influenza and congenital anomalies: a systematic review ... · please describe and define eurocat....

Draft Manuscript For Review. Reviewers should submit their review at http://mc.manuscriptcentral.com/humrep

Influenza and congenital anomalies: a systematic review

and meta-analysis

Journal: Human Reproduction

Manuscript ID: HUMREP-13-1069.R1

Manuscript Type: Meta-Analysis

Date Submitted by the Author: 12-Nov-2013

Complete List of Authors: Luteijn, Johannes; University of Ulster, School of Nursing Brown, Mary; University of Ulster, School of Nursing Dolk, Helen; University of Ulster, School of Nursing

Keywords: Influenza, Congenital Anomalies, Meta-analysis, Observational Studies,

Public Health

Specialty: Reproductive Epidemiology

Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online.

Luteijn_eAppendix Figures 12-51.zip

http://humrep.oupjournals.org

Draft Manuscript Submitted to Human Reproduction for Peer Review

Ref: 20102204 D07-06 OTH UK PS.PDF

1

Please find our responses to the comments made by the reviewers below. Reviewer comments in bold 1

green, our responses in red. 2

Reviewer: 1 3

Comments to the Author 4

This is an interesting manuscript about the association between influenza infection during the first 5

trimester of pregnancy and congenital anomalies. 6

7

The paper is well written. I only have relatively minor but still important comments. 8

1. Search strategy: Please add that MESH terms were used. 9

The fact that the search terms for PubMed and Embase databases were MeSH terms was added. Line 10

89. 11

2. Explain what AOR is in the M&M. 12

The following sentence was added in the methods section: “Meta-analysis was performed combining 13

adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for confounders).” Lines 132-133. 14

3a. Please state in the M&M that you combine OR and AOR. You state in the discussion that 15

adjustment for confounders showed a moderate effect on OR in studies that did report both 16

crude OR and adjusted OR. 17

Initially we did not combine OR and AOR as we extracted 2x2 tables and solely calculated crude ORs. 18

The sentence that was referring to the “moderate effect” was referring to the effect witnessed within 19

studies, rather than aggregate meta-analysis results. Based on comment #2 and comment #3 of 20

reviewer 1, it was decided to redo the meta-analysis on a combination of crude and adjusted ORs. This 21

had a number of consequences: 22

1. We are now able to include the study by Oster (Oster et al. 2011) in the meta-analysis. This 23

was not possible earlier since we did not get a reply from Oster et al to our request for the 24

core 2x2 tables which we required to calculate crude ORs. 25

Page 1 of 134




2

2. Since we now use AORs/ARRs, rather than ORs/RRs, whenever possible (and we include 26

Oster), slight changes to overall pooled meta-analysis estimates have occurred. For example, 27

our estimate for “all non-chromosomal CA” went from 2.25 (95% CI: 1.77-2.85) to 2.00 (1.62-28

2.48). 29

The following sentence was added in the methods section: “Meta-analysis was performed combining 30

adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for confounders). Only four case-control studies 31

provided AORs (Table IV) and none of the cohort studies made statistical adjustments (Table V). 32

Where adjusted estimates were not available, crude estimates were used. ” Lines 132-135. 33

3b. Please do add a subgroup analysis combining only OR and AOR. You do not need to add all 34

figures in my view but you can add a statement in the results section. 35

Solely three studies reported AOR and either crude OR or means of calculating the crude OR.(Botto et 36

al. 2001, Li et al. 2007, Lynberg et al. 1994) Granroth (Granroth et al. 1978) did report OR (2.5) and 37

AOR (1.7), however due to the lack of confidence intervals, we were unable to use Granroths AOR in 38

meta-analysis. Oster solely reported AOR. 39

We decided to do a subgroup analysis comparing adjusted estimates versus crude estimates for all 22 40

studies enrolled in the meta-analysis and enrolling Botto, Li and Lynberg in both categories (with 41

corresponding AOR and OR, respectively). The adjusted estimate pool consisted of solely 4 studies 42

(Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994, Oster et al. 2011), half of which reported 43

exclusively on NTDs. 44

The following statement was added in the results section “No differences were detected between 45

pooled adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77).” Lines 210-211. 46

Results of this subgroup analysis are not included in Table I. 47

4. Can you translate the OR in an absolute value in text or table, at least for the main outcomes? 48

It would be valuable to know how this translates as an RR can imply an increase from 40 to 80% 49

Page 2 of 134




3

or an increase from 0.1% to 0.2%. For instance compared to a risk in the control population of 50

1% the risk after influenza was found to be between 2 and 3%. 51

We recognize the importance of describing public health implications of findings. The following 52

sentence was added at the beginning of the discussion: “The twofold increase in risk of non-53

chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT Central Registry 2013) to 54

3.6% of births among 1st

trimester influenza exposed pregnancies.” Lines 264-266. 55

The EUROCAT congenital anomaly prevalence tables cited here allow for fast calculation of absolute 56

value effects of (A)ORs reported for any congenital anomaly in the systematic review and/or meta-57

analysis. 58

5. Data were pooled when at least 3 studies were available (page 7 line 129). In table II CA's are 59

described that are not included in the meta-analyses but I notice that some of the anomalies 60

are described by three or more studies. 61

This is correct. Not all studies included in the systematic review were enrolled in the meta-analysis. For 62

example, all ecological studies were excluded from meta-analysis since ecological study outcomes 63

cannot be combined into a summary estimate with outcomes of other studies. Some estimates 64

reported in table 2 are from the BWIS study for which we were unable to obtain data after contacting 65

the authors. 66

For additional clarification, “Note for some CA, ≥3 estimates were available from studies that could 67

not be included in the meta-analysis.” was added below Table II. 68

Page 3 of 134




4

Reviewer: 2 69

Comments to the Author 70

The authors pose the question whether first trimester maternal influenza infection increases the risk of 71

non-chromosomal anomalies. Overall, the systematic review was methodical and the study well 72

written. The study question is clearly formulated and answerable in the context of a systematic 73

review. The search for applicable studies was extensive and the selection process only biased by study 74

reporting in PubMed or EMBASE and publication in English or Dutch. Authors of prior studies were 75

contacted to obtain unavailable data. Inclusion and exclusion criteria for the selected studies were 76

clearly stated and appropriate, particularly exclusion of chromosomal anomalies. The quality of the 77

studies selected and risk of bias was assessed using a modification of the Newcastle-Ottawa scale, 78

which is appropriate. Sources of bias were discussed in detail in the text. Limitations were 79

thoughtfully considered and thoroughly discussed in the conclusion. The risk of bias was well 80

explored in the discussion and sources of heterogeneity and weighting discussed. 81

82

I have only a few minor Comments: 83

1. Please check that all abbreviations are defined in the text. It is not clear if “CA” is defined 84

earlier in the text. Was CHD defined? 85

Congenital anomy (CA) is now defined in the main body of text on the first occurrence (Line 78). 86

Abbreviations for specific CA (CHD, CNS, NTD and VSD) were removed from the manuscript as it was 87

felt these abbreviations might be less common. 88

2. Please describe and define EUROCAT. 89

EUROCAT is now defined. EUROCAT is described as “EUROCAT is a network of population-based 90

congenital anomaly registries which surveys over 1.7 million births annually in 23 European 91

countries.” Lines 119-120. 92

3. The sentence on lines 180-181 is confusing. Please reword. 93

Page 4 of 134




5

The sentence has been reworded as follows: 94

“Overall, our meta-analysis involved 29,542 CA cases of which 1,112 were exposed to influenza in the 95

first trimester of pregnancy and 53,089 controls of which 1,382 were exposed to influenza in the first 96

trimester of pregnancy from case-control studies.” Lines 192-194. 97

4. Please reword the sentence on lines 266-268. 98

The sentence has been reworded as follows: 99

“However, as long as maternal reports are collected prospectively prospective to the mother being 100

aware of the malformation (e.g. from medical records or interviews during pregnancy), there is no 101

reason to believe misclassification of influenza exposure will differ between arms cases and controls., 102

and t Therefore prospective maternal reports will not lead to a spurious association, but rather bias 103

the estimate toward the null as cases and controls are subject to similar misclassification.” Lines 289-104

293. 105

5. In line 333, please insert “but” after the comma. 106

“but” was added in the corresponding sentence. Line 366. 107

Page 5 of 134




6

Associate Editor's comments to Author: 108

This is a carefully performed meta-analysis of an important clinical topic. I really do have nothing to 109

add to the two high quality review reports. 110

111

Finally, the format of abstracts published in Human Reproduction has recently changed (please refer 112

to the journal's 'Information for Authors'). Attached is a pro-forma for the new, extended version that 113

should be used in the resubmitted article. 114

Page 6 of 134




7

Botto LD, Lynberg MC and Erickson JD. Congenital heart defects, maternal febrile illness, and 115

multivitamin use: A population-based study. Epidemiology 2001:12:485-490. 116

EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013:2013. 117

Granroth G, Haapakoski J and Saxen L. Defects of the central nervous system in Finland: V. 118

Multivariate analysis of risk indicators. Int J Epidemiol 1978:7:301-308. 119

Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural 120

tube defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 121

Teratol 2007:79:295-300. 122

Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: a 123

population-based case-control study. Am J Epidemiol 1994:140:244-255. 124

Oster ME, Riehle-Colarusso T, Alverson CJ and Correa A. Associations between maternal fever and 125

influenza and congenital heart defects. J Pediatr 2011:158:990-995. 126

Page 7 of 134




1

Manuscript title 1

Influenza and congenital anomalies: a systematic review and meta-analysis 2

Suggestion for a running title 3

Influenza and congenital anomalies 4

Authors full names 5

JM Luteijn. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 6

Ulster, Jordanstown Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 7

MJ Brown. Institute of Nursing Research/School of Nursing, University of Ulster, Jordanstown 8

Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 9

H Dolk. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 10


Page 8 of 134




2

Abstract 12

Study question: Does 1st

trimester maternal influenza infection increase the risk of non-13

chromosomal congenital anomalies? 14

Summary answer: 1st trimester maternal influenza exposure is associated with raised risk of a 15

number of non-chromosomal congenital anomalies, including neural tube defects, hydrocephaly, 16

congenital heart defects, cleft lip, digestive system defects and limb reduction defects. 17

What is known already: Hyperthermia is a well-established risk factor for neural tube defects. 18

Previous studies suggest influenza may be a risk factor not only for neural tube defects, but also 19

other congenital anomalies. No systematic review has previously been undertaken. 20

Study design, size, duration: Systematic review and meta-analysis. A search of EMBASE and PUBMED 21

was performed for English and Dutch studies published up to July 2013. A total of 33 studies (15 22

case-control, 10 cohort and 8 ecological) were included in the systematic review of which 22 studies 23

were included in the meta-analysis. 24

Participants/materials, settings, methods: A total of 29,542 babies with congenital anomaly (1,112 25

exposed) from case-control studies and 1,608 exposed pregnancies resulting in 56 babies with 26

congenital anomaly from cohort studies were included in the meta-analysis. Maternal influenza 27

exposure was defined as any reported influenza, influenza-like illness, or fever with flu, with or 28

without serological or clinical confirmation during the 1st trimester of pregnancy. Data for 24 29

(sub)groups of congenital anomaly available from ≥ 3 studies were analysed using the DerSimonian-30

Laird random effects model. The hypothesis of publication bias was assessed using funnel plots and 31

risk of bias of included studies was assessed using a slightly modified version of the Newcastle-32

Ottawa Scale. 33

Main results and the role of chance: 1st trimester maternal influenza exposure was associated with 34

an increased risk of any congenital anomaly (adjusted odds ratio 2.00, 95% CI: 1.62-2.48), neural tube 35

Page 9 of 134




3

defects (OR 3.33, 2.05-5.40), hydrocephaly (5.74, 1.10-30.00), congenital heart defects (1.56, 1.13-36

2.14), aortic valve atresia/stenosis (AOR 2.59, 1.21-5.54), ventricular septal defect (AOR 1.59, 1.24-37

2.14), cleft lip (3.12, 2.20-4.42), digestive system (1.72, 1.09-2.68) and limb reduction defects (2.03, 38

1.27-3.27). The increased risk for cleft lip (but not for cleft palate), was also reported by ecological 39

studies not included in the meta-analysis. Study outcomes reported for 27 subgroups of congenital 40

anomaly could not be included in the meta-analysis. Visual inspection of funnel plots did not suggest 41

evidence for publication bias. 42

Limitations, reasons for caution: This study enrolled observational studies which can be subject to 43

limitations such as confounding, retrospective maternal exposure reports and non-response of 44

intended participants. Influenza exposed pregnancies can also have been exposed to influenza 45

related medication. 46

Wider implications of the findings: Prevention of influenza in pregnant women may reduce 47

congenital anomaly risk, and would be relevant to more than just neural tube defects. More research 48

is needed to determine whether influenza and/or its related medication is teratogenic, to determine 49

the role of hyperthermia in teratogenicity and the role of other environmental factors such as 50

nutritional status in determining susceptibility. 51

Study funding/ competing interests: Funded by the EC, under the framework of the EU Health 52

Programme 2008-2013, Grant Agreement 2010 22 04 (Executive Agency for Health & Consumers). No 53

competing interests. 54

Page 10 of 134




4

Introduction 55

Both during seasonal influenza and pandemic influenza outbreaks, pregnant women have been at 56

risk of increased morbidity and mortality from influenza infection compared to the general 57

population.(Dodds et al. 2007, Harris 1919, Neuzil et al. 1998, Siston et al. 2010) Women in the later 58

stages of pregnancy are particularly vulnerable to adverse health outcomes after influenza infection 59

(Mak et al. 2008), perhaps because of immunological changes that take place during 60

pregnancy.(Jamieson et al. 2006) 61

Unravelling the question of teratogenicity of influenza is complex. In observational studies, influenza 62

exposure can affect the foetus not only via viral infection of the foetus, influenza-induced 63

hyperthermia and toxic metabolites associated with fever (Edwards 2006), but also via antiviral and 64

antipyretic use. A recent systematic review found strong evidence of an association between 65

maternal hyperthermia and neural tube defects.(Moretti et al. 2005) Evidence on other anomalies 66

with respect to hyperthermia or fever is scarce. Animal models have associated maternal 67

hyperthermia with arthrogryposis (Edwards 1971), congenital heart defects (Cockroft and New 1975, 68

Cockroft and New 1978), club foot(Edwards 1971), microcephaly (Edwards 1969, Edwards 1969), 69

microphthalmos (Germain et al. 1985) and others.(Edwards 2006) 70

The primary method of protecting pregnant women and their unborn child against influenza 71

infection is vaccination. In an increasing number of countries, pregnant women are advised to be 72

vaccinated against seasonal influenza infection.(Mak et al. 2008, Mereckiene et al. 2010) However, 73

vaccination policies in European countries vary both for seasonal influenza vaccination and during 74

the 2009 H1N1 influenza pandemic, especially with respect to trimesters eligible for 75

vaccination.(Luteijn et al. 2011, Mereckiene et al. 2010) In the absence of consensus on whether or 76

not to vaccinate 1st trimester pregnant women, the hypothesis of a causal relationship between 77

congenital anomalies (CA) and influenza virus deserves renewed attention. A better understanding of 78

the possible relationship between influenza and CA will allow for better understanding of the 79

