model; specifying

selection criteria for independent variables; specifying selection strategies for fitting model; and

analysing the data using my knowledge and experiences of water public health in Vietnam and

statistical software.

F.3.2. Multiple linear regression (MLR) model

I used MLR models to start exploring associations between objectively assessed quality

of various sources of on-farm water for drinking and domestic use (i.e., E. coli cfu in various on-

farm water sources – the dependent variables) and subjectively assessed water quality and other

of 340

appropriate on-farm factors (e.g., farmers’ perceptions - potential independent variables). In each

MLR model, a set of appropriate independent variables were regressed against a dependent

variable. The initial potential set of independent variables was determined based on the

conceptual and theoretical model of this research (presented in Chapter Three), the hypotheses of

this research, and the literature. Tables F.1a-d presented pairwise correlations between the

independent variables.

Four selection procedures were used in the MLR models. First, in a full simultaneous

procedure, a dependent variable (e.g., E. coli in stored rainwater for drinking water) was

predicted from the full set of potential independent variables simultaneously (e.g., gender, age,

and years of attending school). Second, in a stepwise regression procedure, I used a syntax that

started with backward elimination then forward selection. After the elimination of each variable

(i.e., the highest non-significant variable with p-value > a predefined α), all the eliminated

variables were checked to see if any of them should be included using forward selection criterion

(i.e., the lowest significant variable p-value <= α). A significance level of α < or => 0.15 was

used for adding and deleting new variables to/from the model respectively. Third, in a forward

selection procedure, Stata was commanded to start a model with only an intercept then

selectively added variables that meet the criterion that I specified (i.e., p-value < a specific α

level). At each adding step, a variable with the lowest significant p-value (p-value <= a

recommended level of α = 0.25) will enter the model. The adding variable process continued

until no more variables meet the adding criteria. Fourth, in a backward elimination procedure,

steps were reversed compared to that of the forward selection procedure. The backward

elimination procedure started with a maximum model then variable(s) that meet the elimination

of 340

criteria were removed. In particular, at each step, a variable with the highest non-significant p-

value (p-value > α) was removed from the model. A recommended significance level of α = 0.25

was used for removing variables from the model.

With respect to fitting the models and determining independent variables that need to be

retained in the final models, I considered two sets of criteria including 1) my research interest

and a priori to be potential confounders and 2) p-values, F-test, and Adjusted R squared before

presenting the results of the final fittings of the MLR model.

The MLR model used in this Chapter is written as the following:

Yi = β0+ βi Xi + ɛi; with i = 1, 2, 3... n

Yi is a continuous dependent variable representing on-farm water quality, considering

bundle i. Each Y is the numbers of E .coli cfu in 100 ml of a source of on-farm water.

Xi is a vector of IVs (continuous and/or dichotomous) believed to be relevant to the

consideration of ith equations (i = 1, 2, 3... n).

βo is a intercept or constant to be estimated,

βi is a vector of unknown coefficients describing the direction and magnitude of the effect

of an Xi on Yi after controlling for the effects of other Xi; βi to be estimated,

ɛi is error term

F.3.3. Seemingly Unrelated Regression (SUR) Model

I assumed residuals of the two equations associating the quality of on-farm drinking

water and on-farm domestic water from other factors are related. Therefore, in addition to the use

of the MLR model, I used a SUR model to explore systematically the associations between a

dependent variable and the corresponding different set of independent variables. The assumption

of 340

of correlated residuals was based on the facts that the SSI farming practice was used among the

participating SSI farmers and that both quality of on-farm drinking water variables and quality of

on-farm domestic water variables come from the same dataset. This assumption was tested when

fitting the SUR model. In addition, some potential factors that may be associated with quality of

on-farm drinking water may not be the same for quality of on-farm domestic water. More

specifically, in the SUR model, two MLR equations were considered simultaneously to explore

associated factors (potential independent variables) of E. coli cfu in on-farm drinking water as

well as domestic water (Long and Freese, 2014). Potential IVs in these two equations were

expected not to be the same. The first MLR equation is for E. coli cfu in on-farm drinking water

verse a set of potential IVs. The second MLR equation is for E. coli cfu in on-farm domestic

water verse other set of Independent variables.

The model for the two equation is written as below with j = 1… m MLR equations:

yj = x’j βj + uj

These two equations are stacked into an SUR model:

[

𝑦1..

𝑦𝑚

] = [𝑥1 0 ⋯ 0⋮ ⋱ ⋮0 ⋯ 𝑥𝑚

] [

β1..

βm

] + [

u1..

um

]

The error terms are assumed to have zero mean and to be independent across farmers and

homoscedastic. For a given farm, the errors are correlated across equations:

E(uij uij’|𝑋) = σjj’ and σjj’ ≠ 0 where j ≠ j’

Cross-equation restrictions were tested, and constraints were imposed for IVs that were

identified in the fitting of both equations and were not significantly different from each other.

The SUR model was re-estimated.

of 340

Table F.1a. Pairwise Correlation between Independent Variables used for Regression

Models

Independent variables a Age DGender Schooling DIncome DChildren

Experie

nce DOff-farm job

Age 1.0

DGender 0.1 1.0

Schooling -0.1* 0.0 1.0

DIncome -0.1* -0.1* -0.2* 1.0

DChildren -0.4* 0.0 0.0 0.* 1.0

Experience 0.3* 0.0 0.1* -0.2* -0.1* 1.0

DOff-farm job -0.2* -0.1* 0.0 0.2* 0.1* 0.0 1.0

DFPR-Livestock-Man -0.1 0.2* 0.1 -0.1 0.1* 0.0 0.0

DFPR-Livestock-Woman 0.0 -0.4* 0.0 0.0 0.0 0.1 0.1

DFPR-Fish-Woman 0.0 -0.3* 0.0 0.0 -0.1 0.0 0.0

DFPR-Fish-Man 0.0 0.1 0.2* -0.2* 0.0 0.1 0.0

DFPR-Health-Man 0.0 0.2* 0.0 -0.1 0.0 0.1* -0.1*

DProvince 0.1* -0.1 0.4* -0.4* -0.3* 0.3* -0.1*

DFish prod 0.0 -0.2* 0.3* -0.2* -0.2* 0.1* 0.0

DRain for animals 0.1* 0.1* 0.0 -0.1 -0.1 0.1* 0.0

DPond for animals 0.0 0.0 0.0 0.0 -0.1 0.0 0.0

DPipe for animals 0.0 -0.1* 0.0 0.3* 0.0 -0.1* 0.1

DMulti-agri activities 0.0 0.1* 0.0 -0.2* 0.0 0.1 -0.1*

DPoultry AI -0.1 0.1 0.0 0.2* 0.1* -0.2* -0.1

Npigs/yr 0.0 0.0 0.2* -0.2* 0.0 0.0 -0.1*

Nducks/yr 0.1 0.1* 0.0 -0.1* -0.2* -0.1 -0.1*

Npoultry/yr 0.0 0.0 0.2* -0.2* -0.1* 0.0 0.0

DFPWater quality 0.0 -0.1* 0.1 -0.2* 0.0 0.1* 0.0

DFPBarr-Cost 0.0 -0.2* 0.0 0.1* 0.1* 0.0 0.2*

DFPRisk rain water 0.0 0.0 0.0 -0.2* 0.0 0.1 0.0

DFPS-AI-Drink 0.0 -0.1* 0.0 0.0 0.0 0.0 0.0

DFPS-Diarrhea-Drink 0.0 0.0 0.1 0.0 -0.1 0.0 0.0

DFPS-Coliform-Drink . . . . . . .

