trends in paediatric injury rates using emergency department based injury surveillance

262 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2010 vol. 34 no. 3© 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

Trends in paediatric injury rates using emergency

department based injury surveillance

Michael ShepherdChildren’s Emergency Department, Auckland District Health Board and Department of Paedeatrics, University of Auckland, New Zealand

Abstract

Objective: The primary aim of this study

was to develop a method of calculating

paediatric injury rate from Emergency

Department injury surveillance data and

use this to describe trends in paediatric

injury. This study also aimed to establish

whether triage category could be used as

an indicator of severity.

Methods: Prospective observational study

of paediatric injury in Brisbane, Australia

from 1998 to 2005 using Emergency

Department injury surveillance data. Injury

incidence was calculated using postcode

restriction, census data and analysis

of injury surveillance data quality and

alternative hospital presentations.

Results: The incidence of Emergency

Department injury presentation

increased by 56% between 1998

and 2005. The incidence of injury

(adjusted for ascertainment and other

hospital presentations) increased from

5,323/100,000/year to 8,316/100,000/

year (p<0.01). The overall incidence of

admission increased from 1,066/100,000/

year to 1,238/100,000/year (p<0.01). The

incidence of injury presentations triaged

as urgent or above increased by 16% over

the study period (2,348/100,000/year to

2,723/100,000/year, p<0.01).

Conclusion: Injury incidence can be

determined using Emergency Department

injury surveillance data and triage category

is a useful indicator of injury severity.

Paediatric emergency department injury

presentations, including serious injury,

increased significantly between 1998 and

2005.

Implications: The methodology used in

this study is easily repeatable and could

be used to evaluate injury prevention

interventions. The prevention and

management of injury should be directed

by accurate injury incidence data.

Key words: wounds and injuries, triage,

incidence, emergencies.

Aust NZ J Public Health. 2010; 34:262-8

doi: 10.1111/j.1753-6405.2010.00524.x

Submitted: June 2009 Revision requested: November 2009 Accepted: December 2009Correspondence to:Dr Michael Shepherd, Children’s Emergency Department, Starship Hospital, Auckland, New Zealand. Fax: +64 9 3757055; e-mail: [email protected]

It has been estimated that at least one-

third of serious unintentional injury is

preventable using application of existing

injury prevention methods.1 Extensive

resources have been applied to the prevention

of paediatric injury with involvement of

a large and diverse range of stakeholders.

Few of these initiatives have been formally

evaluated or published in peer review

literature. A critical step toward developing

an evidence-based approach to injury

prevention is to have a robust method of

measuring injury.

There are a number of possible ways of

measuring injury burden, all of which have

been identified as having limitations.2 These

methods include:

• Mortality review – where potential

limitations include inaccurate diagnostic

or injury data, absence of detail regarding

injury mechanism, difficulty or delay in

accessing information and low incidence

(particularly in paediatrics). In addition,

conditions that cause significant morbidity

without mortality will not be captured

(e.g. caustic ingestion, burns and scalds,

dog bites).2-6

• Hospital admission data – where potential

limitations include low data quality, poor

data availability and variable admission

policies.3-7 A trend in paediatrics to

manage more injury presentations

without admission (e.g. the use of

procedural sedation and increased use of

more advanced radiology allowing early

discharge of children with moderate head

injuries) will result in failure to capture

significant morbidity.

• Trauma registry – where potential

limitations include the limitations of

admissions review outlined above, labour

intensive data collection and retrospective

data collection (potentially affecting data

accuracy).4,5

• Population surveys – which deliver

accurate injury incidence data but require

signif icant commitment of time and

money.5

An alternative approach to measuring

injury is Emergency Department (ED) based

injury surveillance.5 These data are typically

collected prospectively, at the beginning of

a hospital presentation. Data are available

rapidly and information supplied by patients

or their caregivers is not affected by other

interactions that occur during the provision

of healthcare. Potential limitations of such

a database have been identified including

labour intensive data collection, variable

utilisation of EDs by the public and injuries

included are predominantly ‘minor’.2,8

However, paediatric injury may be more

frequently managed in hospital EDs because

of requirements for observation, sedation for

procedures and more specialised imaging

and the influence of social pressures. The

inclusion of ‘minor’ injuries may allow

researchers to develop greater understanding

of the causes of more severe injuries and

allow researchers to identify factors that

influence injury severity in various injury

types. ‘Minor’ injuries may not actually be

minor with significant burden on the health

system, significant morbidity (e.g. burns,

fractures and head injuries) and social

impact.