Page 11 of 134




5

benefit-risk balance of vaccinating 1st

trimester pregnant women and women of childbearing age 80

against influenza. 81

The objective of our review is to identify and summarize the available epidemiologic evidence 82

regarding the risk of CA associated with 1st

trimester exposure to maternal influenza. 83

Materials and methods 84

Search strategy 85

This systematic review was informed by PRISMA guidelines (Moher et al. 2009) and the MOOSE 86

group guidelines.(Stroup et al. 2000) Two of the authors (JML and MJB) conducted the various steps 87

of the review and resolved any disagreements by discussion and consensus. The PubMed® and 88

Embase® databases were searched using the MeSH terms ("Influenza, Human") AND (pregnancy OR 89

congenital abnormality) and (Influenza) AND (pregnancy OR congenital abnormality), respectively on 90

July 1st, 2013. No publication or date restrictions were set. Where the papers’ abstract, title or 91

indexed MeSH terms suggested the possibility of reporting any fetal outcomes after maternal 92

exposure to influenza, the full paper was obtained. Reference lists of enrolled papers were reviewed. 93

Eligibility criteria 94

Case-control, cohort and ecological studies investigating CA outcomes following maternal exposure 95

to influenza were eligible for inclusion. No quality criteria were set for inclusion, although risk of bias 96

analysis was performed (eAppendix). Influenza was defined as any reported influenza, influenza-like 97

illness, or fever with flu, with or without serological or clinical confirmation. 98

Solely studies reporting influenza exposures during the 1st trimester of pregnancy were included in 99

the systematic review and meta-analysis. In order to be included in the meta-analysis, case-control 100

and cohort studies needed to allow for the calculation or report odds ratios (OR) or relative risks 101

(RR). For financial reasons, only English and Dutch language papers were eligible for inclusion. No 102

Dutch papers were included. 103

Page 12 of 134




6

Data extraction 104

Study characteristics were extracted by JML and MJB (eAppendix, Table III-V). Crude and adjusted 105

ORs and RRs and 2x2 tables relevant for meta-analysis were extracted by JML and MJB from cohort 106

and case-control studies. We contacted three authors of studies published ≤10 years ago(Czeizel et 107

al. 2008, Kelly et al. 2012, Oster et al. 2011) to obtain core data in order to create aggregate groups 108

such as orofacial clefts and for crude OR calculation and received the complete dataset for one 109

study.(Czeizel et al. 2008) In case of studies distinguishing between flu with fever and flu without 110

fever where it was impossible to combine the two, such as the stratified study by Lynberg (Lynberg et 111

al. 1994), the flu with fever dataset was extracted. We extracted data regarding malformed controls, 112

if available, rather than non-malformed controls since use of malformed controls reduces the impact 113

of differential recall bias. We recognize this can lead to underestimation of effect size if CA in the 114

control group are related to influenza exposure. In the meta-analysis, influenza exposures outside of 115

the 1st

trimester were added to the non-exposed cohort and for one cohort study without controls 116

(Doll et al. 1960), this allowed us to form a control group. 117

CA were classified into European surveillance of Congenital Anomalies (EUROCAT) defined subgroups, 118

excluding minor anomalies as specified by EUROCAT.(EUROCAT Central Registry 2009) EUROCAT is a 119

network of population-based congenital anomaly registries which surveys over 1.7 million births 120

annually in 23 European countries. CA were classified down to the greatest level of precision possible. 121

For example, a study that reported anomalies only as neural tube defects without further 122

specification contributed data to the analysis of neural tube defects but could not contribute data to 123

an analysis of anencephaly or spina bifida specifically. We excluded chromosomal syndromes since 124

their etiology is not related to maternal exposures. 125

Risk of bias assessment 126

In order to assess the validity of included studies’ findings we assessed the risk of bias of case-control 127

and cohort studies included in the meta-analysis using a slightly modified version of the Newcastle-128

Page 13 of 134




7

Ottawa scale (NOS).(Wells et al. 2004) Judgement criteria and deviations from the NOS are 129

summarized in the eAppendix. 130

Quantitative data summary and synthesis 131

Meta-analysis was performed combining adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for 132

confounders). Only four case-control studies provided AORs (Table IV) and none of the cohort studies 133

made statistical adjustments (Table V). Where adjusted estimates were not available, crude 134

estimates were used. 135

We assumed similarity between OR and RR because CA are rare events.(Davies et al. 1998) Statistical 136

analysis was performed using Stata version 9.2 [StataCorp, College Station, TX]. Meta-analysis was 137

performed on all CA combined and, for EUROCAT defined subgroups of CA (if n studies ≥ 3). The 138

DerSimonian & Laird random effects model (DerSimonian and Laird 1986) was used since the studies 139

in this meta-analysis involved varying countries, time periods and influenza strains. Subgroup analysis 140

was performed based on study type, publication date, risk of differential recall bias and adjustment 141

for confounders in order to assess the impact of these variables on study outcome. Subgroup 142

analyses combined all studies in the relevant categories, using the estimate for all non-chromosomal 143

CA combined where available and if not available, the estimate for the specific CA subgroup studied. 144

Due to scarce numbers and imbalance between some study arms, we used an alternative continuity 145

correction based on the OR of (other) studies with >0 events in both arms and group ratio imbalance 146

as discussed by Sweeting et al.(Sweeting et al. 2004) Heterogeneity between studies was assessed 147

using the I-squared statistic. Values of I-squared equal to 25%, 50% and 75% were considered to 148

represent low, moderate and high levels of heterogeneity, respectively. The hypothesis of 149

publication bias was assessed using funnel plots (eAppendix, Figures 28-51). 150

Results 151

Selection flow 152

Page 14 of 134




8

The PubMed® database search yielded 1369 papers and the Embase® database search yielded 2649 153

papers (Figure 1). After removing 1121 duplicates, a total of 2897 potentially relevant papers were 154

identified by the literature search. After screening by MeSH terms, titles and abstracts, 2615 papers 155

were excluded and full papers were retrieved for the remaining 282 papers. Of these, a total of 40 156

papers covering 27 studies met the inclusion criteria and were included in the systematic review. 6 157

additional eligible papers were detected by reference tracking, leading to a grand total of 46 included 158

papers covering 33 studies. 159

Study characteristics 160

The 46 enrolled papers were classified as 25 papers covering 15 case-control studies, 12 papers 161

covering 10 cohort studies and 9 papers covering 8 ecological studies. Enrolled studies are 162

summarized by study type in eAppendix, Table III-V and evidence provided by enrolled studies that 163

was not included in the meta-analysis is summarized in Table II. For one case-control study 164

information was limited to a conference abstract.(Choi and Klaponski 1970) Included papers were 165

published between 1953 and 2013, with the median year of publication 1971. 166


Visual inspection of the funnel plots did not suggest evidence for publication bias (eAppendix, 168

Figures 28-51). 169

Of the 15 case-control studies (25 papers), 10 studies did not take into account possible confounding 170

by maternal age, socioeconomic class or both. In the majority of these 10 studies some form of 171

matching between cases and controls (usually maternal ward, sex and day of birth) took place. Ten 172

papers relied on retrospective maternal reported influenza episodes (or timing of maternal 173

interviews was unknown), making these studies susceptible to differential recall bias. Of the 174

remaining 5 studies, 2 used serologic confirmation and 3 used prospectively collected antenatal 175

records. The last notable source of possible bias was that in 5 case-control studies over 20% of the 176

Page 15 of 134




9

cases intended for inclusion were not enrolled, making these studies susceptible to non-response 177

bias (eAppendix). 178

For the 10 cohort studies (12 papers), none took into account possible confounding by maternal age, 179

socioeconomic class or both. For 9 cohort studies, infants were not followed up for at least a year (or 180

unclear), making these studies susceptible to misclassification bias for some CA not apparent at birth. 181

6 of the studies used prospectively collected maternal reports for exposure, 3 serologic confirmation, 182

for 1 study exposure ascertainment was not described. The last notable source of possible bias was 183

that in 5 cohort studies the exposed cohort was not drawn from a clearly defined place and time, or 184

failed to enrol ≥80% of the population identified in specified place and time, raising questions over 185

representativeness of the enrolled exposed cohort. 186


Meta-analysis was possible for data from 22 studies, forming groups of ≥3 independent studies for 24 188

(sub)groups of EUROCAT defined major CA: any non-chromosomal major CA, neural tube defects, 189

anencephaly, encephalocele, spina bifida, hydrocephaly, congenital heart defects, orofacial clefts 190

(Figures 2-9) and sixteen other CA subgroups (eAppendix, Figures 12-27). 191

Overall, our meta-analysis involved 29,542 CA cases of which 1,112 were exposed to influenza in the 192

first trimester of pregnancy and 53,089 controls of which 1,382 were exposed to influenza in the first 193

trimester of pregnancy from case-control studies. From cohort studies, 1,608 exposed pregnancies 194

resulting in 56 CA plus 14,613 non-exposed pregnancies resulting in 347 CA were enrolled. The 195

enrolled cohort studies were relatively small with only the Coffey and Jessop study reaching over 50 196

CA (including minor anomalies) following maternal influenza exposure. Case-control studies more 197

readily enrol the large number required for research on CA and 6 out of 15 case-control studies 198

enrolled over 500 cases.(Botto et al. 2001, Czeizel et al. 2008, Granroth et al. 1978, Laurence et al. 199

1968, Oster et al. 2011, Saxen 1975) The larger numbers come at a cost and 11 out of 15 case-control 200

studies gathered exposure data by retrospective maternal reports. 201

Page 16 of 134




10

Meta-analysis discovered statistically significant associations between 1st

trimester influenza 202

exposure and a large number of CA subgroups (Table I) including all non-chromosomal CA combined 203

(OR 2.00, 95% CI: 1.62-2.48). Medium heterogeneity was detected for the pooled estimate of all non-204

chromosomal CA (64%). Subgroup analysis showed lower odds ratios for pooled case-control study 205

outcomes (OR 1.84, 95% CI: 1.49-2.27), than for pooled cohort study outcomes (OR 2.12, 95% CI: 206

1.20-3.75) while pre-1970 studies reported higher odds ratios (OR 2.47, 95% CI: 1.50-4.70) than 207

studies published after 1970 (OR 1.71, 95% CI: 1.41-2.08). Overall, studies susceptible to differential 208

recall bias reported a lower risk (OR 1.92, 95% CI: 1.35-2.72) than studies not susceptible to 209

differential recall bias (OR 2.12, 95% CI: 1.54-2.92). No differences were detected between pooled 210

adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77). 211

Central nervous system defects 212

Associations were found for all neural tube defects (OR 3.33, 2.05-5.40) and the neural tube defect 213

subgroups anencephaly (OR 3.52, 1.69-7.32) and spina bifida (OR 2.20, 1.48-3.28). The majority of 214

the 2,500 neural tube defects were reported by Czeizel (n=1,202, AOR 2.40, 1.30-4.40), Li (344, AOR 215

3.06, 1.40-6.67), Laurence (n=551, OR 3.93, 1.37-11.27) and Lynberg (331, AOR 1.70, 1.10-2.50). The 216

lower OR reported by Lynberg is related to our preference for malformed controls over healthy 217

controls for the study by Lynberg, which lead to more conservative estimates. The study by Lynberg 218

reports higher AOR when using healthy controls for neural tube defects (AOR 3.0, 1.9-4.7). We 219

discovered significant heterogeneity in the aggregate groups and neural tube defects, anencephaly, 220

encephalocele and hydrocephaly (Table I). The heterogeneity for neural tube defects, anencephaly, 221

encephalocele and hydrocephaly seems to be driven by the Coffey, Hirvensalo, Pleydell, Saxen (1960) 222

and Wilson studies which all contributed OR of >10 in at least one CA subgroup. The ecological study 223

on 1957 pandemic influenza by Hakosalo, enrolling 27 neural tube defects reported suggestive 224

evidence for a relationship between influenza and neural tube defects, but two larger ecological 225

studies by Leck (n=2,484 (Leck et al. 1969) and n=162 (Leck 1963)) did not find such evidence for any 226

neural tube defects subgroup. The study by Hakosalo calculated cases back to last menstrual period, 227

Page 17 of 134




11

while none of the studies by Leck had access to gestational length, making these studies susceptible 228

to misclassification of exposure introduced by assumptions around gestational age. Of all 8 ecological 229

studies included, only three corrected for gestational age.(Busby et al. 2005, Hakosalo and Saxen 230

1971, Saxen et al. 1990) 231

Orofacial clefts 232

Our meta-analysis found an association between orofacial clefts and 1st

trimester influenza exposure 233

(OR 1.96, 95% CI: 1.33-2.91). There was a significant association for cleft lip with or without palate 234

(OR 3.12, 2.20-4.42), but not for cleft palate (OR 1.05, 0.60-.1.84) and no heterogeneity was detected 235

in these pooled groups (Table II). The majority of the orofacial clefts (n=2,773) were reported by 236

Czeizel (n=1,956) and Saxen (1975a n=591 and 1975b, n=194) and these studies report ORs between 237

1.90 and 2.32. Four of the ecological studies, all by Leck, also reported the association between 238

orofacial clefts and influenza and two of these four studies reported associations for cleft lip ± cleft 239

palate, but not for isolated cleft palate.(Leck 1963, Leck et al. 1969) 240

Congenital heart defects 241

Our meta-analysis found an association between congenital heart defects and 1st trimester influenza 242

exposure (OR 1.56, 95% CI: 1.13-2.14). The vast majority of the congenital heart defects reported in 243

the meta-analysis were reported by Botto (n=829, AOR 2.1, 0.8-5.5), Czeizel (n=4,479, OR 1.6, 1.3-1.9) 244

and Oster (n=2,361, AOR 1.11, 0.91-1.35). It should be noted the study by Botto suffered from a 2-12 245

year delay between delivery and maternal interview while influenza exposure in the study by Oster 246

could occur from 3 months before pregnancy to the 3rd

month of pregnancy. 247

With respect to specific types of congenital heart defects (Table I), meta-analysis showed aortic valve 248

atresia/stenosis and ventricular septal defect to be associated with 1st


(OR 2.59, 1.21-5.54 and OR 1.59, 1.24-2.04, respectively). No associations were found for atrial septal 250

defect, hypoplastic left heart and transposition of the great vessels. Four congenital heart defect 251

subtypes were only reported by ≤2 studies and therefore not in the meta-analysis (Table II). Of these, 252

Page 18 of 134




12

two studies showed a consistent absence of association for tetralogy of Fallot, and a consistent and 253

high association for tricuspid atresia and stenosis (Botto 7.9, Oster 6.04). 254

Other anomalies 255

Limb reductions were associated with 1st

trimester influenza exposure by the meta-analysis (OR 2.03, 256

1.27-3.27) and this association is supported by several ecological studies (Table II). An association 257

with anophthalmia/microphthalmia is based on a single study (Table II). There was evidence that 258

there is no association with influenza for hypospadias (OR 1.02, 0.75-1.39) and clubfoot (OR 1.03, 259