DFPS-Parasite-Drink 0.0 0.0 0.1 -0.1 0.0 -0.1 0.0

DFPS-AI-Dom 0.1 0.1* 0.0 0.0 0.0 -0.1 0.0

DFPS-Coliform-Dom 0.1 0.3* -0.1* -0.2* -0.1 -0.1 -0.1

DFPS-Diarrhea-Dom 0.1 0.2* 0.1* 0.0 0.0 0.0 0.0

DFPS-Parasite-Dom 0.0 0.2* -0.1 0.0 0.0 -0.1 -0.1

DSatisfy-Drink 0.1 0.1* -0.2* 0.0 0.0 -0.2* 0.0

DSatisfy-Dom 0.1* 0.2* -0.2* -0.2* -0.1 -0.1* 0.0

DAI heard 0.1* 0.0 0.3* -0.1 -0.1* 0.1* 0.0

DFPHarm-Dom 0.0 0.1* 0.1* 0.0 0.0 0.0 0.0

DFPHarm-Drink 0.1 0.0 0.0 -0.1 0.0 0.0 0.0

DDump Wastes -0.2* -0.1* 0.1* 0.1* 0.1* -0.1 0.1*

DLivestock Management 0.1 0.0 0.3* -0.4* -0.1* 0.2* -0.1*

DTest-dom 0.0 0.1* -0.1 0.1 0.0 -0.1 -0.1* a See Tables 4.6 in chapter Four for descriptions of the independent variables

of 340

Table F.1b. Pairwise Correlation between Independent Variables used for

Regression Model (cont’d)

Independent

variables a

DFPR-

Livest

ock-

Man

DFPR-

Livestock-

Woman

DFPR-Fish-

Woman

DFPR-

Fish-Man

DFPR-

Health-Man

DProvin

ce DFish prod

DFPR-Livestock-

Man 1.0

DFPR-Livestock-

Woman -0.4* 1.0

DFPR-Fish-Woman -0.2* 0.6* 1.0

DFPR-Fish-Man 0.5* 0.0 -0.2* 1.0

DFPR-Health-Man 0.5* -0.1* -0.1* 0.4* 1.0

DProvince -0.1 0.1 0.1* 0.3* 0.0 1.0

DFish prod -0.1* 0.1 0.1* 0.3* -0.1 0.7* 1.0

DRain for animals 0.1 -0.1 0.0 0.1 0.1 0.1* 0.1*

DPond for animals -0.1 0.0 0.0 0.1 0.1* 0.2* 0.3*

DPipe for animals -0.2* 0.1 0.0 -0.1* -0.1* -0.1 0.1

DMulti-agri

activities 0.0 -0.1 0.0 0.0 0.0 0.2* 0.3*

DPoultry AI 0.1* -0.1 -0.1 0.0 0.0 -0.2* -0.1*

Npigs/yr -0.1 0.0 0.0 0.1* 0.0 0.5* 0.3*

Nducks/yr -0.1* 0.0 0.0 -0.1 -0.1 0.3* 0.2*

Npoultry/yr 0.0 0.0 0.0 0.2* 0.0 0.4* 0.3*

DFPWater quality -0.1* 0.1 0.1 0.1* -0.1* 0.3* 0.3*

DFPBarr-Cost 0.1* 0.0 -0.1 0.0 -0.1* -0.2* -0.1

DFPRisk rain water 0.1 0.0 0.0 0.0 0.0 0.1* 0.1

DFPS-AI-Drink 0.0 0.1 0.0 -0.1* 0.0 -0.1 -0.1

DFPS-Diarrhea-

Drink 0.0 0.0 0.0 0.0 0.1 0.0 0.0

DFPS-Coliform-

Drink . . . . . . .

DFPS-Parasite-Drink 0.1 -0.1 -0.1* 0.1 0.0 0.0 0.0

DFPS-AI-Dom -0.1* 0.0 0.1 -0.2* -0.1* 0.1 -0.1

DFPS-Coliform-

Dom -0.2* -0.1* 0.0 -0.1* 0.0 0.0 0.0

DFPS-Diarrhea-Dom 0.1 0.0 0.1* 0.1 0.0 0.2* 0.1

DFPS-Parasite-Dom -0.1 -0.1 0.1 -0.1 -0.1 0.0 0.0

DSatisfy-Drink 0.0 0.0 0.0 -0.2* 0.0 -0.3* -0.3*

DSatisfy-Dom 0.0 0.0 0.0 -0.1* -0.1 -0.1 -0.1*

DAI heard 0.1* 0.0 0.1 0.2* 0.1 0.3* 0.2*

DFPHarm-Dom 0.0 0.0 0.0 0.0 0.0 0.2* 0.1

DFPHarm-Drink 0.1 0.0 -0.1 0.0 0.0 0.0 0.0

DDump Wastes 0.1* 0.0 0.0 0.3* 0.0 0.3* 0.2*

DFPE-Livestock

Management 0.0 0.0 0.1 0.2* 0.0 0.6* 0.4*

DTest-dom 0.1 -0.1* -0.1 0.0 0.1 -0.2* -0.2* a See Tables 4.6 in chapter Four for descriptions of the independent variables

of 340

Table F.1c. Pairwise Correlation between Independent Variables used for

Regression Models (cont’d)