Parents and children Article

2010 vol. 34 no. 3 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 263© 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

ED based injury surveillance often samples a proportion of

patients or a selection of EDs within a community, therefore

accurate population incidence data is not able to be calculated. The

primary aim of this study was to develop a method of calculating

paediatric injury rate from ED injury surveillance data and use

this to describe trends in paediatric injury.

The secondary aim was to establish whether triage category

could be used as a valid indicator of severity. Triage is a brief

clinical assessment that determines the urgency of treatment and

the time and sequence in which patients should be seen in the

ED.9 Triage category is assigned on arrival to an ED by a trained

member of the nursing team.10 The triage category assigned should

reflect the clinical urgency.11 The Australasian Triage Scale is

used in Brisbane with five available categories, 1 being the most

urgent.11

Although triage performance has not been investigated in the

two study hospitals, other investigators have found that within an

institution the triage category provides a stable measure of the

acuity of presentation.12 In addition, triage of paediatric patients

within different paediatric emergency departments has been found

to be reliable.13 This study aims to use presentations of triage

category 1 to 3 as a proxy measure of severity. Triage category 1

to 3 patients have injuries that are life threatening or potentially

life threatening.11

MethodsA prospective observational study of children living in

metropolitan Brisbane was completed. Brisbane is the capital

city of the state of Queensland on the east coast of Australia. The

population of Brisbane City at 30 June 2005 was estimated to be

971,757, increasing at approximately 1.5% per year.14

Population incidence data was generated by:

a. Defining a geographical area (based on postcodes) that

represents metropolitan Brisbane.

b. Determining the population of this area using Australian census

data.

c. Assessing the quality of the Queensland Injury Surveillance

Unit (QISU) data over the study period

d. Determining the frequency of injury presentation to a non-

QISU ED by paediatric patients residing within the previously

defined geographical area.

A geographical selection of postcodes of metropolitan Brisbane

was made that centred on the two paediatric EDs – Royal

Children’s Hospital (RCH) and Mater Children’s Hospital (MCH).

RCH and MCH are tertiary children’s hospitals with dedicated

paediatric EDs with a combined annual census of around 70,000

presentations.15,16 The population of this geographical area was

derived from 2001 and 2006 census data from the Australian

Bureau of Statistics.17,18 Age-related postcode population and

gender data were combined to calculate the population of the area

selected. The population of the geographical area was assumed to

have changed linearly between censuses.

The primary dataset was derived from the Queensland Injury

Surveillance Unit (QISU) database. QISU collects injury data from

hospital EDs on behalf of Queensland Health with the support of

the Mater Health Services Brisbane.19 Injury surveillance data was

collected using the National Data Standards for Injury Surveillance

(NDS-IS) version 2c (January 1998)20 and a QISU modification

(from July 2003). The coding of the variables used in this study

did not differ between these two versions.

Injury case definition is “first presentation of trauma, poisoning,

and other conditions of rapid onset to which factors and

circumstances external to the person contributed significantly”.20

Only the first attendance to any data collection site is included.

Electronic data is collected prospectively by nursing and clerical

staff at triage (the point of entry to the ED).

Data qualitySome measures of injury surveillance data quality have been

described:

• Content validity – whether injury information is coded

correctly.21

• Ascertainment – whether all new injuries are recorded (also

called sensitivity by some authors22).

• Bias – whether certain injury types are more likely not to be

recorded or coded correctly (also called representativeness by

some authors22).