0.83-1.27). 260

Discussion 261

This systematic review provides an overview of the published evidence on influenza exposure during 262

the 1st

trimester of pregnancy and CA. Meta-analysis revealed evidence for increases in a wide range 263

of major CA following 1st

trimester influenza exposure. The twofold increase in risk of non-264

chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT Central Registry 2013) to 265

3.6% of births among 1st

trimester influenza exposed pregnancies. 266

Exposure ascertainment 267

Case-control and cohort studies utilized serologic confirmation and maternal reports (prospective 268

and retrospective) for influenza exposure ascertainment. During influenza season the positive 269

predictive value of persons presenting with ILI for influenza is in the order of 66-77%.(Monto et al. 270

2000, Zambon et al. 2001) Serologic confirmation detects clinical and subclinical infections, which 271

might lead to different results since subclinical infections might affect the pregnant women 272

differently from clinical infections. Arguably, serologic confirmation of exposure is more reliable than 273

maternal reports. Five studies based on serologic confirmation were enrolled in the systematic 274

review (Elizan et al. 1969, Hardy et al. 1961, Walker and McKee 1959, Warrell et al. 1981, Wilson et al. 275

1959, Wilson and Stein 1969) of which three were included in the meta-analysis.(Hardy et al. 1961, 276

Warrell et al. 1981, Wilson et al. 1959, Wilson and Stein 1969) The two serologic studies limited to 277

Page 19 of 134




13

systematic review did not detect an association between influenza and neural tube defects (Elizan et 278

al. 1969) and did not detect an increased prevalence of CA among 1957 H2N2 Asian pandemic 279

influenza exposed pregnancies.(Walker and McKee 1959) The three serologic studies in the meta-280

analysis combined contributed 53 out of 27,584 CA, and it was therefore not possible to examine the 281

effect of exposure ascertainment method on effect size. One of these studies reported a possible 282

association between CA and 1957 H2N2 Asian pandemic influenza (Hardy et al. 1961) while a second 283

study did not.(Wilson et al. 1959, Wilson and Stein 1969) The third study did not detect an 284

association between CA and neural tube defects.(Warrell et al. 1981) Note that all studies using 285

serologic confirmation were limited to low numbers. 286

It is apparent that maternal reports lead to misclassification of exposure as women might not recall 287

infection, timing of infection relative to pregnancy or misdiagnose another infection for influenza. 288

However, as long as maternal reports are collected prospective to the mother being aware of the 289

malformation (e.g. from medical records or interviews during pregnancy), there is no reason to 290

believe misclassification of influenza exposure will differ between cases and controls. Therefore 291

prospective maternal reports will not lead to a spurious association, but rather bias the estimate 292

toward the null as cases and controls are subject to similar misclassification. Retrospective maternal 293

reports (e.g. interviews after birth), which are frequently utilized by case-control studies, are 294

susceptible to recall bias where mothers of cases have a different motivation to recall early 295

pregnancy exposure than mothers of non-cases. Mothers may differ not only in their tendency to 296

remember the infection, but in their tendency to misinterpret an illness as influenza.(MacKenzie and 297

Houghton 1974) Some of the studies included in the systematic review give an estimate of the 298

differential recall bias’ effect. For example, in the study by Lynberg, the OR for anencephaly after flu 299

with fever decreased from 3.1 (95% CI: 1.6-6.1) when compared with non-malformed controls to 1.4 300

(95% CI: 0.7-2.6) when compared with malformed controls.(Lynberg et al. 1994) Part of this decrease 301

might also have been related to CA in the control group being related to influenza, thus biasing the 302

OR towards 1. In our meta-analysis, the pooled estimate of OR for studies susceptible to recall bias 303

Page 20 of 134




14

(1.92) was slightly lower than the pooled estimate for other studies (2.00), contrary to expectation. 304

This is related to the low overall OR (1.11) reported by the large Baltimore-Washington Infant Study 305

(Oster et al. 2011), which was susceptible to differential recall bias and contributed over 20% to the 306

pooled estimate of susceptible studies. 307

Case-control, cohort and ecological study designs 308

Cohort studies failed to enrol large numbers of CA and this should not be surprising considering CA 309

only make up for 2-3% of births in a general population and short follow-up time after birth could 310

have led to underascertainment. Investigation of specific CA requires even higher numbers. Due to 311

the possibility of enrolling larger numbers, case-control studies seem better suited for addressing the 312

hypothesis of teratogenicity of influenza, although this comes at a cost as most case-control studies 313

ascertained exposure by retrospective maternal reports. 314

Ecological studies are generally considered a weaker study design than case-control or cohort studies 315

(Evans 2003) due to lack of individual exposure information. It cannot be verified that any excess CA 316

occurred among infected individuals. Correlation between influenza and confounding risk factors at 317

group level may lead to ecological fallacy, for example if influenza and nutritional deficiencies co-318

occur in winter. Ecological studies base exposure status on timing of pregnancy relative to influenza 319

season (or a proxy thereof) and therefore, the cohort defined as “exposed” is diluted by pregnancies 320

that did not have influenza. Population influenza exposure in the eight enrolled ecological studies 321

were derived from influenza incidences, counts or deaths (n=5) and sickness absenteeism rates or 322

claims (n=3). For this reason, the distinguishing power of ecological studies is highly dependent on 323

influenza attack rates and precision used to define influenza exposure. 324

The advantages of ecological studies are that large numbers of patients are enrolled easily and 325

ecological studies are not susceptible to the exposure misclassification or recall bias inherent in 326

individual level studies. Furthermore, they can be free of individual level confounding e.g. if those 327

most susceptible to influenza in the population have other risk factors for CA. Ecological studies 328

Page 21 of 134




15

therefore offer great value for addressing the hypothesis of teratogenicity of infectious diseases and 329

should not be discounted at the bottom of the evidence hierarchy in this area of research. 330

Consistency between study designs lends strength to a causal interpretation.(Hofler 2005) 331

Associations between 1st trimester influenza exposure and CA 332

One of the most striking results of the meta-analysis is the association between 1st

trimester 333

influenza exposure and neural tube defects. Neural tube defects are easily recognized at birth 334

eliminating susceptibility to underascertainment, while a previous meta-analysis reported an 335

association between neural tube defects and hyperthermia (OR 1.92, 95% CI: 1.62-2.29).(Moretti et 336

al. 2005) There also is evidence for neural tube defects following hyperthermia exposure in guinea 337

pigs.(Smith et al. 1992) In more recent human studies, underascertainment of neural tube defects 338

could have been a problem due to terminations of pregnancy for fetal anomaly. According to a 2004 339

EUROCAT analysis, 88% of neural tube defects are detected prenatally of which 88% are 340

aborted.(Boyd et al. 2008) Neural tube defects were more frequently studied than other CA, possibly 341

a result of interest in the 1950s studies by Coffey, suggesting an alarmingly strong link between 342

maternal influenza exposure and neural tube defects (data corresponds to OR 10.58, 4.30-26.02 in 343

the 1963 follow-up).(Coffey and Jessop 1963) This study utilized standardized questionnaires for 344

maternal interview after delivery and therefore was susceptible to differential recall bias, but this 345

limitation would not explain such a high OR. One of the ecological studies found an increase in neural 346

tube defects during the 1957 Asian influenza outbreak in Finland and concluded this might have been 347

caused either by influenza or influenza-related pharmaceuticals.(Hakosalo and Saxen 1971) The study 348

by Li et al had data available both on antiviral and antipyretic use (and other potential confounders, 349

see Table IV) and after adjusting for these co-exposures, OR for neural tube defects following 350

maternal influenza exposure dropped slightly but remained statistically significant (OR 3.93, 95% CI: 351

2.48-6.23). 352

Page 22 of 134




16

Hydrocephaly can sometimes be caused by spina bifida, and we could derive estimates for 353

hydrocephaly not associated with neural tube defects for two studies of the five studies on 354

hydrocephaly.(Coffey and Jessop 1959, Coffey and Jessop 1963, Hirvensalo and Kinnunen 1962) It is a 355

general problem in CA research that one baby may have more than one anomaly, and some of these 356

multiple malformed babies may have “sequences” (EUROCAT Central Registry 2009) that follow from 357

a primary anomaly. The data available for review cannot distinguish different types of diagnoses. 358

The evidence for an association between 1st

trimester influenza exposure and cleft lip with or 359

without palate is strong due to the lack of heterogeneity in the pooled data and consistent positive 360

associations detected across different study designs. It is well known that cleft lip +/- palate differs 361

aetiologically from cleft palate, so this difference is not surprising.(Mossey et al. 2009) The 362

associations detected for congenital heart defects and ventricular septal defects are somewhat 363

contradicted by the results from the BWIS.(Oster et al. 2011) On the other hand congenital heart 364

defects have been associated with hyperthermia in rats.(Cockroft and New 1975, Cockroft and New 365

1978) Club foot has been associated with hyperthermia in guinea pigs (Edwards 1971), but our meta-366

analysis did not find evidence for an association between club foot and influenza. A large number of 367

additional associations for other CA types were detected with more limited underlying evidence. 368

Pathways for mediation of hypothetical teratogenic effect of influenza 369

Influenza can mediate a possible teratogenic effect via multiple pathways and there is a risk of 370

confounding due to the intimate linkage between a disease and its cure. As well as antivirals, 371

antipyretics are also often used during influenza infection and the case-control study by Li et al 372

reported an AOR for antipyretic drugs and neural tube defects of 4.86 (95% CI: 1.33-17.78).(Li et al. 373

2007) Associations not adjusted for antivirals or antipyretics remain of importance since from a 374

vaccination policy perspective, it is less relevant whether the influenza virus or the antivirals are 375

causing any possible anomalies as vaccination will prevent both exposures. We recognize this puts 376

Page 23 of 134




17

limits on generalizability of study findings between populations with different use of 377

antipyretics/antivirals. 378

Direct pathways via which influenza infection can possibly lead to CA are toxic metabolites caused by 379

fever, hyperthermia and the influenza virus crossing the placenta. Hyperthermia has been associated 380

with causing neural tube defects as discussed above.(Moretti et al. 2005) It should be noted this 381

meta-analysis involved a large number of possible causes of hyperthermia, while influenza causes 382

high fever which might be different from general hyperthermia. This could explain the higher OR 383

reported in this meta-analysis for influenza exposure and neural tube defects (Table I). Several of the 384

included studies distinguished between 1st

trimester influenza and 1st

trimester fever.(Aro 1983, 385

Botto et al. 2001, Klemetti 1977, Lynberg et al. 1994, Oster et al. 2011, Saxen 1975, Saxen 1975) One 386

study found an association for influenza, but not for fever (Klemetti 1977), another found an 387

association for influenza with fever and neural tube defects (OR 1.7, 1.1-2.5), which was lowered for 388

influenza without fever (OR 1.3, 0.7-2.5).(Lynberg et al. 1994) A study on congenital heart defects 389

reported associations for fever (OR 1.8, 1.4-2.4) and influenza (OR 2.1, 0.8-5.5).(Botto et al. 2001) 390

while a second study on congenital heart defects reported very similar low and non-significant 391

excesses for fever (OR 1.14, 0.89-1.46) and influenza (OR 1.11, 0.91-1.35).(Oster et al. 2011) The two 392

studies by Saxen on orofacial clefts reported comparable associations for influenza (RR 2.00 for the 393

1st study) and fever (RR 1.96 for the 1

st study) (Saxen 1975, Saxen 1975), and the study by Aro on limb 394

reduction defects reported an OR of 1.6 for fever and 1.9 for influenza.(Aro 1983) It can be concluded 395

that included studies generally reported equivalent or stronger associations for influenza than for 396

fever with respect to CA following 1st

trimester exposure. 397

Another possible pathway by which the influenza virus can mediate a teratogenic effect is placental 398

transmission. Placental transmission has been documented but appears to be rare.(Gu et al. 2007, 399

McGregor et al. 1984) Toxic metabolites associated with fever as a cause of congenital anomalies 400

have also been suggested.(Edwards 2006) 401

Page 24 of 134




18

Limitations of the study 402

The systematic review results should be interpreted in the light of the findings that all of the included 403

studies are observational studies and many were susceptible to several types of bias and 404

ascertainment of exposure might not have been reliable. Although the funnel plots did not provide 405

evidence for publication bias, it should be noted that most of the included studies reported a wide 406

range of positive associations raising the question of whether studies reporting negative results 407

remained unpublished. Reference tracking identified 6 new studies and these studies were missed 408

because they were not indexed as influenza studies. This leaves the possibility open that some, 409

particularly negative, studies were missed due to poorly indexed terms. A possible reason for this is 410

that some case-control studies investigate a wide range of possible causes of CA and may tend to be 411

selectively indexed for the positive associations. For older studies, we could not be sure whether 412

chromosomal CA were excluded from analysis. 413

A weakness of the meta-analysis was that adjustment for confounders was not performed in most 414

included studies. Adjustment for confounders showed a moderate effect on OR within studies that 415

did report both crude OR and adjusted OR.(Acs et al. 2005, Granroth et al. 1978, Li et al. 2007) 416

Subgroup analysis comparing adjusted versus crude estimates did not detect differences, but this 417

could be related to the fact that 50% of the adjusted estimate was composed of neural tube defects 418

data since studies generally reported higher estimates for neural tube defects than for other CA. Due 419

to the limited amount of studies reporting adjusted OR, comparison of crude and adjusted OR for 420

subgroups of the same CA was not possible. Some very large datasets involved matched and/or 421

stratified controls (Botto et al. 2001, Czeizel et al. 2008), and most other datasets were matched by 422

one or more variables (Aro 1983, Granroth 1978, Granroth et al. 1978, Karkinen-Jaaskelainen and 423

Saxen 1974, Laurence et al. 1968, Li et al. 2007, Lynberg et al. 1994, Saxen 1975, Saxen 1975, Warrell 424

et al. 1981) lowering the impact of confounding on the meta-analysis. 425

Statistical limitations 426

Page 25 of 134




19

The study was susceptible to statistical limitations unique to meta-analysis of scarce events. In the 427

context of scarce events like CA and complicated exposure like 1st

trimester influenza, cohort studies 428

can enrol large numbers of exposed and unexposed, but have very few exposed CA outcomes. 429