Independent

variables a

DRain

for

animals

DPond for

animals

DPipe for

animals

DMulti-

agri

activities DPoultry AI Npigs/yr

Nducks/

yr

DRain for animals 1.0

DPond for animals 0.3* 1.0

DPipe for animals 0.0 0.2* 1.0

DMulti-agri

activities 0.1 0.2* -0.2* 1.0

DPoultry AI -0.1* 0.0 0.1* -0.1 1.0

Npigs/yr 0.1* 0.3* 0.0 0.2* -0.1* 1.0

Nducks/yr 0.1* 0.3* 0.0 0.1* -0.1* 0.3* 1.0

Npoultry/yrs 0.1* 0.1* -0.1* 0.1* -0.1* 0.5* 0.2*

DFPWater quality 0.1 0.3* 0.0 0.3* -0.2* 0.2* 0.1*

DFPBarr-Cost 0.0 -0.1* 0.1 -0.2* 0.0 -0.2* -0.2*

DFPRisk rain

water 0.1* 0.0 -0.1* 0.1* -0.1* 0.1 0.0

DFPS-AI-Drink 0.1* 0.1* 0.0 -0.1 0.0 0.0 -0.1*

DFPS-Diarrhea-

Drink 0.1* 0.0 0.0 0.0 0.0 0.0 0.0



DRain for

animals

DPond for

animals

DPipe

for

animals

DMulti-

agri

activities

DPoultr

y AI Npigs/yr Nducks/yr

DFPS-Parasite-Drink 0.0 0.0 0.0 0.0 0.0 0.0 0.0

DFPS-AI-Dom 0.2* 0.0 -0.1* 0.1* 0.1 0.2* 0.1*

DFPS-Coliform-Dom 0.3* 0.4* 0.0 0.3* 0.0 0.3* 0.3*

DFPS-Diarrhea-Dom 0.3* 0.2* -0.1 0.2* 0.0 0.3* 0.2*

DFPS-Parasite-Dom 0.3* 0.2* 0.0 0.2* 0.0 0.2* 0.2*

DSatisfy-Drink -0.1* -0.2* -0.1* -0.1 0.1 -0.1* 0.1

DSatisfy-Dom 0.0 0.0 -0.1* 0.0 -0.1* 0.0 0.1*

0.2* 0.1* 0.1 -0.1 0.0 0.1* 0.1*

DAI heard 0.3* 0.2* 0.0 0.1* 0.0 0.2* 0.2*

DFPHarm-Dom 0.0 0.0 0.0 0.0 0.0 -0.1* 0.0

DFPHarm-Drink 0.0 0.1* 0.1* 0.1* 0.0 0.1 0.0

DDump Wastes 0.1* 0.1 -0.2* 0.1* -0.1* 0.3* 0.2*

DFPE-Livestock

Management 0.0 0.2* 0.0 0.0 0.1* -0.1* -0.1

DTest-dom -0.1 -0.2* -0.1* -0.1 0.1 -0.1* 0.1


DFPS-

Parasite-

Drink

DFPS-

AI-Dom

DFPS-

Coliform-

Dom

DFPS-

Diarrhea-

Dom

DFPS-

Parasite-

Dom

DSatisf

y-Drink

DSatisfy

-Dom

DFPS-Parasite-Drink 1.0

DFPS-AI-Dom 0.1 1.0

DFPS-Coliform-Dom . 0.6* 1.0

DFPS-Diarrhea-Dom 0.1* 0.6* 0.7* 1.0

DFPS-Parasite-Dom 0.0 0.6* 0.8* 0.7* 1.0

DSatisfy-Drink -0.1 0.1* 0.1 0.0 0.0 1.0 a See Tables 4.6 in chapter Four for descriptions of the independent variables

of 340

Table F.1d. Pairwise Correlation between Independent Variables used for

Regression Models (cont’d)


Npoultry/

yr

DFPWater

quality

DFPBarr

-Cost

DFPRisk

rain water

DFPS-

AI-

Drink

DFPS-

Diarrhea-

Drink

DFPS-

Coliform-

Drink

Npoultry/yr 1.0

DFPWater quality 0.1* 1.0

DFPBarr-Cost 0.0 -0.1 1.0

DFPRisk rain water 0.0 0.0 0.0 1.0

DFPS-AI-Drink 0.0 0.0 0.1 0.1* 1.0

DFPS-Diarrhea-Drink 0.0 0.0 0.0 0.1* 0.2* 1.0


DFPS-Parasite-Drink 0.0 0.0 0.0 0.2* 0.2* 0.8* .

DFPS-AI-Dom 0.0 0.0 -0.2* 0.1 0.2* 0.1* .

DFPS-Coliform-Dom 0.1 0.0 -0.4* 0.1 0.0 . .

DFPS-Diarrhea-Dom 0.2* 0.1 -0.2* 0.1* 0.1* 0.1* .

DFPS-Parasite-Dom 0.1* 0.1* -0.4* 0.1 0.0 -0.1 .

DSatisfy-Drink -0.1 -0.3* -0.1 0.0 0.0 -0.1 .

DSatisfy-Dom 0.0 -0.1* 0.0 0.1* 0.0 0.1 .

DAI heard 0.2* -0.1 0.1* 0.1 0.0 0.0 .

DFPHarm-Dom 0.2* 0.0 -0.2* 0.1* 0.1 0.1 .

DFPHarm-Drink 0.0 0.0 0.0 0.1* 0.1* 0.2* .

DDump Wastes 0.1* 0.1* 0.1 0.0 -0.1 0.1 .

DFPE-Livestock

Management 0.3* 0.2* -0.1* 0.1* -0.1 0.0 .

DTest-dom -0.1* -0.1* 0.0 0.0 0.1 0.0 .

Independent

variables a

DFPS-

Parasite-

Drink

DFPS-AI-

Dom

DFPS-

Coliform-

Dom

DFPS-

Diarrhea-

Dom

DFPS-

Parasite-

Dom

DSatisfy-

Drink

DSatisfy-

Dom

DSatisfy-Dom 0.0 0.1 0.0 0.0 0.0 0.3* 1.0

DAI heard 0.0 0.0 0.1 0.2* 0.2* -0.1* 0.0

DFPHarm-Dom 0.1 0.5* 0.6* 0.6* 0.6* 0.0 0.0

DFPHarm-

Drink 0.3* 0.1* 0.1 0.1 0.0 0.0 0.0

DDump Wastes 0.0 0.0 -0.2* 0.0 0.0 -0.2* -0.1*

DFPE-Livestock

Management 0.0 0.1* 0.1 0.2* 0.0 -0.2* 0.0

DTest-dom 0.0 0.0 0.0 0.0 0.0 -0.1* 0.1*

Independent

variables a DAI heard

DFPHarm-

Dom

DFPHarm-

Drink

DDump

Wastes

DLivestock

Management

DTest-

dom

DAI heard 1.0

DFPHarm-Dom 0.2* 1.0

DFPHarm-

Drink 0.0 0.1* 1.0

DDump Wastes 0.2* 0.0 0.0 1.0

DFPE-Livestock

Management 0.2* 0.1* -0.1 0.0 1.0

DTest-dom 0.0 0.0 0.0 0.0 0.0 1.0 a See Tables 4.6 in chapter Four for descriptions of the independent variables

of 340

F.3.4. Detailed Results of Sources of Water Used and Corresponding Frequencies of Use

among SSI Farmers in the Provinces of Thai Binh and An Giang

Table F.2. Mean Frequency of Use of On-Farm Water for Drinking in the Provinces of

Thai Binh and An Giang, Vietnam

Province/Sources of

water

Mean frequency of use for drinking by source of water (*)

Obs Mean Std. Dev. Min Max

Thai Binh

Rain 196 3.4 1.3 1 5

Drilled well 231 3.2 1.3 1 5

Bottled 169 2.4 1.1 1 5

Pipe 61 2.4 1.4 1 5

Dug well 40 1.7 1.0 1 4

Pond 143 1.0 0.2 1 3

River/Canal 143 1.0 0.2 1 3

An Giang

River/Canal 297 3.0 1.3 1 5

Pipe 243 2.7 1.8 1 5

Bottled 280 2.5 1.2 1 5

Rain 277 2.4 1.0 1 5

Pond 271 1.3 0.7 1 4

Drilled well 219 1.0 0.2 1 2

Dug well 217 1.0 0.1 1 2 (*) 1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, to 5 = Very often

of 340

Table F.3. Mean Frequency of Use of On-Farm Water for Domestic Purposes in the


Province/

Sources of water Drink rank+

Mean frequency of use for domestic purposes (*)

Obs Mean Std. Dev. Min Max

Thai Binh

Drilled well 2 251 3.5 1.2 1 5

River/Canal 7 251 3.2 1.2 1 5

Pond 6 209 2.9 1.4 1 4

Rain 1 256 2.5 0.9 1 5

Dug well 5 38 1.9 1.1 1 4

Pipe 4 52 1.8 1.0 1 4

Bottled 3 108 1.2 0.5 1 3

An Giang

River/Canal 1 288 3.8 1.0 1 5

Rain 4 267 2.1 1.0 1 4

Bottled 3 254 1.5 0.8 1 4

Pond 5 260 1.4 0.8 1 4

Pipe 2 245 1.4 0.9 1 5

Drilled well 6 227 1.1 0.6 1 5

Dug well 7 226 1.0 0.2 1 3

(*) 1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, to 5 = Very often

(+) Drink rank = the ranking for drinking for these water sources (see Table 4.1, Chapter Four)

F.4. Detailed Results of Basic Microbial and Related Indicators of On-Farm Water

F.4.1. Quality of On-Farm Water Used for Drinking

The mean results of pH, turbidity, and E. coli cfu/100ml of sources of on-farm drinking

water are presented for both provinces aggregately (Table F.4). The mean pH levels of these

sources of water were from 6.8 to 7.2 (with an exception of bottled water, which had a mean pH

of 9.4). These sources of water (except river water) had mean turbidity levels that were below 5

Nephelometric Turbidity Unit (Powell, 2007). All these sources of on-farm water used for

drinking were contaminated with E. coli. The mean number of E. coli colony-forming units

of 340

(cfu)/100ml in the sources of on-farm water used for drinking ranged from eight to 46.9. Bottled

water had the lowest mean while stored rainwater had the highest.