Electronic data collection has been shown to provide good

quality data with the potential for high ascertainment.21,23,24 The

QISU electronic collection method has been validated and has

demonstrated high ascertainment rates.24 The ascertainment

profile for the two EDs is measured routinely by QISU using a

comparison with mandatory electronic discharge diagnosis coding

data from Queensland Health database (which is required to be

submitted for funding).

Bias was assessed by completing further analysis of the profile

of cases not ascertained – including age and type of injury.

Inclusion criteria were; a completed injury surveillance record,

age less than 14 years (this age range was used as the hospitals had

differing policies regarding 14 year olds presenting with injury),

usual residence within the previously defined geographical area,

injured patient presented to hospital during the eight-year period

from 1 January 1998 to 31 December 2005.

Other hospitals in the region supplied paediatric injury data

including arrival date, date of birth, diagnosis or ICD code

(S00-T98) and postcode. These data were available for all of the

Queensland Health hospitals in greater Brisbane (five hospitals) for

2005 and for two of the five private hospitals in greater Brisbane

for 2005. These data were then restricted according to the inclusion

criteria detailed above.

The data were divided into four chronological groups, each

spanning a two-year period in order to measure the change

in incidence over the eight year period, while ensuring a case

frequency that would allow for meaningful analysis.

To calculate estimated incidence of hospital presentation with

injury, the population denominator was estimated using 2001 and

2006 census data and the numerator was adjusted for ascertainment

Parents and children Trends in paediatric injury rates


and for other hospital presentations (both by increasing the

numerator accordingly).

Statistical methodsQISU data were manipulated using Microsoft Office Excel

2003 (Microsoft Corporation, Seattle, Washington, US). Epi

Info Version 3.4 (Centers for Disease Control and Prevention,

Atlanta, Georgia, US) was used to generate summary tables and

statistics.

The comparison of two incidence rates was carried out using

Pearson’s Χ2 test. The comparison of injury incidence rates over

the eight year time period (grouped into four two-year bins)

was carried out using the Cochran-Armitage test for trend.25

Pearson’s Χ2 test is reported as ‘p’ (p<0.05 considered statistically

significant), Χ2 for trend is reported as “ptrend

” (ptrend

<0.05 considered

statistically significant) and Χ2 for departure is reported as “pdepart

”

(pdepart

>0.05 considered to suggest a linear trend). Statistical tests

were conducted using EpiBasic version1.0 (University of Aarhus,

Nordre Ringgade, Denmark).

EthicsSpecific ethics approval has not been sought for this study.

QISU has standing ethical approval to collate, analyse and publish

de-idenitifed injury surveillance data gathered in Queensland

hospitals.

ResultsThe total population of metropolitan Brisbane (0 to 14 years)

increased by 7,410 (7.4%) from 2001 to 2006. The greatest

increase was in the 0 to 4 age group, which increased by 12.4%

over this period. The gender ratio changed little, with males making

up around 51% of the population across all ages.

The ascertainment profile (sensitivity) for RCH and MCH for

the study period is shown in Table 1. Ascertainment (the number

of injured children with complete injury surveillance information

collected) decreased over the study period. The percentage

of injured children not waiting to be seen by a doctor and the

proportion of children presenting to the ED with an injury related

problem did not change significantly over the study period.

The age profile and ICD diagnosis of the patients with an injury

diagnosis in 2005 in metropolitan Brisbane (Queensland Health

Emergency Department data) was very similar to that of patients

with completed injury surveillance records in 2005 (QISU data)

suggesting little bias. Allergy was under-represented in the QISU

data.

Only 195 injured children residing in the study area attended

public hospital EDs other than RCH or MCH in 2005 and 461

injured children from metropolitan Brisbane presented to the two

private hospitals in 2005 (Table 2). The age distribution of the

injured patients presenting to other hospitals (both private and

other Queensland Health) was skewed toward the older child.

An age group-based conversion factor was calculated (in bold

in Table 2).