Current statistical methods will assign these studies a lot of weight compared to case-control studies 430

with many cases and greater numbers of exposed cases. An example from this systematic review is 431

the cohort study by Pleydell, which reported 1 case of hydrocephaly among 12 1st


exposed pregnancies and 1 case among 1071 unexposed pregnancies leading to an OR of 97.27. The 433

heterogeneity model favours outliers and smaller studies and provided this study with 20% weight in 434

the overall hydrocephaly estimate, compared to 43% weight for the case-control study by Czeizel 435

which enrolled 314 cases of hydrocephaly (16 exposed). The weight allocated by the heterogeneity 436

model to the Pleydell study is clearly disproportionate. 437

A second problem lies in the continuity correction which is used to address zero events in one of 438

both arms of a study when pooling odds ratios. For rare events like CA, zero events in one or both 439

arms are not uncommon. The standard value for continuity correction is 0.5, but this arbitrary value 440

causes problems in studies with uneven arms and can even dominate the other arm. We addressed 441

this problem as proposed by Sweeting (Sweeting et al. 2004) by letting the continuity correction 442

depend on the OR of (other) studies with >0 events in both arms and group ratio imbalance. 443

However, this still led to OR > 1 for studies reporting 0 events in the exposed arm such as the 444

hydrocephaly data by Hirvensalo. 445

Conclusions and implications for CA prevention 446

Given the risk of congenital anomaly associated with influenza we show here, prevention of influenza 447

by vaccinating women who are planning to get pregnant may reduce congenital anomaly risk. 448

However, before evidence based policy can be implemented, further safety data on use of influenza 449

vaccines in pregnancy with respect to congenital anomalies is required (Kallen and Olausson 2012, 450

Page 26 of 134




20

Pasternak et al. 2012). Other methods for preventing influenza in early pregnancy include improving 451

nutritional and general health status and adopting behaviours which prevent interpersonal spread. 452

In conclusion, prevention of influenza in pregnant women may reduce congenital anomaly risk, and 453

would be relevant to more than just neural tube defects. More research is needed to determine 454

whether influenza and/or its related medication is teratogenic, to determine the role of 455

hyperthermia in teratogenicity and the role of other environmental factors such as nutritional status 456

in determining susceptibility. 457

Declaration of authors roles 458

J.M.L. (corresponding author, first author) was involved in study design, data collection and synthesis, 459

manuscript preparation and revision, construction of tables and figures, statistical analysis and 460

submission of the manuscript. M.J.B. (co-author) was involved in data collection and manuscript 461

revision. H.D. (co-author) was involved in study design, writing and revising the manuscript. 462

Acknowledgements 463

We would like to thank Dr. I. Barisic for assistance with case classification and we would like to thank 464

Dr. Julia Métneki, Dr. Erzsébet Puhó, Professor Andrew Czeizel and Dr. Annukka Ritvanen for sending 465

additional data. We would like to thank Professor Lolkje de Jong-van den Berg and Dr. Gordon 466

Marnoch for their helpful comments. 467

Funding 468

Funded by the EC, under the framework of the EU Health Programme 2008-2013, Grant Agreement 469

2010 22 04 (Executive Agency for Health & Consumers). 470

Conflict of interest 471

No competing interests 472

Page 27 of 134




21

Figure 1: Flow chart of systematic review 473

[Figure 1] 474

Figure 2: Forest plot of non-chromosomal CA following 1st


[Figure 2] 476

Figure 3: Forest plot of neural tube defects following 1st trimester influenza exposure 477

[Figure 3] 478

Figure 4: Forest plot of anencephaly following 1st


[Figure 4] 480

Figure 5: Forest plot of encephalocele following 1st


[Figure 5] 482

Figure 6: Forest plot of spina bifida following 1st


[Figure 6] 484

Figure 7: Forest plot of hydrocephaly following 1st


[Figure 7] 486

Figure 8: Forest plot of congenital heart defects following 1st


[Figure 8] 488

Figure 9: Forest plot of orofacial clefts following 1st


[Figure 9] 490

Page 28 of 134




22

References 491

Abramowitz LJ. The effect of Asian influenza on pregnancy. S Afr Med J 1958:32:1155-1156. 492

Acs N, Banhidy F, Puho E and Czeizel AE. Maternal influenza during pregnancy and risk of 493

congenital abnormalities in offspring. Birth Defects Res A Clin Mol Teratol 2005:73:989-996. 494

Adams MM, Mulinare J and Dooley K. Risk factors for conotruncal cardiac defects in Atlanta. J Am 495

Coll Cardiol 1989:14:432-442. 496

Aro T. Maternal diseases, alcohol consumption and smoking during pregnancy associated with 497

reduction limb defects. Early Hum Dev 1983:9:49-57. 498



Boyd PA, Devigan C, Khoshnood B, Loane M, Garne E, Dolk H and EUROCAT Working Group. 501

Survey of prenatal screening policies in Europe for structural malformations and chromosome 502

anomalies, and their impact on detection and termination rates for neural tube defects and Down's 503

syndrome. BJOG 2008:115:689-696. 504

Buck C. Exposure to virus diseases in early pregnancy and congenital malformations. Can Med Assoc 505

J 1955:72:744-746. 506

Busby A, Dolk H and Armstrong B. Eye anomalies: seasonal variation and maternal viral infections. 507

Epidemiology 2005:16:317-322. 508

Campbell WA. Influenza in early pregnancy; effects on the foetus. Lancet 1953:1:173-174. 509

Choi NW and Klaponski FA. On neural-tube defects: an epidemiological elicitation of etiological 510

factors. Neurology 1970:20:399-400. 511

Cockroft DL and New DA. Abnormalities induced in cultured rat embryos by hyperthermia. 512

Teratology 1978:17:277-283. 513

Cockroft DL and New DA. Effects of hyperthermia on rat embryos in culture. Nature 1975:258:604-514

606. 515

Coffey VP and Jessop WJ. Maternal influenza and congenital deformities. A follow-up study. Lancet 516

1963:1:748-751. 517

Coffey VP and Jessop WJ. Maternal influenza and congenital deformities: a prospective study. Lancet 518

1959:2:935-938. 519

Coffey VP and Jessop WJ. Congenital abnormalities. Ir J Med Sci 1955:6:30-48. 520

Csaky-Szunyogh M, Vereczkey A, Kosa Z, Urban R and Czeizel AE. Association of maternal diseases 521

during pregnancy with the risk of single ventricular septal defects in the offspring--a population-based 522

case-control study. J Matern Fetal Neonatal Med 2013:26:738-747. 523

Page 29 of 134




23

Czeizel AE, Puho EH, Acs N and Banhidy F. Use of specified critical periods of different congenital 524

abnormalities instead of the first trimester concept. Birth Defects Res A Clin Mol Teratol 2008:82:139-525

146. 526

Davies HT, Crombie IK and Tavakoli M. When can odds ratios mislead?. BMJ 1998:316:989-991. 527

DerSimonian R and Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986:7:177-188. 528

Dodds L, McNeil SA, Fell DB, Allen VM, Coombs A, Scott J and MacDonald N. Impact of influenza 529

exposure on rates of hospital admissions and physician visits because of respiratory illness among 530

pregnant women. CMAJ 2007:176:463-468. 531

Doll R, Hill AB and Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 532

1960:14:167-172. 533

Edwards MJ. Review: Hyperthermia and fever during pregnancy. Birth Defects Res A Clin Mol 534

Teratol 2006:76:507-516. 535

Edwards MJ. The experimental production of Arthrogryposis multiplex congenita in guinea-pigs by 536

maternal hyperthermia during gestation. J Pathol 1971:104:221-229. 537

Edwards MJ. The experimental production of clubfoot in guinea-pigs by maternal hyperthermia during 538

gestation. J Pathol 1971:103:49-53. 539

Edwards MJ. Congenital defects in guinea pigs: fetal resorptions, abortions, and malformations 540

following induced hyperthermia during early gestation. Teratology 1969:2:313-328. 541

Edwards MJ. Congenital defects in guinea pigs: prenatal retardation of brain growth of guinea pigs 542

following hyperthermia during gestation. Teratology 1969:2:329-336. 543

Elizan TS, Ajero-Froehlich L, Fabiyi A, Ley A and Sever JL. Viral infection in pregnancy and 544

congenital CNS malformations in man. Arch Neurol 1969:20:115-119. 545

EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013. 546

EUROCAT Central Registry. EUROCAT Guide 1.3 and reference documents. Instructions for the 547

Registration and Surveillance of Congenital Anomalies. Revised: August 2013. 548

Evans D. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare 549

interventions. J Clin Nurs 2003:12:77-84. 550

Germain MA, Webster WS and Edwards MJ. Hyperthermia as a teratogen: parameters determining 551

hyperthermia-induced head defects in the rat. Teratology 1985:31:265-272. 552

Granroth G. Defects of the central nervous system in Finland: III. Disease and drugs in pregnancy. 553

Early Hum Dev 1978:2:147-162. 554



Gu J, Xie Z, Gao Z, Liu J, Korteweg C, Ye J, Lau LT, Lu J, Gao Z, Zhang B et al. H5N1 infection of 557

the respiratory tract and beyond: a molecular pathology study. Lancet 2007:370:1137-1145. 558

Page 30 of 134




24

Hakosalo J and Saxen L. Influenza epidemic and congenital defects. Lancet 1971:2:1346-1347. 559

Hardy JM, Azarowicz EN, Mannini A, Medearis DN,Jr and Cooke RE. The effect of Asian influenza 560

on the outcome of pregnancy, Baltimore, 1957-1958. Am J Public Health Nations Health 561

1961:51:1182-1188. 562

Harris J. Influenza occurring in pregnant women: a statistical study of thirteen hundred and fifty cases. 563

JAMA 1919;72:978–80 1919:. 564

Hirvensalo M and Kinnunen O. Influenza and pregnancy. Duodecim 1962:78:740-747. 565

Hofler M. The Bradford Hill considerations on causality: a counterfactual perspective. Emerg Themes 566

Epidemiol 2005:2:11. 567

Jamieson DJ, Theiler RN and Rasmussen SA. Emerging infections and pregnancy. Emerg Infect Dis 568

2006:12:1638-1643. 569

Kallen B and Olausson PO. Vaccination against H1N1 influenza with Pandemrix((R)) during 570

pregnancy and delivery outcome: a Swedish register study. BJOG 2012:119:1583-1590. 571

Karkinen-Jaaskelainen and Saxen L. Maternal influenza, drug consumption, and congenital defects of 572

the central nervous system. Am J Obstet Gynecol 1974:118:815-818. 573

Kelly S, Keuhl K and Leffredo C. Tricupsid Atresia: an epidemiologic investigation in the Baltimore-574

Washington area. American Journal of Epidemiology 2012:175:S59. 575

Klemetti A. Definition of congenital malformations and detection of associations with maternal 576

factors. Early Hum Dev 1977:1:117-123. 577

Laurence KM, Carter CO and David PA. Major central nervous system malformations in South Wales. 578

II. Pregnancy factors, seasonal variation, and social class effects. Br J Prev Soc Med 1968:22:212-222. 579

Leck I. Further tests of the hypothesis that influenza in pregnancy causes malformations. HSMHA 580

Health Rep 1971:86:265-269. 581

Leck I. Examination of the Incidence of Malformations for Evidence of Drug Teratogenesis. Br J Prev 582

Soc Med 1964:18:196-201. 583

Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963:17:70-80. 584

Leck I, Hay S, Witte JJ and Greene JC. Malformations recorded on birth certificates following A2 585

influenza epidemics. Public Health Rep 1969:84:971-979. 586

Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural 587

tube defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 588

Teratol 2007:79:295-300. 589

Luteijn JM, Dolk H and Marnoch GJ. Differences in pandemic influenza vaccination policies for 590

pregnant women in Europe. BMC Public Health 2011:11:819. 591

Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: 592

a population-based case-control study. Am J Epidemiol 1994:140:244-255. 593

Page 31 of 134




25

MacKenzie JS and Houghton M. Influenza infections during pregnancy: association with congenital 594

malformations and with subsequent neoplasms in children, and potential hazards of live virus 595

vaccines. Bacteriol Rev 1974:38:356-370. 596

Mak TK, Mangtani P, Leese J, Watson JM and Pfeifer D. Influenza vaccination in pregnancy: current 597

evidence and selected national policies. Lancet Infect Dis 2008:8:44-52. 598

McGregor JA, Burns JC, Levin MJ, Burlington B and Meiklejohn G. Transplacental passage of 599

influenza A/Bangkok (H3N2) mimicking amniotic fluid infection syndrome. Am J Obstet Gynecol 600

1984:149:856-859. 601

Medveczky E, Puho E and Czeizel AE. An evaluation of maternal illnesses in the origin of neural-tube 602

defects. Arch Gynecol Obstet 2004:270:244-251. 603

Mereckiene J, Cotter S, D'Ancona F, Giambi C, Nicoll A, Levy-Bruhl D, Lopalco PL, Weber JT, 604

Johansen K, Dematte L et al. Differences in national influenza vaccination policies across the 605

European Union, Norway and Iceland 2008-2009. Euro Surveill 2010:15:19700. 606

Metneki J, Puho E and Czeizel AE. Maternal diseases and isolated orofacial clefts in Hungary. Birth 607

Defects Res A Clin Mol Teratol 2005:73:617-623. 608

Moher D, Liberati A, Tetzlaff J, Altman DG and PRISMA Group. Preferred reporting items for 609

systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009:339:b2535. 610

Monto AS, Gravenstein S, Elliott M, Colopy M and Schweinle J. Clinical signs and symptoms 611

predicting influenza infection. Arch Intern Med 2000:160:3243-3247. 612

Moretti ME, Bar-Oz B, Fried S and Koren G. Maternal hyperthermia and the risk for neural tube 613

defects in offspring: systematic review and meta-analysis. Epidemiology 2005:16:216-219. 614

Mossey PA, Little J, Munger RG, Dixon MJ and Shaw WC. Cleft lip and palate. Lancet 615

2009:374:1773-1785. 616

Neuzil KM, Reed GW, Mitchel EF, Simonsen L and Griffin MR. Impact of influenza on acute 617

cardiopulmonary hospitalizations in pregnant women. Am J Epidemiol 1998:148:1094-1102. 618



Park CH, Stewart W, Khoury MJ and Mulinare J. Is there etiologic heterogeneity between upper and 621

lower neural tube defects?. Am J Epidemiol 1992:136:1493-1501. 622

Pasternak B, Svanstrom H, Molgaard-Nielsen D, Krause TG, Emborg HD, Melbye M and Hviid A. 623

Risk of adverse fetal outcomes following administration of a pandemic influenza A(H1N1) vaccine 624

during pregnancy. JAMA 2012:308:165-174. 625

Pleydell MJ. Anencephaly and other congenital abnormalities. An epidemiological study in 626

Northamptonshire. Br Med J 1960:1:309-315. 627

Record RG. Anencephalus in Scotland. Br J Prev Soc Med 1961:15:93-105. 628

Page 32 of 134




26

Rockenbauer M, Olsen J, Czeizel AE, Pedersen L, Sorensen HT and EuroMAP Group. Recall bias in a 629

case-control surveillance system on the use of medicine during pregnancy. Epidemiology 630

2001:12:461-466. 631

Saxen I. The association between maternal influenza, drug consumption and oral clefts. Acta Odontol 632

Scand 1975:33:259-267. 633

Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev 634

Soc Med 1975:29:103-110. 635

Saxen L, Hjelt L, Sjostedt JE, Hakosalo J and Hakosalo H. Asian influenza during pregnancy and 636

congenital malformations. Acta Pathol Microbiol Scand 1960:49:114-126. 637

Saxen L, Holmberg PC, Kurppa K, Kuosma E and Pyhala R. Influenza epidemics and anencephaly. 638