Table F.4. Mean pH, Turbidity, and E. coli cfu of On-Farm Water Used for Drinking in the


Variable N Mean Std. Dev. Min Max

pH

Rain water 148 6.8 3.0 4.9 42.8

Pipe water 118 7.2 2.6 6.0 32.6

Well water 172 6.9 5.3 5.5 76.3

Bottled water 43 9.4 14.0 6.4 99.2

River water (fl.) 119 7.0 0.4 5.8 9.6

Turbidity (measured in Nephelometric Turbidity Unit)

Rain water 148 0.9 0.9 0.0 3.2

Pipe water 114 1.7 2.7 0.0 18.2

Well water 172 3.7 9.2 0.0 72.4

Bottled water 43 0.8 0.4 0.3 1.5

River water (fl.) 120 7.6 20.9 0.2 167.0

E. coli (measured in number of E. coli colony-forming units per 100 mL of sample water)

Rain water 150 46.9 74.0 0.0 232.0

Pipe water 117 10.1 17.1 0.0 50.0

Well water 156 24.1 40.0 0.0 213.0

Bottled water 42 8.0 20.1 0.0 110.0

River water (fl.) 117 12.8 30.5 0.0 260.0

F.4.2. Quality of On-Farm Water Used for Domestic Purposes

The mean levels of pH in well domestic water in both provinces were within the range of

6.8 – 7.3 (Table F.5). The mean turbidity level was greatest in river water (i.e., 61.4) compared

to that in well water and pond water (i.e., 15.3 and 39.9 respectively). Pond water was used the

most for domestic purposes and had much greater mean numbers of E. coli cfu (i.e., 2596.6)

compared to that in well water and river water (i.e., 348.1 E. coli cfu and 817.8 E. coli cfu

respectively).

of 340

Table F.5. Mean pH, Turbidity, and E. coli cfu of On-Farm Water Used for Domestic

Purposes in the Provinces of Thai Binh and An Giang, Vietnam

Variable N Mean Std. Dev. Min Max

pH

Well water 31 6.8 1.2 0.8 7.9

Pond water 354 6.7 0.7 0.6 8.5

River water 208 7.3 0.4 6.0 8.1

Turbidity (measured in Nephelometric Turbidity Units)

Well water 29 15.3 37.4 0.0 185.0

Pond water 352 39.9 27.1 0.0 155.0

River water 211 61.4 42.6 5.6 300.0

E. coli (measured in number of E. coli colony-forming units per 100 mL of sample water)

Well water 30 348.1 536.9 0.0 1,800.0

Pond water 353 2,596.6 3,237.5 0.0 12,400.0

River water 213 817.8 1,077.9 0.0 3,200.0

F.4.3. Detailed Results of Drinking Water Quality Assessed by CHW/CVW using ColiplateTM

Table F.6. Mean MPN a of E. coli cfu Using ColiplateTM Tests in On-Farm Water for

Drinking by Sources and by Number of Incubation Day in the Provinces of Thai Binh and

An Giang, Vietnam

Province/Water source

Incubation time (day)

1 2 2.5 3 4

Thai Binh

Rain - - - 47.3 19.5

Pipe - - - 11.8 7.0

Well - - - 27.5 46.9

Bottle - - - 2.7 -

An Giang

Rain - 77.5 950.1 105.2 -

Pipe 87.0 124.7 1,685.6 109.2 34.6

Well - 18.0 - - -

Bottle 39.0 5.9 - 56.9 -

River with flocculation 59.1 461.9 704.8 94.8 151.0

Pipe (stored in reservoir) - 32.6 - 132.0 - a MPN: Most Probable Number (MPN) of E. coli colony-forming units per 100mL sample water

of 340

F.5. Detailed Results of MLR Models for E. coli in Sources of Water On Farms

Table F.7a. Summary of the Best MLR Model Fittings for E. coli cfu in Stored Rainwater for Drinking by Selection

Procedures in the Provinces of Thai Binh and An Giang, Vietnam

Independent variables

a Fullb Stepc Forwd Bacwe


DGender -5.66 0.76 - - - - - -

Age 0.79 0.22 0.76 0.17 1.01* 0.05 0.76 0.17

Schooling -0.37 0.88 -2.86 0.23 - - -2.86 0.23

Experience -1.02 0.20 -0.89 0.22 -1.08 0.13 -0.89 0.22

DIncome -37.80* 0.09 -34.52* 0.09 -28.20 0.14 -34.52* 0.09

DChildren -13.49 0.36 - - - - - -

DOff-farm job 11.14 0.42 - - - - - -

Npoultry/yr 0.04 0.74 - - - - - -

DFish prod - - 50.07** 0.00 48.70** 0.00 50.07** 0.00

DFPR-Livestock-Man 7.63 0.63 - - - - - -

DFPR-Livestock-

Woman 25.80 0.20 17.93 0.24 - - 17.93 0.24

DFPWater quality -7.55 0.73 - - - - - -

DFPHarm-Drink 13.90 0.42 20.33 0.20 - - 20.33 0.20

DFPS-AI-Drink -12.28 0.44 - - - - - -

Note: AI = Avian Influenza, D = Dummy, FPS = farmers’ perceived susceptibility to, FP = Farmers’ perceptions of, a

Independent Variables described in Table 4.6; bFull = full simultaneous regression model; cStep = Stepwise selection

regression model; dForw = forward selection regression model; eBacw = backward selection regression model; (*) significant at

p<0.1; (**) significant at p<0.01

of 340

Table F.7b. Summary of the Best MLR Model Fittings for E. coli cfu in Stored Rainwater for Drinking by Selection

Procedures in the Provinces of Thai Binh and An Giang, Vietnam (cont’d)

Independent variables a Fullb Stepc Forwd Bacwe


DFPS-Diarrhea-Drink -38.00* 0.06 -23.75 0.18 - - -23.75 0.18

DFPS-Coliform-Drink -14.26 0.38 -19.00 0.20 -25.12* 0.06 -19.00 0.20

DFPS-Parasite-Drink 6.16 0.74 - - - - -

DSatisfy-Drink -49.65 0.00 -47.07** 0.00 -44.88** 0.00 -47.07** 0.00

constant 74.16 0.15 28.34 0.43 0.30 0.99 28.34 0.43

Obs 135 - 135 - 135 - 135 -

Prob > F 0.04* - 0.00** - 0.00** - 0.00** -

R-squared 0.20 - 0.24 - 0.21 - 0.24 -

Adj R-squared 0.09 - 0.17 - 0.17 - 0.17 -

Root MSE 70.31 - 66.82 - 66.96 - 66.82 -

Number of significant

variables 2 - 3 4 3 Notes: D = Dummy, FPS = farmers’ perceived susceptibility to, a Independent Variables described in Table 4.6; bFull = full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection regression model; (*) significant at

p<0.1; (**) significant at p<0.01

of 340

Table F.8a. Summary of the Best MLR Model Fittings for E. coli cfu in Pipe Water for Drinking by Selection Procedures in