Injury burdenTotal injury presentations increased by 24%, 14% and 5%

between each two-year period. The incidence of injury (adjusted

for ascertainment and other hospital presentations) increased from

5,323/100,000/year in 1998/99 to 8,316/100,000/year in 2004/05

(Figure 1). This was an increase of 56% over the study period

(ptrend

<0.01, pdepart

=0.07), with increases of 14%, 19% and 12%

between each two-year period.

Between 57 and 58% of children and young people presenting

with an injury were male, this did not change significantly over

the study period.

Table 1: Ascertainment profile.

1998-1999 2000-2001 2002-2003 2004-2005

Mean ascertainment % 90% 92% 86% 79%

Mean % injury presentations “did not wait” N/A 0.9% 1.4% 1.3%Mean % of ED presentations injury-related N/A 26% 27% 26%

Table 2: Other hospital presentations, children residing in metropolitan Brisbane, 2005.

QISU cases Age Injured, % Injured, % Total injured, % adjusted for group presented to other presented to presented to ascertainment public hospitals private hospitals ‘other’ hospitals

3,593 0 to 4 57 1.59 201 5.59 258 7.182,298 5 to 9 62 2.7 146 6.35 208 9.051,999 10 to 13 77 3.85 114 5.7 191 9.557,890 total 195 2.47 461 5.84 656 8.31

0

2000

4000

6000

8000

10000

12000

14000

16000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 Total

Age (years)

Rat

e (p

er 1

0000

0/ye

ar)

1998 to 1999

2000 to 2001

2002 to 2003

2004 to 2005

Figure 1 – Incidence of injury by age and year

4

Figure 1: Incidence of injury by age and year.

Shepherd Article



AdmissionsThe percentage of all injury presentations that required

admission decreased between 1998 and 1999 (20%) and 2000

and 2001 (15%) with little change over the following four years.

Admission percentage was highest among youngest children.

The overall incidence of admission decreased from 1998/99

(1,066/100,000 /year) to 2000/01 (943/100,000/year) (p<0.01) and

then increased over the following six year period to 1,238/100,000/

year over the study period (ptrend

<0.01, pdepart

<0.01) (Figure 2).

The incidence of admission for the under five year age group

increased significantly between 1998 and 1999 and 2004 and 2005

(1,394 to 1,756/100,000/year) (ptrend

<0.01, pdepart

<0.01) (Figure 2).

Between 1998 and 1999 and 2004 and 2005 there was no change

in the incidence of admission for the five to nine year (916 to

946/100,000/year, ptrend

=0.26) and 10 to 13 year age groups (844

to 886/100,000/year, ptrend

=0.52) (Figure 2).

Triage categoryThe incidence of injuries assigned a triage category of 2 and 3

increased by 44% (ptrend

<0.01, pdepart

<0.01) and 12% (ptrend

<0.01,

pdepart

<0.01) respectively over the study period. The incidence of

injuries assigned a triage category of 4 increased by 97% over the

study period (ptrend

<0.01, pdepart

<0.01).

The proportion of injury presentations triaged as urgent or above

(triage category of ‘1’, ‘2’, and ‘3’) were 44% in 1998/99, 29%

in 2000/01, 33% in 2002/03 and 34% in 2004/05. The incidence

of injury presentations triaged as urgent or above increased by

16% over the study period (Figure 3), from 2,348/100,000/year

to 2,723/100,000/year (ptrend

<0.01, pdepart

<0.01). The incidence of

injury presentation triaged as urgent or above increased across all

age groups (ptrend

<0.01, pdepart

<0.01) (Figure 3). The incidence of

injuries triaged as urgent or above was much higher for 1 and 2

year olds, decreasing in school age children (Figure 3).

Cause of injuryAn analysis of the external cause of injuries that required

admission was performed (Table 3). An analysis of the external

cause of injuries that were triaged as urgent or above were the same

as those that most frequently resulted in admission (Table 3).