Am J Public Health 1990:80:473-475. 639

Saxen L, Klemetti A and Haro AS. A matched-pair register for studies of selected congenital defects. 640

Am J Epidemiol 1974:100:297-306. 641

Siston AM, Rasmussen SA, Honein MA, Fry AM, Seib K, Callaghan WM, Louie J, Doyle TJ, 642

Crockett M, Lynfield R et al. Pandemic 2009 influenza A(H1N1) virus illness among pregnant women 643

in the United States. JAMA 2010:303:1517-1525. 644

Smith MS, Upfold JB, Edwards MJ, Shiota K and Cawdell-Smith J. The induction of neural tube 645

defects by maternal hyperthermia: a comparison of the guinea-pig and human. Neuropathol Appl 646

Neurobiol 1992:18:71-80. 647

Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA 648

and Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. 649

Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000:283:2008-650

2012. 651

Sweeting MJ, Sutton AJ and Lambert PC. What to add to nothing? Use and avoidance of continuity 652

corrections in meta-analysis of sparse data. Stat Med 2004:23:1351-1375. 653

Walker WM and McKee AP. Asian influenza in pregnancy; relationship to fetal anomalies. Obstet 654

Gynecol 1959:13:394-398. 655

Warrell MJ, Tobin JO and Wald NJ. Examination for influenza IgA and IgM antibodies in pregnancies 656

associated with fetal neural-tube defects. J Med Microbiol 1981:14:159-162. 657

Wells GA, Shea B and O'Connell D. The Newcastle-Ottawa Scale (NOS) for assessing the quality of 658

nonrandomised studies in meta-analyses 2004:. 659

Wilson MG, Heins HL, Imagawa DT and Adams JM. Teratogenic effects of Asian influenza. J Am 660

Med Assoc 1959:171:638-641. 661

Wilson MG and Stein AM. Teratogenic effects of asian influenza. A n extended study. JAMA 662

1969:210:336-337. 663

Zambon M, Hays J, Webster A, Newman R and Keene O. Diagnosis of influenza in the community: 664

relationship of clinical diagnosis to confirmed virological, serologic, or molecular detection of 665

influenza. Arch Intern Med 2001:161:2116-2122. 666

Page 33 of 134




1

Table I: First trimester maternal influenza exposure and risk of congenital anomalies: studies, total 1

number of cases of congenital anomaly, pooled odds ratio and heterogeneity. 2

Group Participating

studies (n)

I2

Statistic for

heterogeneity (%)

Pooled OR (95%

CI)

Total

number of

CA (n)

Any congenital anomaly 22 64 2.00 (1.62-2.48) 29,945c

-Susceptible to differential

recall bias

9 65 1.92 (1.35-2.72) 5,426

-Not susceptible to

differential recall bias

13 64 2.12 (1.54-2.91) 24,519c

-Case-control studies 13 60 1.84 (1.49-2.27) 29,542c

-Cohort studies 9 62 2.12 (1.21-3.72) 403

-Any type, published 1955-

1969

11 58 2.47 (1.50-4.70) 1,171

-Any type, Published 1975-

2011

11 55 1.71 (1.41-2.08) 28,774

-Adjusted estimates only a 4 87 2.15 (1.05-4.42) 3,865

-Crude estimates only a 21 61 2.22 (1.78-2.77) 27,584

Neural Tube Defects 11 50 3.33 (2.05-5.40) 2,500

-Anencephaly 10 44 3.52 (1.69-7.32) 608

-Encephalocele 4 63 2.95 (0.78-11.13) 225

-Spina Bifida 7 0 2.20 (1.48-3.28) 1,093

Hydrocephaly 5 45 5.74 (1.10-30.00) 323

Congenital heart defects 10 41 1.56 (1.13-2.14) 7,715

-Aortic Valve 3 31 2.59 (1.21-5.54) 167

Page 34 of 134




2

a Three studies (Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994) were able to provide estimates 3

both for adjusted and crude OR.

4

b Note that a single study contributed over 90% of the total weight to this pooled estimate

5

c Note that for some studies such as Czeizel et al 2008, the “any congenital anomaly” group also 6

included CA which were not included in any other analysis.

7

Atresia/Stenosis

-Atrial Septal Defect 3 0 0.82 (0.45-1.51) 429

-Hypoplastic Left Heart 3 0 1.58 (0.94-2.64) 203

-Transposition of the Great

Vessels

3 0 1.40 (0.90-2.17) 321

-Ventricular Septal Defect 4 0 1.59 (1.24-2.04) 1,434

Orofacial Clefts 10 37 1.96 (1.33-2.91) 2,773d

-Cleft Lip +- Palate b 7 0 3.12 (2.20-4.42) 1,404

-Cleft Palate b 3 0 1.05 (0.60-1.84) 584

Digestive System b 4 0 1.71 (1.09-2.69) 1,195

Urinary 5 0 1.45 (0.90-2.34) 48

Hypospadias b 4 0 1.02 (0.75-1.39) 3,041

Limb Reduction 3 0 2.03 (1.27-3.27) 1,002

Club Foot b 4 0 1.11 (0.93-1.34) 2,430

Hip Dislocation/Dysplasia 3 0 0.31 (0.00-37.62) 37

Polydactyly b 4 0 1.72 (0.85-3.48) 1,094

Syndactyly 3 71 1.98 (0.19-

20.563)

662

Musculo-Skeletal 3 0 1.05 (0.16-6.97) 776

Page 35 of 134




3

d Note that 591 orofacial clefts from the Saxen 1975a study (Saxen 1975) and 194 orofacial clefts 8

from the Saxen 1975b study (Saxen 1975) were not specified and therefore solely included in the 9

overall orofacial clefts analysis. 10

Page 36 of 134




1

Table II: Evidence included in the systematic review relating to 1st

trimester influenza exposure and 1

CA not included in meta-analysis. 2

Anomaly Evidence reported with respect to 1st

trimester influenza

exposure

Any anomaly Visual inspection: no convincing evidence that infant death rates

from CA are associated with 1st

trimester influenza

exposure(Buck 1955); RR 1.10(Leck 1963); single defect RR 1.00,

multiple defects RR 0.9(Leck 1964); single defect RR 1.03,

multiple defects RR 1.07(Leck et al. 1969)

Central Nervous System defects 0.81% in influenza exposed group vs. 0.30% and 0.32% in control

groups(Hakosalo and Saxen 1971)

Neural Tube Defects p=0.03, no data(Choi and Klaponski 1970)

Anencephaly RR 0.82(Leck 1963); sRR 1.32(Leck et al. 1969); RR 1.0, 0.8-

1.3(Saxen et al. 1990) No rise in anencephaly rates following

influenza epidemics detected(Record 1961)

Spina Bifida + Encephalocele RR 0.99(Leck 1963), sRR 1.13(Leck et al. 1969)

Hydrocephaly sRR 0.93(Leck et al. 1969)

Microcephaly OR 1.1, 0.3-4.1(Czeizel et al. 2008); sRR 0.24(Leck et al. 1969)

Anophthalmos/Microphthalmos OR 1.26, 1.02-1.57(Busby et al. 2005)

Congenital Cataract OR 0.8, 0.3-2.2(Czeizel et al. 2008)

Congenital Glucaoma OR 2.0, 0.4-10.9(Czeizel et al. 2008)

Congenital heart defects (suggestive but not significant; 0.93% in influenza exposed

group vs. 0.47% and 0.73% in control groups)(Hakosalo and

Saxen 1971)

Page 37 of 134




2

Atrioventricular Septal Defect OR 2.0, 0.3-15.3(Botto et al. 2001); AOR 1.29, 0.82-2.01(Oster et

al. 2011)

Tetralogy of Fallot OR 0.5, 0.1-3.6(Botto et al. 2001); AOR 0.78, 0.37-1.62(Oster et

al. 2011)

Triscuspid atresia and stenosis AOR 7.9, 0.3-29.6(Botto et al. 2001); AOR 6.04, 2.36-15.42(Oster

et al. 2011)

Ebstein's anomaly OR 3.0, 0.4-23.9(Botto et al. 2001)

Pulmonary valve stenosis AOR 1.21, 0.71-2.04(Oster et al. 2011)

Pulmonary valve atresia AOR 2.71, 1.16-6.32(Oster et al. 2011)

Coarctation of aorta AOR 3.8, 1.6-8.8(Botto et al. 2001); AOR 0.41, 0.13-1.33(Oster et

al. 2011)

Total anomalous pulm venous OR 2.2, 0.3-16.9(Botto et al. 2001)

Cleft Lip RR 1.55, p<0.05(Leck 1963); isolated RR 1.4, with other defects

RR 6.3(Leck 1964); cleft lip without cleft palate sRR 1.47,

p<0.05(Leck et al. 1969); cleft lip without cleft palate RR 1.64,

p<0.05(Leck 1971); cleft lip without cleft palate sRR 1.08, cleft

lip with cleft palate sRR 1.01(Leck et al. 1969); cleft lip with cleft

palate sRR 1.10(Leck et al. 1969), cleft lip with cleft palate RR

1.35(Leck 1971)

Cleft Palate RR 0.81(Leck 1963); sRR 0.94(Leck et al. 1969); sRR 0.93(Leck et

al. 1969)

Page 38 of 134




3

Digestive System 1 in 63 exposed during 1st

trimester of pregnancy and 4 in 1106

non-exposed, RR 4.4(Coffey and Jessop 1959, Coffey and Jessop

1963); 1 in 171 exposed during 1st

trimester of pregnancy and 13

in 6720 non-exposed, RR 3.0(Hirvensalo and Kinnunen 1962)

Oesophageal atresia

with/without

tracheo-oesophageal fistula

OR 1.6, 0.2-11.7(Czeizel et al. 2008); RR 2.41, p<0.01(Leck 1963);

isolated RR 1.5, with other defects RR 4.8(Leck 1964); sRR

1.12(Leck et al. 1969)

Atresia/Stenosis of the small

intestine

OR 2.1, 0.5-8.0(Czeizel et al. 2008)

Ano-rectal atresia and stenosis OR 1.1, 0.4-3.0(Czeizel et al. 2008); RR 2.12, p<0.05(Leck 1963);


0.80(Leck et al. 1969)

Hirschsprung's disease OR 0.7, 0.2-3.0(Czeizel et al. 2008)

Diaphragmatic Hernia OR 3.2, 1.2-8.8(Czeizel et al. 2008); 0 in 171 exposed during 1st

trimester of pregnancy and 3 in 6720 non-exposed(Hirvensalo

and Kinnunen 1962); RR 1.05(Leck 1963); sRR 1.01(Leck et al.

1969)

Abdominal Wall Defects OR 2.8, 1.1-6.9(Czeizel et al. 2008)

Omphalocele RR 1.92, <0.05(Leck 1963); isolated RR 1.3, with other defects RR

2.3(Leck 1964); sRR 1.16(Leck et al. 1969)

Urinary 1 in 171 exposed during 1st trimester of pregnancy and 42 in

6720 non-exposed(Hirvensalo and Kinnunen 1962)

Bilateral renal agenesis including

Potter syndrome

OR 0.8, 0.2-2.8(Czeizel et al. 2008); RR 1.38(Leck 1963)

Page 39 of 134




4

Congenital Hydronephrosis RR 1.63, p=0.05(Leck 1963)

Hypospadias RR 1.51(Leck 1963); sRR 0.93(Leck et al. 1969)

Limb sRR 0.75(Leck et al. 1969)

Limb Reduction Limited to thumbs or radii RR 2.20(Leck 1963); both arms or legs

RR 1.30(Leck 1963); one arm or leg RR 1.37(Leck 1963); sRR

1.32(Leck et al. 1969); RR 1.22(Leck 1971)

Upper Limb Reductions sRR 1.91, p<0.01(Leck et al. 1969)

Lower Limb Reductions sRR 1.23(Leck et al. 1969)

Club Foot sRR 0.89(Leck et al. 1969)

Hip Dislocation and/or Dysplasia sRR 0.77(Leck et al. 1969)

Disorders of skin 0 in 171 exposed during 1st trimester of pregnancy and 3 in 6720

non-exposed(Hirvensalo and Kinnunen 1962)

Down’s Syndrome sRR 1.16(Leck et al. 1969)

Solely data involving >2 CA cases was included and some reported CA could not be translated to 3

EUROCAT defined subgroups. We did not distinguish between “positive” and “negative” findings 4

since some of the studies were severely underpowered. Note for some CA, ≥3 estimates were 5

available from studies that could not be included in the meta-analysis. 6

Page 40 of 134




4,018 potentially relevant articles identified by Embase (2,649) and PubMed® (1,369)

database search

6 potentially relevant articles identified by

reference tracking

2,897 unique articles -1,121 duplicates within and across databases

2,897 unique articles screened

2,615 articles excluded -No abstract. Title, MeSH, publication

type ruled out relevance 843 - Title, MeSH and abstract ruled out

relevance 1,772

282 + 6 articles selected for full review

242 articles excluded -Relevance unknown due to foreign

language 39 - Non relevant original research 134

-No original data in article 50 -Exposure timing outside 1st trimester/non specified 13

-Exposure not sufficiently specified 6

46 articles included in systematic review

Screening

Eligibility

Included

32 articles (22 datasets) included in meta-analysis

13 articles excluded -Ecological studies 9

- Cohort studies without controls 1 -Studies with insufficient information 4

Page 41 of 134




239x202mm (72 x 72 DPI)

Page 42 of 134




311x225mm (72 x 72 DPI)

Page 43 of 134




283x202mm (72 x 72 DPI)

Page 44 of 134




306x202mm (72 x 72 DPI)

Page 45 of 134




298x202mm (72 x 72 DPI)

Page 46 of 134




306x202mm (72 x 72 DPI)

Page 47 of 134




285x202mm (72 x 72 DPI)

Page 48 of 134




280x202mm (72 x 72 DPI)

Page 49 of 134




1

eAppendix 1

Quality assessment of case-control and cohort studies enrolled in meta-analysis 2

For quality assessment of studies included in the meta-analysis, we used a slightly modified version 3

of the Newcastle-Ottawa scale (NOS).(Wells et al. 2004) We did not distinguish between community-4

based or hospital-based controls since most studies would recruit from antenatal clinics it is safe to 5

assume that this population is similar to a community-recruited population. In addition, the risk of 6

referrals would not have been present in studies from before the late 1980s before prenatal 7

screening was widely available. There is some risk of bias introduced from high risk pregnancies 8

being more likely to book at specific hospitals. The following NOS items were not assessed: due to 9

the unique nature of CA we did not require case-control studies to demonstrate that controls did not 10

have a history of the outcome of interest and we did not require cohort studies to provide 11

demonstration that the outcome of interest was not present at start of study. Quality scores were 12

not used since weighting and calculating overall quality scores is arbitrary and the importance of 13

individual items and direction of bias depends on the context in which they are applied. 14