the Provinces of Thai Binh and An Giang Provinces, Vietnam

Independent

variables a

Full b Step c Forw d Bacw e


DGender -5.64 0.19 -5.24* 0.10 - - -5.24* 0.10

Age 0.19 0.32 0.19 0.24 - - 0.19 0.24

Schooling -0.69 0.26 -0.66 0.23 - - -0.66 0.23

Experience 0.28 0.36 - - - - - -

DIncome 6.86* 0.11 5.26 0.15 5.04 0.67 5.26 0.15

DChildren 6.67 0.14 6.16 0.13 - - 6.16 0.13

DOff-farm job 1.03 0.81 - - - - - -

Npoultry/yr 0.27** 0.03 0.29*** 0.01 0.37*** 0.00 0.29 0.01

Nducks/yr 0.08* 0.00 0.08*** 0.00 0.06*** 0.00 0.08*** 0.00

Npigs/yr 0.10 0.53 - - - - - -

DFish prod - - - - - - - -

DFPR-

Livestock-Man -1.28 0.78

-

- - - - -

DFPR-

Livestock-

Woman -2.12 0.66

-

- - - - -

DFPWater

quality -0.52 0.91

-

- - - - - Notes: D = Dummy, FPS = farmers’ perceived responsibility to, FP = Farmers’ perceptions of, N = Number, a Independent Variables described in Table 4.6; bFull = full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward

selection regression model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.8b. Summary of the Best MLR Model Fittings for E. coli cfu in Pipe Water for Drinking by Selection Procedures in

the Provinces of Thai Binh and An Giang, Vietnam (cont’d)

Independent

variables a



DFPHarm-

Drink 6.11 0.21 5.47 0.21 - - 5.47 0.21

DFPS-AI-Drink 7.27* 0.11 7.00 0.06 - - 7.00 0.06

DFPS-Diarrhea-

Drink -0.94 0.86

-

- - - - -

DFPS-Coliform-

Drink 0.77 0.88

-

- - - - -

DFPS-Parasite-

Drink -7.59* 0.10 -6.16* 0.11 - - -6.16* 0.11

DSatisfy-Drink 2.34 0.57 - - - - - -

constant -12.80 0.36 -9.60 0.38 - - -9.60 0.38

Obs 90 - 90 - 90 - 90 -

Prob > F 0.01*** - 0.00*** - 0.00*** - 0.00*** -

R-squared 0.37 - 0.35 - 0.26 - 0.35 -

Adj R-squared 0.20 - 0.27 - 0.23 - 0.27 -

Root MSE 14.16 - 13.56 - 13.84 - 13.56 -

Number of

significant

variable 5 4 2 3 Notes: AI = Avian Influenza, D = Dummy, FPS = farmers’ perceived susceptibility to, FP = Farmers’ perceptions of, a Independent Variables described in

Table 4.6; bFull = full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw =

backward selection regression model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.9. Summary of the Best MLR Model Fittings for E. coli cfu in Well Water for Drinking by Selection Procedures in the


Independent

variables a



DGender 1.90 0.89 -17.25* 0.06 -14.64* 0.11 -17.25* 0.06

Age -0.58 0.30 - - - - - -

Schooling -4.04** 0.04 - - -1.96 0.17 - -

Experience 0.07 0.93 - - - - - -

DIncome 60.98*** 0.00 48.17*** 0.01 37.48** 0.03 48.17*** 0.01

DChildren -0.08 0.99 -9.71 0.24 - - -9.71 0.24

DOff-farm job -14.85 0.15 -10.47 0.18 -9.85 0.22 -10.47 0.18

Npoultry/yr - - - - - - -

Npigs/yr -0.22* 0.09 - - - - - -

DFish prod - -76.53*** 0.00 -70.33*** 0.00 -76.53*** 0.00

DFPR-Livestock-Man -0.12 0.99 - - - - - -

DFPR-Livestock-

Woman 26.33* 0.07 18.13* 0.09 17.83* 0.10 18.13* 0.09

DFPWater quality -2.91 0.88 18.54 0.16 - - 18.54 0.16 Notes: D = Dummy, FPS = farmers’ perceived responsibility to, N = Number, a Independent Variables described in Table 4.6; bFull = full simultaneous regression

model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection regression model; ere = removed

from the model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.10. Summary of the Best MLR Model Fittings for E. coli cfu in Well Water for Drinking by Selection Procedures in

the Provinces of Thai Binh and An Giang, Vietnam

Independent

variables a



DFPHarm-Drink 11.84 0.40 - - - - - -

DFPS-AI-Drink -9.28 0.43 -13.21 0.14 -12.48* 0.16 -13.21 0.14

DFPS-Diarrhea-

Drink 19.64 0.18 17.13* 0.09 16.19* 0.10 17.13* 0.09

DFPS-Coliform-

Drink -8.66 0.45 - - - - - -

DFPS-Parasite-Drink -26.73** 0.05 -19.40** 0.05 -17.41** 0.07 - -

DAI severe - - - - - -19.40** 0.05

DSatisfy-Drink -1.82 0.87 - - - - - -

constant 100.55** 0.02 110.97*** 0.00 129.68*** 0.00 110.97*** 0.00

Obs 115 - 138 - 138 - 138 -

Prob > F 0.00*** - 0.00*** - 0.00*** - 0.00*** -

R-squared 0.34 - 0.40 - 0.39 - 0.40 -

Adj R-squared 0.23 - 0.35 - 0.35 - 0.35 -

Root MSE 49.19 - 44.18 - 44.24 - 44.18 -


variables 5 - 6 - 7 - 7 - Notes: AI = Avian Influenza, D = Dummy, FPS = farmers’ perceived susceptibility to, a Independent Variables described in Table 4.6; bFull = full simultaneous

regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection regression model; ere =

removed from the model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.11a. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for Drinking by Selection Procedures in

the Provinces of Thai Binh and An Giang Provinces, Vietnam

Independent

variables a



DGender

No sig

nifican

t variab

le found in

this fu

ll model fittin

g

pro

cedure

14.12** 0.05 14.12** 0.05 14.12** 0.0

Age - - - - - -

Schooling -1.61* 0.07 -1.61* 0.07 -1.61 0.07

Experience - - - - - -

DIncome 8.55 0.16 8.55 0.16 8.55 0.16

DChildren - - - - - -

DOff-farm job - - - - - -

Npoultry/yr 0.75*** 0.01 0.75* 0.01 0.75*** 0.01

Nducks/yr -0.02 0.15 -0.02 0.15 -0.02 0.15

Npigs/yr - - - - - -

DFish prod - - - - - -

DFPR-Livestock-Man - - - - - -

DFPR-Livestock-

Woman - - - - - -

DFPWater quality - - - - - - Notes: D = Dummy, FPS = farmers’ perceived main responsibility, N = Number, a Independent Variables described in Table 4.6; bFull = full simultaneous

regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection regression model; ere

= removed from the model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.11b. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for Drinking by Selection Procedures in

the Provinces of Thai Binh and An Giang, Vietnam (cont’d)

Independent

variables a



DFPHarm-Drink

No sig

nifican

t variab

le found in

this fu

ll model

fitting p

roced

ure

- - - - - -

DFPS-AI-Drink - - - - - -

DFPS-Diarrhea-Drink - - - - - -

DFPS-Coliform-Drink - - - - - -

DFPS-Parasite-Drink - - - - - -

DSatisfy-Drink - - - - - -

constant 1.79 0.82 1.79 0.82 1.79 0.82

Obs 111 111 - 111 -

Prob > F 0.03** - 0.03** - 0.03** -

R-squared 0.11 - 0.11 - 0.11 -

Adj R-squared 0.07 - 0.07 - 0.07 -

Root MSE 30.15 - 30.15 - 30.15 -


variables 0 3 3 2 Notes: D = Dummy, FPS = farmers’ perceived susceptibility to, FP = Farmers’ perceptions of, a Independent Variables described in Table 4.6; bFull = full

simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection

regression model; ere = removed from the model;(*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.12a. Summary of the Best MLR Model Fittings for E. coli cfu in River Water for Domestic Purposes by Selection