Playing was the most frequently recorded activity, involved in

33% to 39% of injuries. The incidence of sporting-related injury

increased from 1,150/100,000/year in 1998/99 to 1,580/100,000/

year in 2004/05 (ptrend

<0.01, pdepart

=0.17). The incidence of

sporting-related injury triaged as urgent or above did not change

significantly over the study period, 495/100,000/year in 1998/99

and 537/100,000/year in 2004/05 (p=0.07).

Injury locationAround one-third of all injuries involved the upper limbs,

between 16 and 19% involved the head, around 15% involved the

lower limbs and between 11 to 13% involved the face. This did

not change over the study period.

Place injury occurredThe majority of injuries occurred in or around the child or

young person’s home (58 to 59% of all injuries). Injuries also

occurred frequently at school or preschool (15 to 16%) and in

parks (12 to 13%).

Table 3: External cause of injuries requiring admission in 2004/05.

% of all admissions

Under 5

Low fall 37.7

Burn 13.4

Collision 11.3

High fall 9.9

Poison 9.2

5 to 9

Low fall 35.9

High fall 21.8

Collision 12.6

Transport-related 9.8

Burn 6.6

10 to 13

Low fall 36.3

Collision 19.7

Transport-related 16.9High fall 8.3

0

500

1000

1500

2000

2500

3000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 Overall

Age (years)

Adm

issi

ons

(per

100

000/

year

)

1998 to 19992000 to 20012002 to 20032004 to 2005

Figure 2 – Incidence of injury admission by age and year

5

Figure 2: Incidence of injury admission by age and year.

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 to 4 5 to 9 10 to13

Overall

Age (years)

Rat

e of

inju

ry (p

er 1

0000

0/ye

ar)

1998 to 1999

2000 to 20012002 to 2003

2004 to 2005

Figure 3– Incidence of injury triaged as urgent or above, by age and year

6

Figure 3: Incidence of injury triaged as urgent or above, by age and year.


Unintentional injuryAnalysis of the nature of injury variable for all injury

presentations showed that all types of injuries increased over the

study period with the exception of poisoning. The intent of injury

was recorded as unintentional in 96.8% of cases in 2004/05. This

changed little over the study period with unintentional injuries

making up 97.7% of all injuries in 1998/99, 96.0% in 2000/01

and 95.2% in 2002/03.

DiscussionA significant increase in the incidence of paediatric ED injury

presentations (56% over eight years) has not previously been

reported in Australia or New Zealand. Victorian data suggests

a similar magnitude of ED presentation increase, although

an incidence rate was not reported.26 The incidence of injury

presentation in 1998/99 (5,323/100,000/year) was similar to

a previous study of paediatric injury in Brisbane in 1985 to

1991 (6,013/100,000/year).27 In Victoria, the overall incidence

of paediatric injury presentation in 1989 was approximately

5,400/100,000/year, which was very similar to the Brisbane

incidence in 1998/99.28 In the US, the incidence of injury

presentation in 2006 was estimated to be 10,600/100,000/year,

somewhat higher than that of Brisbane.29

This increase in injury presentation may result from an increase

in injury incidence, an increase in ED utilisation or a change in

identification of injury presentations. A change in identification

of injury presentations is very unlikely to be the cause as injury

presentations (as a percentage of all presentations) did not increase

(Table 1). An increase in ED utilisation could be caused by reduced

use of other local EDs, reduced use of general practitioners (GPs)

and greater ‘medicalisation’ of less serious injury. Reduced use of

other local EDs seems unlikely to be a major contributor as the

study area was centred on the two public EDs studied. Reduced

use of GPs for paediatric injury is a possible contributor, although

most likely to effect presentations of less serious injury.

Greater ‘medicalisation’ of less serious injury refers to a

community wide change in injury management. Children with

injuries that were previously managed without medical input

(e.g. by caregivers) may be more likely to present to medical

practitioners. This is most likely to result in increases in

presentation of the least serious injuries.