Case-control studies 15

Case definition was considered low risk of bias if the source reported diagnosis by healthcare 16

professional, hospital or primary records, or utilized independent validation. Representativeness of 17

cases were considered low risk of bias if cases were drawn from a clearly defined period of time and 18

catchment area, or consisted of an appropriate random sample of cases. We considered selection of 19

controls low risk of bias if controls could have been in the case group had the outcome of interest 20

been present as opposed to controls from different time periods or locations than cases. For 21

assessing risk of confounding (comparability in NOS) we considered any comparative study which did 22

match or adjust for confounding by at least maternal age and socioeconomic class low risk of bias. 23

For socioeconomic class, several proxies such as malnutrition for older studies and race for American 24

studies were also considered satisfactory. Fever and antiviral use were not considered confounders 25

Page 50 of 134




2

for reasons stated in the introduction. For ascertainment of exposure, we considered serological 26

confirmation, diagnosis by healthcare professionals and prospective interviews low risk of bias. Post 27

delivery maternal interviews were considered high risk of bias since mothers of infants born with 28

anomalies might make a more determined effort to recall illnesses during pregnancy than other 29

mothers leading to differential recall bias. In case-control designs, recall bias has been estimated to 30

bias OR up to a factor of 1.9.(Rockenbauer et al. 2001) We considered attrition or non-response 31

rates ≥20% for cases or controls high risk of bias. 32

Cohort studies 33

Representativeness of the exposed cohort was considered low risk of bias if the exposed cohort was 34

drawn from a clearly defined period of time and catchment area, or consisted of an appropriate 35

random sample of eligible cohort subjects. If ≤80% of the identified individuals in this place and time 36

failed to enrol, representativeness was also considered high risk of bias. We considered selection of 37

the unexposed cohort low risk of bias if individuals in the unexposed cohort could have been in the 38

exposed cohort had they been exposed to 1st

trimester influenza infection. Ascertainment of 39

exposure was handled the same way as for the case-control studies (see above). Length of follow-up 40

is only relevant for certain congenital anomalies such as congenital heart defects and we considered 41

a length of follow up of 1 year satisfactory. With respect to adequacy of follow up, we considered 42

≤20% attrition low risk of bias. 43

Figure 10: Risk of bias assessment of included case-control studies. 44

[Figure 10] 45

Green represents low risk of bias, red represents high risk of bias, yellow represents unknown risk of 46

bias. 47

Figure 11: Risk of bias assessment of included cohort studies. 48

[Figure 11] 49

Page 51 of 134




3


bias. 51

Supplementary Forest Plots 52

Figure 12: Forest plot of aortic valve atresia/stenosis following 1st


[Figure 12] 54

Figure 13: Forest plot of atrial septal defect following 1st


[Figure 13] 56

Figure 14: Forest plot of hypoplastic left heart following 1st


[Figure 14] 58

Figure 15: Forest plot of transposition of the great vessels following 1st


[Figure 15] 60

Figure 16: Forest plot of ventricular septal defects following 1st


[Figure 16] 62

Figure 17: Forest plot of cleft lip ± cleft palate following 1st


[Figure 17] 64

Figure 18: Forest plot of cleft palate following 1st trimester influenza exposure 65

[Figure 18] 66

Figure 19: Forest plot of digestive system defects following 1st


[Figure 19] 68

Figure 20: Forest plot of urinary defects following 1st


[Figure 20] 70

Figure 21: Forest plot of hypospadias following 1st


[Figure 21] 72

Figure 22 Forest plot of limb reduction defects following 1st


[Figure 22] 74

Page 52 of 134




4

Figure 23: Forest plot of club foot following 1st trimester influenza exposure 75

[Figure 23] 76

Figure 24: Forest plot of hip dislocation and/or dysplasia following 1st


[Figure 25] 78

Figure 25: Forest plot of polydactyly following 1st


[Figure 25] 80

Figure 26: Forest plot of syndactyly following 1st trimester influenza exposure 81

[Figure 26] 82

Figure 27: Forest plot of musculo-skeletal defects following 1st


[Figure 27] 84

Supplementary Funnel Plots 85

Funnel plot asymmetry is indicative for publication bias as for example small studies reporting 86

negative findings might not be published. Visual inspection of funnel plots did not provide evidence 87

for publication bias. 88

Figure 28: Funnel plot of non-chromosomal CA following 1st


[Figure 28] 90

Figure 29: Funnel plot of neural tube defectsfollowing 1st trimester influenza exposure 91

[Figure 29] 92

Figure 30: Funnel plot of anencephaly following 1st


[Figure 30] 94

Figure 31: Funnel plot of encephalocele following 1st


[Figure 31] 96

Figure 32: Funnel plot of spina bifida following 1st


[Figure 32] 98

Figure 33: Funnel plot of hydrocephaly following 1st


Page 53 of 134




5

[Figure 33] 100

Figure 34: Funnel plot of congenital heart defects following 1st trimester influenza exposure 101

[Figure 34] 102

Figure 35: Funnel plot of orofacial clefts following 1st


[Figure 35] 104

Figure 36: Funnel plot of aortic valve atresia/stenosis following 1st


[Figure 36] 106

Figure 37: Funnel plot of atrial septal defect following 1st


[Figure 37] 108

Figure 38: Funnel plot of hypoplastic left heart following 1st


[Figure 38] 110

Figure 39: Funnel plot of transposition of the great vessels following 1st


[Figure 39] 112

Figure 40: Funnel plot of ventricular septal defects following 1st


[Figure 40] 114

Figure 41: Funnel plot of cleft lip ± cleft palate following 1st


[Figure 41] 116

Figure 42: Funnel plot of cleft palate following 1st


[Figure 42] 118

Figure 43: Funnel plot of digestive system defects following 1st


[Figure 43] 120

Figure 44: Funnel plot of urinary defects following 1st


[Figure 44] 122

Figure 45: Funnel plot of hypospadias following 1st


[Figure 45] 124

Figure 46: Funnel plot of limb reduction defects following 1st trimester influenza exposure 125

Page 54 of 134




6

[Figure 46] 126

Figure 47: Funnel plot of club foot following 1st trimester influenza exposure 127

[Figure 47] 128

Figure 48: Funnel plot of hip dislocation and/or dysplasia defects following 1st


exposure 130

[Figure 48] 131

Figure 49: Funnel plot of polydactyly following 1st trimester influenza exposure 132

[Figure 49] 133

Figure 50: Funnel plot of syndactyly following 1st trimester influenza exposure 134

[Figure 50] 135

Figure 51: Funnel plot of musculo-skeletal defects following 1st


[Figure 51] 137

Page 55 of 134




1

Table III: Overview of 9 papers from 8 ecological studies included in the systematic review. 1

Authors Publication

year (study

years) and

location

Study outcome Influenza

exposure and

critical period

No. of cases and

methodology

Buck

(Buck

1955)

1955

(Jan 1944-Jul

1951), Canada

Infant death due to CA Influenza

incidence per

calendar month

coinciding with

1st trimester of

pregnancy

Unclear population

size, infant deaths

only. Analysis by visual

inspection.

Busby et

al (Busby

et al.

2005)

2005 (1988-

1994), England

Anophthalmos and

microphthalmos

Reported

weekly

influenza

infection counts

coinciding with

6th-10th

week

of gestation.

275 malformed infants,

live births and

stillbirths. Analysis by

Poisson regression

comparing cases to all

births.

Hakosalo

and Saxen

(Hakosalo

and Saxen

1971)

1971 (Jan 15th

-

Oct 31st, 1958),

Finland

Anencephaly and CA of

the CNS, circulatory

system and urogenital

system.

Sickness

absenteeism

rates coinciding

with 5th

-11th

week of

gestation.


live births and


χ2 test, comparing high

risk versus low risk

pregnancies. High risk

pregnancies were

Page 56 of 134




2

exposed to Pandemic

Asian Flu in 5th

-11th

week of gestation. Low

risk pregnancies were

conceived earlier

(Group 1) or conceived

later (Group 2)

Leck (Leck

1963, Leck

1964)

1963 and 1964

(1957-

1961(Leck

1963), 1957-

1963(Leck

1964)),

Birmingham,

UK

Anencephaly, Spina

Bifida, Cleft Lip, Cleft

Palate, Oesophageal

Atresia, Anal Atresia,

Renal Agenesis or

Hypoplasia,

Hydronephrosis,

Hypospadias, Limb

Reductions (Thumb or

Radii, Bilateral arms or

legs, single arm or leg),

Exomphalos and

Diaphragmatic

Hernia.(Leck 1963)

Weekly new

claims to

sickness benefit

coinciding with

26 to 40 weeks

before delivery.


live births and

stillbirths.(Leck 1963)

15 additional 1957-

1961 cases and

unknown amount of

1962-1963 cases.(Leck

1964)

Analysis by incidence

ratio’s of high risk

versus low risk

pregnancies.

Leck et al

(Leck et al.

1969)

1969 (1956-

1965), Several

USA states

Orofacial Clefts (1956-

1965) and any CA (1961-

196%). Anencephalus,

Spina Bifida,

Weekly number

of deaths

attributable to

influenza

For 1956-1961: 1,496

isolated cleft palate,

1,418 isolated cleft lip,

2,298 cleft lip with

Page 57 of 134




3

Hydrocephalus,

Microcephalus, Cleft

Palate, Cleft Lip, Cleft Lip

with Cleft Palate,

Esophageal Defects,

Anoreactal Defects,

Hypospadias, Clubfoot,

Reduction Defects

(Upper Limb, Lower

Limb, Upper and Lower

Limb, Limbs

Unspecified), Congenital

Hip Dislocation,

Diaphragmatic Hernia,

Down’s Syndrome and

Exomphalos.

coinciding with

26 to 40 weeks

before delivery.

cleft palate. For 1962-

1965: 9,980 CAs. Live

births only. Analysis by

crude and

standardized (by

season and year)

incidence ratios of high


pregnancies and χ2

test.

Leck (Leck

1971)

1971

(Birmingham:

1954-1965,

England and

Wales 1964-

1968, USA:

1955-1965).

Birmingham: Cleft Lip,

Cleft Lip with Cleft Palate

and Limb Reductions.

England and Wales:

Anencephalus, Spina

Bifida, Encephalocele,

Cleft Lip, Cleft Lip with

Cleft Palate, Esophageal

Atresia and Stenosis,

Weekly new

claims to

sickness benefit

coinciding with

26 to 40 weeks

before delivery.

Birmingham: 106 cleft

lip, 200 cleft lip with

cleft palate and 136

limb reductions, live

births and stillbirths.

England and Wales:

13,707 CAs, live births

and stillbirths. USA:

Unknown amount of

Page 58 of 134




4

Anorectal Atresia and

Stenosis, Limb

Reduction, Exomphalos.

USA: Cleft Lip and Limb

Reductions.

CAs, live births only.

Analysis by crude and

standardized (by

season and year)



pregnancies and χ2

test.

Record

(Record

1961)

1961 (1938-

1958),

Scotland.

Anencephaly Winters (Oct-

Mar) with

influenza

deaths >500

were

considered

epidemic

winters and

linked with

anencephaly

incidences the

next May-Oct.

5,039 Anencephaly

cases. Analysis by

visual inspection.

Saxen et

al (Saxen

et al.

1990)

1990 (1968-

1982), Finland

Anencephaly High risk

months were

identified via

the Central

Public Health

248 Anencephaly

cases. Analysis by rate

ratio’s between high

risk and low risk

pregnancies.

Page 59 of 134




5

Laboratory and

coinciding

pregnancies in

the 1st

trimester

were

considered high

risk.

Results from these studies are included in Table II. 2

Page 60 of 134




1

Table IV: Overview of 25 papers from 15 case-control studies sorted by year. 1

1st

Author Publication

year (study

years) and

location

Exposure

ascertainment

Birth types Controls, matching

and adjustment (if

any)

Coffey (Coffey and

Jessop 1955)

1955 (1953-

1954),

Ireland

Exposure was based on

maternal reports,

retrospective to

delivery.

Live and

stillbirths.

Random sample of

non-malformed

controls.

Laurence (Laurence

et al. 1968)

1968 (1956-

1962), UK

The study suggests

exposure was based on

maternal reports,

retrospective to

delivery for 1956-1960

and mainly prospective

to delivery after 1960.

Live and

stillbirths.

Controls without

neural tube defects.

Matched by date of

birth, place of

residence, sex of

infant.

Elizan (Elizan et al.

1969)

1969 (1959-

1964), USA


complement fixation

test.

Autopsied

stillbirths.

Controls without

central nervous

system defects.

Matched by

gravidity, institution,

last menstrual

period, maternal age

and race.

Choi (Choi and 1970 (1963- Exposure was based on It is Controls without

Page 61 of 134




2

Klaponski 1970) 1968),

Canada

maternal interviews

after birth

supplemented by

medical records.

unknown

which birth

types were

included.

neural tube defects.

Matched by date of

birth, hospital, sex of

infant.

Saxen (Saxen 1975) 1975a

(1967-1971),

Finland

Finnish registry of

congenital

malformations (FRCM).


maternal reports,

retrospective to

delivery.

Live and

stillbirths.

Controls without

orofacial clefts.

Matched by place of

residence and time

of birth.

Saxen (Saxen 1975) 1975b

(1972-1973),

Finland

Finnish registry of

congenital



maternal reports,

retrospective to

delivery.

Live and

stillbirths.

Controls without

orofacial clefts.

Matched by place of

residence and time

of birth.

Klemetti (Klemetti

1977)

1977 (1963-

1965),

Finland

Finnish registry of

congenital



maternal reports,

prospective to delivery.

Live and

stillbirths.

Non-malformed

controls and cases

originated from a

series of consecutive

pregnancies

registered in the

same county.

Page 62 of 134




3

Granroth (Granroth

1978, Granroth et al.

1978) (+earlier

papers by Karkinen-

Jääskeläinen

(Karkinen-

Jaaskelainen and

Saxen 1974) and

Saxen (Saxen et al.

1974)).

1978 (1965-

1973),

Finland

Finnish registry of

congenital



maternal reports,

retrospective to

delivery and maternity

centres, prospective to

delivery.

Live and

stillbirths.

Non-malformed

controls. Matched

by date of delivery

and Maternity

Welfare District.

Warrell (Warrell et

al. 1981)

1981 (1972-

1976), UK


fluorescent-antibody

technique.

It is

unknown

which types

of births

were

included.

Non-malformed

controls.

Matched (as far as

possible) by data of

blood serum

collection,

gestational age and

parity.

Aro (Aro 1983) 1983 (1964-

1977),

Finland

Finnish registry of

congenital


Exposure was mainly

based antenatal

records, collected


Live and

stillbirths.

Non-malformed

control.

Matched by date of

delivery and place.

Page 63 of 134




4

Lynberg (Lynberg et

al. 1994) (+ earlier

paper by Park (Park

et al. 1992))

1994 (1968-

1980), USA

Atlanta Birth Defects

Case Control Study

(ABDCCS). Exposure

was based on maternal

reported episodes of

>2 days of flu and fever

from 1 month before

to 3 months after

conception,

retrospective to

delivery.