Independent

variables a



DGender -488.94* 0.07 -338.80* 0.08 -328.88* 0.09 -338.80* 0.08

Age 9.58 0.23 11.06* 0.11 10.86 0.12 11.06* 0.11

Schooling 7.65 0.76 - - - - - -

Experience 9.13 0.39 - - - - - -

DIncome -181.19 0.33 -264.47* 0.10 -266.40 0.12 -264.47* 0.10

DChildren -17.04 0.94 - - - - - -

DOff-farm job 267.70* 0.10 264.33* 0.08 265.29* 0.08 264.33* 0.08

Npoultry/yr - - - - - - - -

Nducks/yr - - - - - - - -

Npigs/yr 38.07*** 0.01 36.97*** 0.01 37.54*** 0.01 36.97*** 0.01

DFish prod -31.03 0.87 - - - - - -

DFPR-Livestock-

Man -404.43*** 0.03 -368.40** 0.03 -365.00** 0.03 -368.40** 0.03

DFPR-Livestock-

Woman -263.49 0.39 - - - - - - Notes: D = Dummy, FPS = farmers’ perceived responsibility to, FP = Farmers’ perceptions of, N = Number, \a Independent Variables described in Table

4.6; bFull = full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw =

backward selection regression model; *) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.12b. Summary of the Best MLR Model Fittings for E. coli cfu in River Water Used for Domestic Purposes by Selection


Independent

variables a




DFPHarm-Dom -233.61 0.26 -278.42* 0.10 -233.06 0.24 -278.42* 0.10

DFPS-AI-Dom 312.53 0.13 279.94 0.16 295.27 0.14 279.94 0.16

DFPS-Diarrhea-Dom -95.02 0.69 - - -102.39 0.65 - -

DFPS-Coliform-

Dom 101.43 0.78 - - - - - -

DFPS-Parasite-Dom -504.85** 0.02 -482.03*** 0.01 -447.77** 0.03 -482.03*** 0.01

DSatisfy-Dom 186.27 0.36 - - - - - -

constant 847.70 0.12 824.62** 0.03 832.28** 0.02 824.62** 0.03

Obs 189 - 188 - 188 - 188 -

Prob > F 0.01*** - 0.00*** - 0.00*** - 0.00*** -

R-squared 0.18 - 0.16 - 0.17 - 0.16 -

Adj R-squared 0.09 - 0.12 - 0.12 - 0.12 -

Root MSE 1008.80 - 990.43 - 992.64 - 990.43 -

Number of

significant variables 5 - 9 - 6 - 8 - Notes: AI = Avian Influenza, D = Dummy, Dom = Domestic, FPS = farmers’ perceived susceptibility to, FP = Farmers’ perceptions of, a Independent

Variables described in Table 4.6; bFull = full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection

regression model; eBacw = backward selection regression model; *) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.13a. Summary of the Best MLR Model Fittings for E. coli in Pond Water for Domestic Purposes by Selection





DGender 14.97 0.98 - - - - - -

Age -7.28 0.73 - - - - - -

Schooling 118.27 0.13 213.66** 0.02 213.66** 0.02 213.66** 0.02

Experience -41.08 0.14 - - - - - -

DIncome

-

1,139.09* 0.11 - - - - - -

DChildren 268.30 0.55 - - - - - -

DOff-farm job 719.17* 0.08 - - - - - -

Npoultry/yr -7.18 0.18 -17.79* 0.10 -17.79* 0.10 -17.79* 0.10

Nducks/yr - -1.03 0.20 14.78* 0.11 -1.03 0.20

Npigs/yr - 14.78* 0.11 -1.03 0.20 14.78* 0.11

DFish prod -266.54 0.68 - - - - - -

DFPR-Livestock-Man -56.70 0.90 - - - - - -

DFPR-Livestock-

Woman -383.79 0.52 - - - - - -


DFPHarm-Dom 352.23 0.53 - - - - - - Notes: D = Dummy, FPR = farmers’ perceived responsibility to, FP = Farmers’ perceptions of, a Independent Variables described in Table 4.6; bFull = full

simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection regression

model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

Table F.13b. Summary of the Best MLR Model Fittings for E. coli in Pond Water for Domestic Purposes by Selection


Independent

variables a



DFPS-AI-Dom 211.36 0.68 1,044.65** 0.05 1,044.65** 0.05 1,044.65** 0.05

DFPS-Diarrhea-

Dom 206.70 0.72 - - - - - -

DFPS-Coliform-

Dom 544.62 0.29 - - - - - -

DFPS-Parasite-

Dom -488.10 0.39 - - - - - -

DSatisfy-Dom 619.99 0.17 1,028.56* 0.09 1,028.56* 0.09 1,028.56* 0.09

constant 2,427.87 0.12 210.21 0.81 210.21 0.81 210.21 0.81

Obs 299 - 197 - 197 - 197 -

Prob > F 0.11 - 0.02 - 0.02 - 0.02 -

R-squared 0.08 - 0.08 - 0.08 - 0.08 -

Adj R-squared 0.03 - 0.05 - 0.05 - 0.05 -

Root MSE 3,285.70 - 3,433.80 - 3,433.80 - 3,433.80 -

Number of

significant

variables 2 - 5 - 5 - 5 - Notes: AI = Avian Influenza, D = Dummy, Dom = Domestic, FPS = Farmers’ perceived susceptibility to, a Independent Variables described in Table 4.6; bFull =

full simultaneous regression model; cStep = Stepwise selection regression model; dForw = forward selection regression model; eBacw = backward selection

regression model; (*) significant at p<0.1; (**) significant at p<0.01

of 340

F.6. Detailed Results of SUR Model for E. coli in Sources of Water On Farms

The best fit of the SUR model for estimating two equations of E. coli cfu in sources of

drinking water and sources of domestic water is presented in Tables F.14-16. Overall, the two

equations for E. coli cfu in drinking and domestic water were significant at p < 0.1. The SUR

model fitting for E. coli cfu in drinking water has four associated factors of the levels of E. coli

cfu in drinking water among which one was positive significant predictor. The SUR model

fitting for E. coli cfu in domestic water has five significant associated factors. Variable

Nducks/yr on farm showed up as a factor in both equations with different coefficients and p-

values. Table F.15 shows the correlation matrix on the residuals of the two equations for E. coli

cfu in drinking and domestic water. The Breusch-Pagan test of independence shows that the

residuals of the two equations were significantly related (p < 0.05). Therefore, the use of SUR is

justified.