Irrespective of the cause of greater injury presentations to the

EDs in this study, this increase has important implications for the

EDs and the health system in general. This increase should be

factored into any resource decisions made around infrastructure

and staffing of these metropolitan EDs.

It is not clear why there was a decrease in the incidence of injury

triaged as urgent or above and the incidence of admission between

1998/99 and 2000/01. However the incidence of injury admission

increased over the study period from 1,066 to 1,238/100,000/year.

Comparisons of the incidence of admission are complicated by

differences in the definition of an admission.26,30 It is likely that

this difference in definition will affect the incidence of admission

of older children as they are more likely to be managed using a

short-stay admission. Differences in the medical management of

injured children also affect the incidence of admission.

Age specific incidences in Australia, Victoria and Metropolitan

Brisbane and California are presented in Table 4. Wide variations

are seen and the large effect of removing same day admissions

is seen in the Victorian incidence. These definition and medical

management differences and resultant wide variations make

dataset comparisons very difficult.

The incidence of patient presentation with an injury triaged as

urgent or above increased over the study period. This increase

was of similar magnitude to the increase in the incidence of

admission but was distributed evenly across the age range (with

significant increases in all age groups). This increase is likely to be

independent of changes to the process of patient care and therefore

probably reflects a real increase in the incidence of serious injury of

16% over the study period. As this measure of acuity has not been

reported elsewhere in the literature, comparison across datasets

or populations was not possible.

Further studies could compare triage category with measures

of severity like the Abbreviated Injury Scale (AIS). The AIS has

been demonstrated to be a valid method of measuring injury

severity in paediatrics.31 Calculation of the AIS requires detailed

injury diagnosis information using ICD diagnosis coding that is

then mapped to AIS, a process that is likely to be more difficult

to apply to ED injury surveillance.31 It was not used in this study

as the injury surveillance collection system used only accepts a

single diagnosis.

The rate of triage category 4 presentations increased by 97%

over the study period, a much greater increase than presentations

triaged as urgent or above. This shows that while serious injury

presentations have increased, less serious injury presentations

increased much more. This is mostly likely to represent a change

Table 4: Age specific incidence of injury admission (/100,000/year).

Australia Australia Victoria Metro California, 1998-2000 2003/0429 200526 Brisbane US Urban30 2004/05 199731

0 to 4 years 1,865a 1,472 1,412 1,756 357a

5 to 9 years 1,479 1,351 1,292 946 243

10 to 14 years 1,518 1,551 1,319 886b 298Total 1,564 Not reported 1,374 (686c) 1,238 300d

Notes: a) 1 to 4 years b) 10 to 13 year age group c) With ‘same-day’ admissions excluded d) Estimated from published data

Shepherd Article


in healthcare usage, with greater proportion of less seriously

injured children presenting to ED, rather than being managed in

the community.

Children less than five years of age have the highest injury

mortality incidence32 and this study shows that they have the

highest incidence of injury presentation and serious injury. This

study also shows that serious injury in this age group has increased

over the past eight years. This increase in morbidity has occurred

despite a wide range of injury prevention initiatives.

LimitationsSome population data assumptions were required including;

estimating population numbers using census data (likely

underestimating population numbers by a small amount

– approximately 1.8% in 2001),33 assuming a linear change in

population between census dates; assuming a linear population

spread within the 10 to 14 year age group in order to exclude 14

year olds.

It was assumed that use of private hospitals remained constant

over the study period, as data were only available for 2005. It is

likely that private hospital use has increased in Brisbane over the

study period.34 A small underestimation of injury incidence may

have occurred as data were only available from two of the five

private hospitals in Brisbane that treat children in an ED. A small

overestimation of injury incidence may have occurred if some

children who are recorded as attending a private ED subsequently

presented to RCH or MCH, therefore essentially being recorded

twice.

Over the study period the percentage of children with injuries

who had complete injury surveillance information collected

decreased from 90% to 79%. This decrease in ascertainment

increases the risk of ascertainment bias. This study assumes that

the cases not ascertained by the injury surveillance process have a

similar profile to those collected. This assumption is supported by

literature21,22,24 and the age and ICD10 mapping detailed above.