Live and

stillbirths.

Non-malformed and

malformed controls

(with CA “outside

the hyperthermia

spectrum”).

Matched by hospital

of birth, race, year

and quarter of birth.

Adjusted for alcohol

use, education,

maternal age,

multivitamin use

and smoking.

Botto (Botto et al.

2001) (+ earlier

paper by Adams

(Adams et al. 1989))

2001 (1968-

1980), USA

Atlanta Birth Defects

Case Control Study

(ABDCCS). Exposure


reports, retrospective

to delivery.

Live and

stillbirths.

Non-malformed

controls.

Matched by

hospital, race and

time of birth.


use, chronic illness,

education, race,

multivitamin use,

period of birth and

smoking.

Page 64 of 134




5

Li(Li et al. 2007) 2006 (2003-

2005), China


maternal reports,

retrospective to

delivery.

Live and

stillbirths.

Non-malformed

controls.

Matched by county,

estimated date of

conception, ethnic

group and sex.

Adjusted for

antibiotics,

antipyretics, diet,

folic acid, history of

major congenital

anomaly affected

pregnancy, maternal

education and

infectious disease

other than influenza.

Czeizel (+ earlier

papers by Acs,

Medveczky and

Metneki) (Acs et al.

2005, Czeizel et al.

2008, Medveczky et

al. 2004, Metneki et

al. 2005) (+ later

paper by Csáky-

2007 (1980-

1996),

Hungary

Hungarian Case-

Control Surveillance of

Congenital Anomalies

(HCCSCA). Exposure


reports, retrospective

to delivery.

Live and

stillbirths.

Non-malformed

controls.

Matched by district

of residence, sex,

year and week of

birth.

Page 65 of 134




6

Szunyogh (Csaky-

Szunyogh et al.

2013))

Oster (Oster et al.

2011) (+ later paper

by Kelly (Kelly et al.

2012))

2011 (1981-

1989), USA

Baltimore-Washington

Infant Study (BWIS).


maternal reports,

retrospective to

delivery.

Live births. Random sample of

live-bortn controls

without congenital

heart disease.

Frequency matched

on age at interview,

month, year and

hospital of birth.


use, BMI, family

history of congenital

heart disease,

gestational diabetes,

race, sex and

smoking.

Results from these studies are included in the meta-analysis (Table I) and Table II. 2

Page 66 of 134




1

Table V: Overview of 12 papers from 10 cohort studies sorted by year. 1

1st

Author Publication

year (study

years) and

location

Recruitment Exposure ascertainment Matching

additional to

location and

adjustment (if

any)

Campbell

(Campbell

1953)

1953 (1950-

1951), UK

Mothers attending

antenatal clinics

were enrolled.


maternal reports,


None

Abramowitz

(Abramowitz

1958)

1958 (1957),

South Africa

Women attending

gynaecology wards

were enrolled.

Exposure ascertainment

was not described.

None

Walker (Walker

and McKee

1959)

1959 (1957),

USA

Ward patients who

delivered at

University hospitals

were enrolled.


hemagglutination

inhibition and maternal

reports, retrospective to

delivery.

None

Doll (Doll et al.

1960)

1960 (1957-

1958), UK

Mothers attending

an antenatal clinic

for the first time

during the study

period were

enrolled.


maternal reports,

prospective to delivery

and some of these were

confirmed with

professional diagnosis.

None

Pleydell 1960 (1957),

UK

Expecting mothers

reported to suffer


maternal reports,

None

Page 67 of 134




2

(Pleydell 1960) from ILI were

included and

exposed and non-

exposed were

recruited via local

midwives.


Saxen (Saxen et

al. 1960)

1960 (1958),

Finland

Women delivering

in the University

Central Hospital of

Helsinki were

enrolled.


maternal reports,


None

Hardy (Hardy

et al. 1961)

1961 (1957-

1958), USA

Patients visiting an

obstetrical prenatal

clinic were enrolled.


complement fixation,

hemagglutinin inhibition

and maternal reports,


None

Hirvensalo

(Hirvensalo and

Kinnunen 1962)

1962 (1957-

1959),

Finland

Women attending

the Maternity

Health Centres were

enrolled.


maternal reports,


None

Coffey (Coffey

and Jessop

1959, Coffey

and Jessop

1963)

1959 and

1963 follow-

up (1957-

1958), Ireland

Women attending

antenatal clinics

were enrolled.


maternal reports,


Stage of

pregnancy.

Page 68 of 134




3

Wilson (Wilson

et al. 1959,

Wilson and

Stein 1969)

1959 and

1969 follow-

up (1957),

USA

Prenatal registrants

were enrolled.


hemagglutinin inhibition.

None


Page 69 of 134




77x213mm (96 x 96 DPI)

Page 70 of 134




87x165mm (96 x 96 DPI)

Page 71 of 134




1

Manuscript title 1

Influenza and congenital anomalies: a systematic review and meta-analysis 2

Suggestion for a running title 3

Influenza and congenital anomalies 4

Authors full names 5

JM Luteijn. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 6


MJ Brown. Institute of Nursing Research/School of Nursing, University of Ulster, Jordanstown 8

Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 9

H Dolk. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 10


Page 72 of 134




2

Abstract 12

Study question: Does 1st

trimester maternal influenza infection increase the risk of non-13

chromosomal congenital anomalies? 14

Summary answer: 1st trimester maternal influenza exposure is associated with raised risk of a 15

number of non-chromosomal congenital anomalies, including neural tube defects, hydrocephaly, 16

congenital heart defects, , cleft lip, digestive system defects and limb reduction defects. 17

What is known already: Hyperthermia is a well-established risk factor for neural tube defects. 18

Previous studies suggest influenza may be a risk factor not only for neural tube defects, but also 19

other congenital anomalies. No systematic review has previously been undertaken. 20

Study design, size, duration: Systematic review and meta-analysis. A search of EMBASE and PUBMED 21

was performed for English and Dutch studies published up to July 2013. A total of 33 studies (15 22

case-control, 10 cohort and 8 ecological) were included in the systematic review of which 2122 23

studies were included in the meta-analysis. 24

Participants/materials, settings, methods: A total of 27,18129,542 babies with congenital anomaly 25

(9191,112 exposed) from case-control studies and 1,608 exposed pregnancies resulting in 56 babies 26

with congenital anomaly from cohort studies were included in the meta-analysis: 919 from case-27

control studies and 56 from cohort studies. Maternal influenza exposure was defined as any reported 28

influenza, influenza-like illness, or fever with flu, with or without serological or clinical confirmation 29

during the 1st

trimester of pregnancy. Data for 2024 (sub)groups of congenital anomaly available 30

from ≥ 3 studies were analysed using the DerSimonian-Laird random effects model. The hypothesis 31

of publication bias was assessed using funnel plots and risk of bias of included studies was assessed 32

using a slightly modified version of the Newcastle-Ottawa Scale. 33

Main results and the role of chance: 1st trimester maternal influenza exposure was associated with 34

an increased risk of any congenital anomaly (adjusted odds ratio 2.252.00, 95% CI: 1.77-2.851.62-35

Page 73 of 134




3

2.48), neural tube defects (OR 3.613.33, 2.23-5.822.05-5.40), hydrocephaly (5.775.74, 1.10-36

30.291.10-30.00), congenital heart defects (1.841.56, 1.50-2.261.13-2.14), aortic valve 37

atresia/stenosis (AOR 2.59, 1.21-5.54), ventricular septal defect (AOR 1.59, 1.24-2.14), cleft lip 38

(3.003.12, 2.12-4.252.20-4.42), digestive system (1.731.72, 1.11-2.691.09-2.68) and limb reduction 39

defects (2.062.03, 1.29-3.291.27-3.27). The increased risk for cleft lip (but not for cleft palate), was 40

also reported by ecological studies not included in the meta-analysis. Study outcomes reported for 41

31 27 subgroups of congenital anomaly could not be included in the meta-analysis. Visual inspection 42

of funnel plots did not suggest evidence for publication bias. 43

Limitations, reasons for caution: This study enrolled observational studies which can be subject to 44

limitations such as confounding, retrospective maternal exposure reports and non-response of 45

intended participants. Influenza exposed pregnancies can also have been exposed to influenza 46

related medication. 47

Wider implications of the findings: Prevention of influenza in pregnant women may reduce 48

congenital anomaly risk, and would be relevant to more than just neural tube defects. More research 49

is needed to determine whether influenza and/or its related medication is teratogenic, to determine 50

the role of hyperthermia in teratogenicity and the role of other environmental factors such as 51

nutritional status in determining susceptibility. 52

Study funding/ competing interests: Funded by the EC, under the framework of the EU Health 53

Programme 2008-2013, Grant Agreement 2010 22 04 (Executive Agency for Health & Consumers). No 54

competing interests. 55

Page 74 of 134




4

Introduction 56

Both during seasonal influenza and pandemic influenza outbreaks, pregnant women have been at 57

risk of increased morbidity and mortality from influenza infection compared to the general 58

population.(Dodds et al. 2007, Harris 1919, Neuzil et al. 1998, Siston et al. 2010) Women in the later 59

stages of pregnancy are particularly vulnerable to adverse health outcomes after influenza infection 60

(Mak et al. 2008), perhaps because of immunological changes that take place during 61

pregnancy.(Jamieson et al. 2006) 62

Unravelling the question of teratogenicity of influenza is complex. In observational studies, influenza 63

exposure can affect the foetus not only via viral infection of the foetus, influenza-induced 64

hyperthermia and toxic metabolites associated with fever (Edwards 2006), but also via antiviral and 65

antipyretic use. A recent systematic review found strong evidence of an association between 66

maternal hyperthermia and neural tube defects.(Moretti et al. 2005) Evidence on other anomalies 67

with respect to hyperthermia or fever is scarce. and aAnimal models have associated maternal 68

hyperthermia with arthrogryposis (Edwards 1971), congenital heart defects (Cockroft and New 1975, 69

Cockroft and New 1978), club foot(Edwards 1971), microcephaly (Edwards 1969, Edwards 1969), 70

microphthalmos (Germain et al. 1985) and others.(Edwards 2006) Evidence on other anomalies with 71

respect to hyperthermia or fever is scarce. 72

The primary method of protecting pregnant women and their unborn child against influenza 73

infection is vaccination. In an increasing number of countries, pregnant women are advised to be 74

vaccinated against seasonal influenza infection.(Mak et al. 2008, Mereckiene et al. 2010) Despite the 75

health benefits of vaccination by protecting pregnant women against influenza infectionHowever, 76

vaccination policies in European countries vary both for seasonal influenza vaccination and during 77

the 2009 H1N1 influenza pandemic, especially with respect to trimesters eligible for 78

vaccination.(Luteijn et al. 2011, Mereckiene et al. 2010) In the absence of consensus on whether or 79

not to vaccinate 1st trimester pregnant women, the hypothesis of a causal relationship between 80

Page 75 of 134




5

congenital anomalies (CA) and influenza virus deserves renewed attention. A better understanding of 81

the possible relationship between influenza and congenital anomaliesCA will allow for better 82

understanding of the benefit-risk balance of vaccinating 1st trimester pregnant women and women of 83

childbearing age against influenza. 84

The objective of our review is to identify and summarize the available epidemiologic evidence 85

regarding the risk of congenital anomalyCA associated with 1st trimester exposure to maternal 86

influenza. 87

Materials and methods 88

Search strategy 89

This systematic review was informed by PRISMA guidelines (Moher et al. 2009) and the MOOSE 90

group guidelines.(Stroup et al. 2000) Two of the authors (JML and MJB) conducted the various steps 91

of the review and resolved any disagreements by discussion and consensus. The PubMed® and 92

Embase® databases were searched using the MeSH terms ("Influenza, Human") AND (pregnancy OR 93

congenital abnormality) and (Influenza) AND (pregnancy OR congenital abnormality), respectively on 94

July 1st, 2013. No publication or date restrictions were set. Where the papers’ abstract, title or 95

indexed MeSH terms suggested the possibility of reporting any fetal outcomes after maternal 96

exposure to influenza, the full paper was obtained. Reference lists of enrolled papers were reviewed. 97

Eligibility criteria 98

Case-control, cohort and ecological studies investigating congenital anomalyCA outcomes following 99

maternal exposure to influenza were eligible for inclusion. No quality criteria were set for inclusion, 100

although risk of bias analysis was performed (eAppendix). Influenza was defined as any reported 101

influenza, influenza-like illness, or fever with flu, with or without serological or clinical confirmation. 102

Solely studies reporting influenza exposures during the 1st trimester of pregnancy were included in 103

the systematic review and meta-analysis. In order to be included in the meta-analysis, case-control 104

and cohort studies needed to allow for the calculation or ofreport odds ratios (OR) or relative risks 105

Page 76 of 134




6

(RR). For financial reasons, only English and Dutch language papers were eligible for inclusion. No 106

Dutch papers were included. 107

Data extraction 108

Study characteristics were extracted by JML and MJB (eAppendix, Table III-V). Data Crude and 109

adjusted ORs and RRs and 2x2 tables relevant for meta-analysis waswere extracted by JML and MJB 110

from cohort and case-control studies in 2x2 tables. We contacted three authors of studies published 111

≤10 years ago(Czeizel et al. 2008, Kelly et al. 2012, Oster et al. 2011) to obtain core unavailable data 112

in order to create aggregate groups such as orofacial clefts and for crude OR calculation and received 113

the complete dataset for one study.(Czeizel et al. 2008) In case of studies distinguishing between flu 114

with fever and flu without fever where it was impossible to combine the two, such as the stratified 115

study by Lynberg (Lynberg et al. 1994), the flu with fever dataset was extracted. We extracted data 116

regarding malformed controls, if available, rather than non-malformed controls since use of 117

malformed controls reduces the impact of differential recall bias. We recognize this can lead to 118

underestimation of effect size if CA in the control group are related to influenza exposure. In the 119

meta-analysis, influenza exposures outside of the 1st

trimester were added to the non-exposed 120

cohort and for one cohort study without controls (Doll et al. 1960), this allowed us to form a control 121

group. 122

Congenital anomaliesCA were classified into European surveillance of Congenital Anomalies 123

(EUROCAT) defined subgroups, excluding minor anomalies as specified by EUROCAT.(EUROCAT 124

Central Registry 2009) EUROCAT is a network of population-based congenital anomaly registries 125

which surveys over 1.7 million births annually in 23 European countries. CA were classified down to 126

the greatest level of precision possible. For example, a study that reported anomalies only as neural 127

tube defects without further specification contributed data to the analysis of neural tube defects but 128

could not contribute data to an analysis of anencephaly or spina bifida specifically. We excluded 129

chromosomal syndromes since their etiology is not related to maternal exposures. 130

Page 77 of 134




7


In order to assess the validity of included studies’ findings we assessed the risk of bias of case-control 132

and cohort studies included in the meta-analysis using a slightly modified version of the Newcastle-133