Drinking


(… results of iteration 2-5 omitted)

Iteration 6: tolerance = 3.22E-08

Seemingly unrelated regression, iterated

Equation Obs Parms RMSE R-sq chi2 P

NE.coli Drink 243 4 14078.81 0.03 8.26 0.08

NE.coli Dom 243 26 2761.06 0.17 50.45 0.00


Drinking (cont'd)

Correlation matrix of residuals:

NE.coli Drink NE.coli Dom

NE.coli Drink 1

NE.coli Dom 0.14 1

Breusch-Pagan test of independence: chi2(1) = 4.60, Pr = 0.03

of 340

Table F.16. SUR Model Estimation of the Two Equations of E. coli cfu in On-Farm Water

Coef. Std. Err. z P>z [95% Conf. Interval]

NE.coli Drink

DOff-farm job 3163.55** 1873.55 1.69 0.09 -508.54 6835.64

NDucks/yr 5.51 4.48 1.23 0.22 -3.26 14.28

DSatisfy-Drink -1706.41 1856.55 -0.92 0.36 -5345.19 1932.37

DFPHarm-Drink -3422.72 2243.79 -1.53 0.13 -7820.47 975.04

_cons 1870.37 2292.16 0.82 0.42 -2622.18 6362.92

NE.coli Dom

Age 0.11 19.03 0.01 1.00 -37.19 37.41

DGender 160.58 494.00 0.33 0.75 -807.66 1128.81

Schooling 195.04* 70.12 2.78 0.01 57.62 332.46

DChildren -354.00 457.98 -0.77 0.44 -1261.63 543.64

Experience -45.37 29.34 -1.55 0.12 -102.88 12.14

DFPR-Livestock-Woman -941.67 737.85 -1.28 0.20 -2387.82 504.49

DFPR-Livestock-Man -987.56** 543.15 -1.82 0.07 -2052.13 77.00

DFPR-Fish-Woman -619.64 930.68 -0.67 0.51 -2443.75 1204.46

DFPR-Fish-Man 301.12 596.29 0.50 0.61 -867.59 1469.83

DFPR-Health-Man 275.55 646.56 0.43 0.67 -991.68 1542.78

DProvince 43.29 697.51 0.06 0.95 -1323.81 1410.39

DFish prod 1217.73* 611.38 1.99 0.05 19.45 2416.00

DRain for animals 204.62 548.58 0.37 0.71 -870.59 1279.82

DPond for animals 21.48 451.33 0.05 0.96 -863.11 906.07

DPipe for animals -365.64 687.70 -0.53 0.60 -1713.51 982.23

DMulti-agri activities -523.14 640.61 -0.82 0.41 -1778.72 732.44

DPoultry AI 2281.77* 848.94 2.69 0.01 617.88 3945.66

NDucks/yr -0.01 0.99 -0.01 0.99 -1.95 1.93

NPoultry/yr -5.40 7.98 -0.68 0.50 -21.05 10.26

DFPWater quality -720.46 523.83 -1.38 0.17 -1747.14 306.22

DFPBarr-Cost 230.33 480.15 0.48 0.63 -710.75 1171.40

DFPRisk rain water -1266.58 1153.19 -1.09 0.28 -3516.79 1003.62

DFPS-Parasite-Dom 68.45 586.73 0.12 0.91 -1081.53 1218.42

DAI heard -2682.34** 1693.55 -1.58 0.10 -6001.65 636.97

DFPHarm-Dom 508.12 493.76 1.03 0.30 -459.63 1475.87

DFPE-Livestock

Management -361.33 548.19 -0.66 0.51 -1435.76 713.10

_cons 5472.15 2344.87 2.33 0.02 876.28 10068.02

Notes: AI = Avian Influenza, D = Dummy, Dom = Domestic, FPS = Farmers’ perceived susceptibility to, FP =

Farmers’ perceptions of, * significant at p < 0.05; ** significant at p < 0.1

of 340

The variable of number of ducks (Nducks/yr) was presented in the two individual

equations for E. coli cfu in drinking water and domestic water respectively (Table F.16).

Therefore, a test of constraint for the Nducks/yr variable was conducted. The test of cross-

equation constraint showed the Nducks/yr variable is not significantly different from the two

individual equations at p = 0.22 .The SUR model was run again with the cross-equation

constraint of the Nducks/yr variable.

of 340

APPENDIX G: CHAPTER FIVE’S APPENDICES

G.1. Summary Statistics of Variables Used in the Probit Models

Table G.1. Summary Statistics of Demographic Variables Used in Probit Models of Factors

Associated with Small-Scale Integrated (SSI) Farmers’ Engagement in Mitigating

Strategies to Reduce WRZD Transmission

Variable Description Mean s.d. Count (=1) % (=1)

DProvince

Two provinces in the research (1 = Thai

Binh; 0 = An Giang) - - 300 50.2

Age Age of the participating SSI farmers 45.9 11.3 - -

Schooling Years of attending school 6.9 3.2 - -

Experience Years of farming 9.6 8.3 - -

DChildren

Having household members under 18 years

old (1= Yes; 0 = No) - - 421 70.4

Mainly responsible by for:

DFPR-Health-

Man family health by man (1 = Yes; 0 = No) - - 131 22.2

DFPR-Health-

Woman family health by women (1 = Yes; 0 = No) - - 104 17.6

DFPR-Livestock-

Man

Mainly responsible for livestock production

by gender (1 = Yes; 0 = No) - - 244 41.4

DFPR-Livestock-

Woman

Mainly responsible for livestock production

by gender (1 = Yes; 0 = No) - - 91 15.5

DIncome

On-farm DIncome per person per year

below and above poverty line (1= Yes; 0 =

No) - - 118 20.0

Number of livestock by type/year

Npoultry/yr chickens 24.3 21.2 - -

Nducks/yr ducks 36.2 47.9 - -

Npigs/yr pigs 18.9 18.2 - -

Ncattle/yr cattle 3.1 2.1 - -

Number of E. coli cfu in:

NE.coli Drink on-farm drinking water 12.8 21.0 - -

NE.coli Dom on-farm domestic water 1,506.3 1,994.6 - -

LogE.coli Drink on-farm drinking water (log transformed) 1.9 2.0 - -

LogE.coli Dom on-farm domestic water (log transformed) 6.2 2.1 - -

Presence/Absence of E. coli cfu in:

DE.coli Domestic

on-farm domestic water (1 = Presence; 0 =

Absence) - - 574 96.0

DE.coli Drink

on-farm drinking water (1 = Presence; 0 =

Absence) - - 314 58.7

Turbidity levels in:

Turbidity drinking on-farm drinking water 1.1 0.9 - -

Turbidity

domestic on-farm domestic water 43.6 28.6 - -

Notes: D = Dummy, N = Number, FPR = Farmers’ perceived main responsibility

of 340

G.2. Summary Statistics of SSI Farmers’ Perceptions Related Variables Used in the Probit Models

Table G.2. Summary Statistics of SSI Farmers’ Perceptions Related Variables Used in Probit Models to Explore Factors

Associated with SSI Farmers’ Engagement in Mitigating Strategies to Reduce WRZD Transmission

Variable name Description – farmers’ perceptions of: Count (=”1”) % (=”1”)

DFPS-AI-Drinking

susceptibility to avian influenza of untreated drinking water (1 = highly susceptible, 0 = low

susceptible) 313 52.3

DFPS-AI-Dom

susceptibility to avian influenza of untreated domestic water (1 = highly susceptible, 0 = low

susceptible) 289 48.3

DFPS-AI-Wastewater susceptibility to avian influenza of untreated wastewater (1 = highly susceptible, 0 = low susceptible) 435 88.0

DFPS-AI-Pond susceptibility to avian influenza of pond water (1 = highly susceptible, 0 = low susceptible) 401 83.7

DAI severe severity of avian influenza if infected water (1 = very severe, 0 = not very serve) 481 80.4

DFPSafe-Drink safety of untreated water for drinking (1 = Yes, it is safe; 0 = No, it is not safe) 54 9.0

DFPBarr-Cost cost is a barrier to taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 239 40.0

DFPBarr-Not know “do not know” as a barrier to taking actions (1 = Yes; 0 = No) 232 38.9

DFPBarr-Not understand “do not understanding” as a barrier to taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 121 20.2

DFPBarr-Peer peer pressure as a barrier to taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 28 4.7