ConclusionThis study illustrates the value of quality, prospectively collected

ED-based injury surveillance data. These data can be used to

determine incidences of the more serious spectrum of paediatric

injury (admissions or higher triage category presentations) and

changes in presentation patterns. Triage category is a useful

additional measure of injury severity.

There have been significant increases in injury related ED

presentations, injury related admissions and serious injuries (based

on triage category) in Brisbane between 1998 and 2005.

Implications Increasing incidence of paediatric injury presentation will require

further resources, including training and research in paediatric

emergency medicine and increased ED capacity. Injury prevention

priorities should be guided by injury incidence, for example focusing

on serious injuries sustained by 1 to 3 year olds.

The methodology of this study is easily repeatable and incidences

of injury should be calculated at regular intervals. This will allow

evaluation of the effects of further injury prevention interventions,

particularly on the more serious spectrum of paediatric injury

(admissions or higher triage category presentations).

AcknowledgementsStaff of the Queensland Injury Surveillance Unit and the

clerical and nursing staff of Mater Children’s and Royal Children’s

Brisbane Emergency Departments.

References1. Philippakis A, Hemenway D, Alexe DM, Dessypris N, Spyridopoulos T, Petridou

E. A quantification of preventable unintentional childhood injury mortality in the United States. Inj Prev. 2004;10(2):79-82.

2. Langley J, Cryer C. Indicators for injury surveillance. Australasian Epidemiologist. 2000;7(1):5-9.

3. Hendrie D, Miller TR. Assessing the burden of injuries: competing measures. Inj Control Saf Promot. 2004;11(3):193-9.

4. Horan JM, Mallonee S. Injury surveillance. Epidemiol Rev. 2003;25:24-42.5. Holder Y, Peden M, Krug E, et al, editors. Injury Surveillance Guidelines.

Geneva (CHE): World Health Organisation; 2001.6. Kreisfeld R, Harrison JE. Use of Multiple Causes of Death Data for Identifying

and Reporting Injury Mortality. Canberra (AUST): Australian Institute of Health and Welfare; 2007.

7. Langley J, Stephenson S, Cryer C, Borman B. Traps for the unwary in estimating person based injury incidence using hospital discharge data. Inj Prev. 2002;8(4):332-7.

8. Stone DH, Morrison A, Smith GS. Emergency department injury surveillance systems: the best use of limited resources? Inj Prev. 1999;5(3):166-7.

9. College of Emergency Nursing Australasia. Position Statement Triage Nurse. Australasian Emergency Nursing Journal. 2007;10(2):93-5.

10. Pardey TGM. The clinical practice of Emergency Department Triage: Application of the Australasian Triage Scale--An extended literature review: Part I: Evolution of the ATS. Australasian Emergency Nursing Journal. 2006;9(4):155-62.

11. Australasian College of Emergency Medicine. P06 - Australasian Triage Scale [policy on the Internet]. Melbourne (AUST): ACEM; 2006 [cited 2007 Sep]. Available from: http://www.acem.org.au/

12. Vance J, Sprivulis P. Triage nurses validly and reliably estimate emergency department patient complexity. Emerg Med Australas. 2005;17(4):382-6.

13. Durojaiye L, O’Meara M. A study of triage of paediatric patients in Australia. Emerg Med Australas. 2002;14(1):67-76.

14. Australian Bureau of Statistics. Queensland - National Regional Profile 2000 to 2004. Canberra (AUST): ABS; 2005.

15. Mater Children’s Hospital. Paediatric Emergency Medicine. Mater Hospital 1999. Available from: http://www.mater.org.au/getdoc/a423f7ed-a33e-43b1-9542-7da6a2fcdf7f/Paediatric-Emergency-Medicine.aspx [cited October 2007].

16. Royal Children’s Hospital. Department of Emergency Medicine. Queensland Government July 2008. Available from: http://www.health.qld.gov.au/rch/professionals/critcare_dem.asp [Cited October 2007]

17. Australian Bureau of Statistics [census data on the Internet]. Canberra (AUST): ABS; 2007 [cited June 2007]. 2001 Census Data by Location. Available from: http://www.censusdata.abs.gov.au/

18. Australian Bureau of Statistics [census data on the Internet]. Canberra (AUST): ABS; 2007 [cited June 2007]. 2006 Census Data by Location. Available from: http://www.censusdata.abs.gov.au/



19. Queensland Injury Surveillance Unit. Welcome to the Queensland Injury Surveillance Unit website. Queensland Injury Surveillance Unit; 2009. Available from: http://www.qisu.org.au/ModCoreFrontEnd/index.asp?pageid=109 [cited November 2007]

20. National Injury Surveillance Unit. National Data Standards for Injury Surveillance. Version 2.1. Canberra (AUST): Australian Institute of Health and Welfare; 1998 January.

21. Stokes M, Ozanne-Smith J, Harrison J, Steenkamp M. Validation of an injury surveillance epidemiological data system used within emergency departments. Int J Inj Contr Saf Promot. 2000;7(4):219-32.

22. Macarthur C, Pless IB. Sensitivity and representativeness of a childhood injury surveillance system. Inj Prev. 1999;5(3):214-6.

23. Gillam C, Meuleners L, Versluis L, Hendrie D, Sprivulis P. Electronic injury surveillance in Perth emergency departments: Validity of the data. Emerg Med Australas. 2007;19(4):309-14.

24. Hockey R, Horth A, Pitt WR. Validation study of injury surveillance data collected through Queensland hospital emergency departments. Emerg Med Australas. 2000;12(4):310-6.

25. Berry JG, Harrison JE. A Guide to Statistical Methods for Injury Surveillance. Adelaide (AUST): Australian Institute of Health and Welfare; 2006 October. Injury Technical Paper Series No.: 5 (AIHW Catalogue No.: INJCAT 72).

26. Cassell E, Clapperton A. Preventing unintentional injury in Victorian children aged 0-14 years: a call to action. Hazard (VISU). 2007 Autumn; 65:1-19.

Shepherd Article

27. Pitt WR, Balanda KP, Nixon J. Child injury in Brisbane South 1985-91: implications for future injury surveillance. J Paediatr Child Health. 1994;30(2):114-22.

28. Nolan T, Penny M. Epidemiology of non-intentional injuries in an Australian urban region: results from injury surveillance. J Paediatr Child Health. 1992;28(1):27-35.

29. National Center for Injury Prevention and Control [injury, violence and safety page on the Internet]. Atlanta (GA): CDC; 2006 [cited October 2007]. WISQARS Leading Causes of Nonfatal Injury Reports. Available from: http://webappa.cdc.gov/sasweb/ncipc/nfilead2001.html

30. Berry JG, Harrison JE. Hospital Separations Due to Injury and Poisoning, Australia, 2003–04. Adelaide (AUST): Australian Institute of Health Welfare; 2007.

31. Durbin DR, Localio AR, MacKenzie EJ. Validation of the ICD/AIS MAP for pediatric use. Inj Prev. 2001;7(2):96-9.

32. Commission for Children and Young People and Child Guardian. Chapter 3 − Deaths of children and young people 2004–05. In: Annual Report: Deaths Of children and Young People Queensland 2005-06. Brisbane (AUST): CCYPCG; 2006.

33. Australian Bureau of Statistics [homepage on the Internet]. Canberra (AUST): ABS; 2006 [cited 2007 Oct]. Undercount. In: 2901.0 - Census Dictionary, 2006 (Reissue). Available from: http://www.abs.gov.au/

34. Hargreaves J, Grayson N, Titulaer I. Trends in hospital service provision. Aust Health Rev. 2002;25(5):2-18.

trends in paediatric injury rates using emergency department based injury surveillance

Documents