DFPBarr-Busy busy as a barrier to taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 151 25.3

DFPBen-Action benefit of taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 452 77.9

DFPBen-No Action benefit of not taking actions to reduce WRZD transmission (1 = Yes; 0 = No) 132 24.1

DFPE-Livestock Management self-capacity in managing livestock (1 = Strong; 0 = Weak) 263 44.4

DFPE-Point of Use self-capacity in using point of use methods (e.g., washing hand with soap) (1 = Strong; 0 = Weak) 250 44.9

DFPE-Water Protection self-capacity in protecting source water (1 = Strong; 0 = Weak) 255 42.8

DFPE-Water Distribution self-capacity in water storage, treatment, distribution (1 = Strong; 0 = Weak) 239 40.4

DFPT-Health Advice advice from health workers as a trigger of actions to reduce WRZD transmission (1 = Yes; 0 = No) 295 49.3

DFPT-Loss Income loss DIncome as a trigger of actions to reduce WRZD transmission (1 = Yes; 0 = No) 165 27.6

DFPT-Media media as a trigger of actions to reduce WRZD transmission (1 = Yes; 0 = No) 405 67.7

DFPT-Peer peer pressure as a trigger of actions to reduce WRZD transmission (1 = Yes; 0 = No) 101 16.9

DFPT-Worry worry as a trigger of actions to reduce WRZD transmission (1 = Yes; 0 = No) 374 62.5

Notes: D = Dummy, FPS = Farmers’perceived susceptibility, AI = Avian Influenza, FPSafe = Farmers’ perceived safety of, FPBarr = Farmers’ perceived

barriers of, FPBen = Farmers’ perceived benefits of, FPE = Farmers’ perceived self-efficacy, FPT = Farmers’ perceived triggers of

of 340

References

Bartram, J. and Pedley, S. (1996). Microbial Analysis. In J.Bartram and R. Ballance (Eds.),

Water Quality Monitoring - A Practical Guide to the Design and Implementation of

Freshwater Quality Studies and Monitoring Programmes ( World Health Organization.

Bluewater Biosciences (2013). Coliplate (TM): A quantitative 24-hour test for Coliforms and E.

coli. Bluewater Biosciences [On-line]. Available:

http://bluewaterbiosciences.com/index.php?main_page=pageandid=16

Dohoo, I., Martin, W., and Stryhn, H. (2003). Veterinary Epidemiologic Research.

Edberg, S. C., Rice, E. W., Karlin, R. J., and Allen, M. J. (2000). Escherichia coli: the best

biological drinking water indicator for public health protection. Journal of Applied

Microbilogy, 88, 106S-116S.

EPA (1993). Method 180.1: Determination of Turbidity by Nephelometry. EPA [On-line].

Available:

http://water.epa.gov/scitech/methods/cwa/bioindicators/upload/2007_07_10_methods_me

thod_180_1.pdf

FPTCW (2013). Guideline Technical Document: Escherichia coli. Ottawa, Ontario: Federal-

Provincial-Territorial Committee on Drinking Water of the Federal-Provincial-Territorial

Committee on Health and the Environment, Health Canada.

Fricker, C. R., DeSarno, M., Warden, P. S., and Eldred, B. J. (2008). False negative ß-D-

glucuronidase reactions in membrane lactose glucuronide agar medium used for the

simultaneous detection of coliforms and Escherichia coli from water. Letters in Applied


of 340

Greene, W. H. (2008a). System of Regression Equations. In Econometric Analysis (6th Edition

ed., pp. 252-280). Pearson Prentice Hall.

Greene, W. H. (2008b). The Classical Multiple Linear Regression Model. In Econometric

Analysis (6th Edition ed., pp. 8-17). Pearson Prentice Hall.



Hall, T. A. (1999). BioEdit: a user-friendly biological sequence alignment editor and

analysisprogram for Windows 95/98/NT. Nucleic Acids Symposium Series, 95-98.

IDRC (2012). Ecohealth Research in Practice. (Insight and Innovation in International

Development ed.) (vols. Innovative Applications of an Ecosystem Approache to Health)

International Development Centre (IDRC); Springer.

IHPH (2013a). Technical Guidelines: Procedure for Determination of pH in Water. Institute of

Hygiene and Public Health Ho Chi Minh City.

IHPH (2013b). Technical Guidelines: Procedure for Determination of Turbidity in Water.

Institute of Hygiene and Public Health Ho Chi Minh City.




Knudsen, L. G., Pham, D. P., Nguyen, H. T., Helle, S., Jensen, P. K., Dalsgaard, A. et al. (2008).

The fear of awful smell: risk perceptions among farmers in Vietnam using wastewater

and human excreta in agriculture. The Southeast Asian journal of tropical medicine and

public health, 39, 341-352.

of 340









MoH and MARD (2011). The Vietnam Integrated National Operational Program on Avian

Inlfuenza, Pandemic Preparedness and Emerging Infectious Diseases (APEID) 2010-

2015: Strengthening responses and improving prevention through a One Health

approach. Hanoi: Vietnam Ministry of Agriculture and Rural Development and Ministry

of Health.

Payment, P. and Pintar, K. (2006). Waterborne Pathogens: A Critical Assessment of Methods,

Results and Data Analysis. Journal of Water Science, 19, 233-245.




Powell, C. (2007). The Perception of Risk and Risk Taking Behavior: Implications for Incident

Prevention Strategies. Wilderness and environmental medicine [On-line]. Available:

http://linkinghub.elsevier.com/retrieve/pii/S1080603207701965?showall=true

TCVN 6187 (2009). Water quality: Detection and enumeration of coliform organisms

thermotolerant coliform organisms and presumptive Escherichia coli. Part 2: Membrane

of 340

filtration method. Directorate for Standards, Metrology and Quality, Ministry of Science

and Technology of the Socialist Republic of Vietnam [On-line]. Available:

http://www.tcvn.vn/sites/head/en/tim-kiem-tieu-chuan.aspx?sk=andcat=tcvn

TCVN 6492 (2011). Water quality. Determination of pH. Directorate for Standards, Metrology

and Quality, Ministry of Science and Technology of the Socialist Republic of Vietnam

[On-line]. Available: http://www.tcvn.vn/sites/head/en/tim-kiem-tieu-

chuan.aspx?sk=andcat=tcvn

TCVN 6663 (2011). TCVN 6663 2011 - ISO 5667-1:2006: Water quality - Sampling - Part 1:

Guidance on the design of sampling programmes and sampling techniques. Directorate

for Standards, Metrology and Quality, Ministry of Science and Technology of the

Socialist Republic of Vietnam [On-line]. Available:

http://www.tcvn.vn/sites/head/en/tim-kiem-tieu-chuan.aspx?sk=andcat=tcvn




Ward, H., Toledano, M. B., Shadick, G., Davies, B., and Elliott, P. (2012). Oxford Handbook of

Epidemiology for Clinicians. (1 ed.) Oxford University Press.

WHO (2011a). Guidelines for drinking-water quality. (Fourth ed.) World Health Organisation.

WHO (2011b). Microbial aspects. In Guidelines for drinking-water quality (pp. 117-153). World




of 340

WHO and OECD (2003a). Assessing Microbial Safety of Drinking Water: Improving


WHO and OECD (2003b). Assessing Microbial Safety of Drinking Water: Improving


Winfield, M. D. and Groisman, E. A. (2003). Role of nonhost environments in the lifestyles of

Salmonella and Escherichia coli. Applied and Environmental Microbiology, 67, 3678-

3694.

perceptions of risk factors and mitigating strategies for

Documents