Ottawa scale (NOS).(Wells et al. 2004) Judgement criteria and deviations from the NOS are 134

summarized in the eAppendix. 135


Meta-analysis was performed combining adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for 137

confounders). Only four case-control studies provided AORs (Table IV) and none of the cohort studies 138

made statistical adjustments (Table V). Where adjusted estimates were not available, crude 139

estimates were used. 140

We assumed similarity between OR and RR because congenital anomaliesCA are rare events.(Davies 141

et al. 1998) Statistical analysis was performed using Stata version 9.2 [StataCorp, College Station, TX]. 142

Meta-analysis was performed on all anomalies CA combined and, for EUROCAT defined subgroups of 143

anomalies CA (if n studies ≥ 3). The DerSimonian & Laird random effects model (DerSimonian and 144

Laird 1986) was used since the studies in this meta-analysis involved varying countries, time periods 145

and influenza strains. Subgroup analysis was performed based on study type, publication date, and 146

risk of differential recall bias and adjustment for confounders in order to assess the impact of these 147

variables on study outcome. Subgroup analyses combined all studies in the relevant categories, using 148

the estimate for all non-chromosomal CA combined where available and if not available, the 149

estimate for the specific CA subgroup studied. 150

Due to scarce numbers and imbalance between some study arms, we used an alternative continuity 151

correction based on the OR of (other) studies with >0 events in both arms and group ratio imbalance 152

as discussed by Sweeting et al.(Sweeting et al. 2004) Heterogeneity between studies was assessed 153

using the I-squared statistic. Values of I-squared equal to 25%, 50% and 75% were considered to 154

Page 78 of 134




8

represent low, moderate and high levels of heterogeneity, respectively. The hypothesis of 155

publication bias was assessed using funnel plots (eAppendix, Figures 2428-4351). 156

Results 157

Selection flow 158

The PubMed® database search yielded 1369 papers and the Embase® database search yielded 2649 159

papers (Figure 1). After removing 1121 duplicates, a total of 2897 potentially relevant papers were 160

identified by the literature search. After screening by MeSH terms, titles and abstracts, 2615 papers 161

were excluded and full papers were retrieved for the remaining 282 papers. Of these, a total of 40 162

papers covering 27 studies met the inclusion criteria and were included in the systematic review. 6 163

additional eligible papers were detected by reference tracking, leading to a grand total of 46 included 164

papers covering 33 studies. 165

Study characteristics 166

The 46 enrolled papers were classified as 25 papers covering 15 case-control studies, 12 papers 167

covering 10 cohort studies and 9 papers covering 8 ecological studies. Enrolled studies are 168

summarized by study type in eAppendix, Table III-V and evidence provided by enrolled studies that 169

was not included in the meta-analysis is summarized in Table II. For one case-control study 170

information was limited to a conference abstract.(Choi and Klaponski 1970) Included papers were 171

published between 1953 and 2013, with the median year of publication 1971. 172


Visual inspection of the funnel plots did not suggest evidence for publication bias (eAppendix, 174

Figures 2428-4351). 175

Of the 15 case-control studies (25 papers), 10 studies did not take into account possible confounding 176

by maternal age, socioeconomic class or both. In the majority of these 10 studies some form of 177

matching between cases and controls (usually maternal ward, sex and day of birth) took place. Ten 178

papers relied on retrospective maternal reported influenza episodes (or timing of maternal 179

Page 79 of 134




9

interviews was unknown), making these studies susceptible to differential recall bias. Of the 180

remaining 5 studies, 2 used serologic confirmation and 3 used prospectively collected antenatal 181

records. The last notable source of possible bias was that in 5 case-control studies over 20% of the 182

cases intended for inclusion were not enrolled, making these studies susceptible to non-response 183

bias (eAppendix). 184

For the 10 cohort studies (12 papers), none took into account possible confounding by maternal age, 185

socioeconomic class or both. For 9 cohort studies, infants were not followed up for at least a year (or 186

unclear), making these studies susceptible to misclassification bias for some congenital anomaliesCA 187

not apparent at birth. 6 of the studies used prospectively collected maternal reports for exposure, 3 188

serologic confirmation, for 1 study exposure ascertainment was not described. The last notable 189

source of possible bias was that in 5 cohort studies the exposed cohort was not drawn from a clearly 190

defined place and time, or failed to enrol ≥80% of the population identified in specified place and 191

time, raising questions over representativeness of the enrolled exposed cohort. 192


Meta-analysis was possible for data from 21 22 studies, forming groups of ≥3 independent studies for 194

20 24 (sub)groups of EUROCAT defined major CA: any non-chromosomal major CA, neural tube 195

defects, anencephaly, encephalocele, spina bifida, hydrocephaly, congenital heart defects, orofacial 196

clefts (Figures 2-9) and twelve sixteen other CA subgroups (eAppendix, Figures 12-2327). 197

Overall, our meta-analysis involved 27,18129,542 CA cases of which 9191,112 were exposed to 198

influenza in the first trimester of pregnancy and 49,65453,089 controls of which 11211,382 were 199

exposed to influenza in the first trimester of pregnancy from case-control studies. From cohort 200

studies, 1,608 exposed pregnancies resulting in 56 CA plus 14,613 non-exposed pregnancies resulting 201

in 347 CA were enrolled. The enrolled cohort studies were relatively small with only the Coffey and 202

Jessop study reaching over 50 congenital anomaliesCA (including minor anomalies) following 203

maternal influenza exposure. Case-control studies more readily enrol the large number required for 204

Page 80 of 134




10

research on CA and 6 out of 15 case-control studies managed to enrolled over 500 cases.(Botto et al. 205

2001, Czeizel et al. 2008, Granroth et al. 1978, Laurence et al. 1968, Oster et al. 2011, Saxen 1975) 206

The larger numbers come at a cost and 11 out of 15 case-control studies gathered exposure data by 207

retrospective maternal reports. 208

Meta-analysis discovered statistically significant associations between 1st


exposure and a large number of CA subgroups (Table I) including all non-chromosomal CA combined 210

(OR 2.252.00, 95% CI: 1.77-2.851.62-2.48). Medium heterogeneity was detected for the pooled 211

estimate of all non-chromosomal CA (70.364%). Subgroup analysis showed similar outcomeslower 212

odds ratios for pooled case-control study outcomes (OR 2.251.84, 95% CI: 1.721.49-2.932.27), and 213

than for pooled cohort study outcomes (OR 2.12, 95% CI: 1.20-3.75) while pre-1970 studies reported 214

slightly higher odds ratios (OR 2.47, 95% CI: 1.50-4.70) than studies published after 1970 (OR 215

2.111.71, 95% CI: 1.62-2.761.41-2.08). Overall, studies susceptible to differential recall bias reported 216

a greater lower risk (OR 2.421.92, 95% CI: 1.71-3.411.35-2.72) than studies not susceptible to 217

differential recall bias (OR 2.12, 95% CI: 1.54-2.92). No differences were detected between pooled 218

adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77). 219

Central nervous system defects 220

Associations were detected found for all neural tube defects (OR 3.613.33, 2.23-5.822.05-5.40) and 221

the NTD neural tube defect subgroups anencephaly (OR 4.083.52, 2.14-7.081.69-7.32) and spina 222

bifida (OR 2.682.20, 1.79-4.021.48-3.28). The majority of the 2,500 neural tube defects were 223

reported by Czeizel (n=1,202, AOR 2.492.40, 1.361.30-4.574.40), Li (344, AOR 4.463.06, 3.061.40-224

6.506.67), Laurence (n=551, OR 3.93, 1.37-11.27) and Lynberg (331, AOR 1.721.70, 1.151.10-225

2.572.50). The lower OR reported by Lynberg is related to our preference for malformed controls 226

over healthy controls for the study by Lynberg, which lead to more conservative estimates. The study 227

by Lynberg reports higher AOR when using healthy controls for neural tube defects (AOR 3.43.0, 2.1-228

5.51.9-4.7). We discovered significant heterogeneity in the aggregate groups and NTDneural tube 229

Page 81 of 134




11

defects, anencephaly, encephalocele and hydrocephaly (Table I). The heterogeneity for neural tube 230

defects, anencephaly, encephalocele and hydrocephaly seems to be driven by the Coffey, Hirvensalo, 231

Pleydell, Saxen (1960) and Wilson studies which all contributed OR of >10 in at least one CA subgroup. 232

The ecological study on 1957 pandemic influenza by Hakosalo, enrolling 27 NTD neural tube defects 233

reported suggestive evidence for a relationship between influenza and NTD neural tube defects, but 234

two larger ecological studies by Leck (n=2,484 (Leck et al. 1969) and n=162 (Leck 1963)) did not find 235

such evidence for any neural tube defectsNTD subgroup. The study by Hakosalo calculated cases 236

back to last menstrual period, while none of the studies by Leck had access to gestational length, 237

making these studies susceptible to misclassification of exposure introduced by assumptions around 238

gestational age. Of all 8 ecological studies included, only three corrected for gestational age.(Busby 239

et al. 2005, Hakosalo and Saxen 1971, Saxen et al. 1990) 240

Orofacial clefts 241

Our meta-analysis detected found an association between orofacial clefts and 1st


exposure (OR 2.151.96, 95% CI: 1.711.33-2.702.91). There was a significant association for cleft lip 243

with or without palate (OR 3.003.12, 2.122.20-4.254.42), but not for cleft palate (OR 1.101.05, 244

0.630.60-.1921.84) and no heterogeneity was detected in these pooled groups (Table II). The 245

majority of the orofacial clefts (n=2,773) were reported by Czeizel (n=1,956) and Saxen (1975a n=591 246

and 1975b, n=194) and these studies report ORs between 1.90 and 2.32. Four of the ecological 247

studies, all by Leck, also reported the association between orofacial clefts and influenza and two of 248

these four studies reported associations for cleft lip ± cleft palate, but not for isolated cleft 249

palate.(Leck 1963, Leck et al. 1969) 250

Congenital heart defects 251

Our meta-analysis detected found an association between congenital heart defectsCHD and 1st

252

trimester influenza exposure (OR 1.841.56, 95% CI: 1.501.13-2.262.14), with very low heterogeneity. 253

The vast majority of the congenital heart defectsCHD reported in the meta-analysis were reported 254

Page 82 of 134




12

by Botto (n=829, AOR 2.1, 0.8-5.5), Czeizel (n=4,479, OR 1.6, 1.3-1.9) and Oster (n=2,361, AOR 1.11, 255

0.91-1.35). Data reported by Botto (AOR 1.8, 1.1-2.9) and Czeizel (OR 2.0, 1.3-3.2) were suggestive of 256

an association between ventricular septal defects and influenza. It should be noted the study by 257

Botto suffered from a 2-12 year delay between delivery and maternal interview. These findings are 258

contradicted by the Baltimore-Washington Infant Study (BWIS), for which we were unable to obtain 259

core data in 2x2 tables (and therefore could not be included in the meta-analysis, Table II). The BWIS 260

did not find an association with CHD (AOR 1.11, 0.91-1.35), nor for VSD (AOR 1.30, 0.92-1.83). while 261

The BWIS did detect associations for tricuspid atresia (AOR 6.04, 2.36-15.42) and pulmonary valve 262

atresia (AOR 2.71, 1.16-6.32). It should be noted influenza exposure in the BWIS studythe study by 263

Oster could occur from 3 months before pregnancy to the 3rd

month of pregnancy. 264

With respect to specific types of congenital heart defects (Table I), meta-analysis showed aortic valve 265

atresia/stenosis and ventricular septal defect to be associated with 1st


(OR 2.59, 1.21-5.54 and OR 1.59, 1.24-2.04, respectively). No associations were found for atrial septal 267

defect, hypoplastic left heart and transposition of the great vessels. Four congenital heart defect 268

subtypes were only reported by ≤2 studies and therefore not in the meta-analysis (Table II). Of these, 269

two studies showed a consistent absence of association for tetralogy of Fallot, and a consistent and 270

high association for tricuspid atresia and stenosis (Botto 7.9, Oster 6.04). 271

Other anomalies 272

Limb reductions were associated with 1st

trimester influenza exposure by the meta-analysis (OR 273

2.062.03, 1.291.27-3.293.27) and this association is supported by several ecological studies (Table II). 274

An association with anophthalmia/microphthalmia is based on a single study (Table II). There was 275

evidence that there is no association with influenza for hypospadias (OR 0.951.02, 0.700.75-1.281.39) 276

and clubfoot (OR 1.03, 0.83-1.27). 277

Page 83 of 134




13

Discussion 278

This systematic review provides an overview of the published evidence on influenza exposure during 279

the 1st

trimester of pregnancy and congenital anomaliesCA. Meta-analysis revealed evidence for 280

increases in a wide range of major CA following 1st trimester influenza exposure. The twofold 281

increase in risk of non-chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT 282

Central Registry 2013) to 3.6% of births among 1st

trimester influenza exposed pregnancies. 283

Exposure ascertainment 284

Case-control and cohort studies utilized serologic confirmation and maternal reports (prospective 285

and retrospective) for influenza exposure ascertainment. During influenza season the positive 286

predictive value of persons presenting with ILI for influenza is in the order of 66-77%.(Monto et al. 287

2000, Zambon et al. 2001) Serologic confirmation detects clinical and subclinical infections, which 288

might lead to different results since subclinical infections might affect the pregnant women 289

differently from clinical infections. Arguably, serologic confirmation of exposure is more reliable than 290

maternal reports. Five studies based on serologic confirmation were enrolled in the systematic 291

review (Elizan et al. 1969, Hardy et al. 1961, Walker and McKee 1959, Warrell et al. 1981, Wilson et al. 292

1959, Wilson and Stein 1969) of which three were included in the meta-analysis.(Hardy et al. 1961, 293

Warrell et al. 1981, Wilson et al. 1959, Wilson and Stein 1969) The two serologic studies limited to 294

systematic review did not detect an association between influenza and neural tube defectsNTD 295

(Elizan et al. 1969) and did not detect an increased prevalence of CA among 1957 H2N2 Asian 296

pandemic influenza exposed pregnancies.(Walker and McKee 1959) The three serologic studies in the 297

meta-analysis combined contributed 53 out of 27,584 CA, and it was therefore not possible to 298

examine the effect of exposure ascertainment method on effect size. One of these studies reported a 299

possible association between CA and 1957 H2N2 Asian pandemic influenza (Hardy et al. 1961) while 300

a second study did not.(Wilson et al. 1959, Wilson and Stein 1969) The third study did not detect an 301

association between CA and neural tube defectsNTD.(Warrell et al. 1981) Note that all studies using 302

serologic confirmation were limited to low numbers. 303

Page 84 of 134




14

It is apparent that maternal reports lead to misclassification of exposure as women might not recall 304

infection, timing of infection relative to pregnancy or misdiagnose another infection for influenza. 305

However, as long as maternal reports are collected prospectivelyprospective to the mother being 306

aware of the malformation (e.g. from medical records or interviews during pregnancy), there is no 307

reason to believe misclassification of influenza exposure will differ between armscases and controls. , 308

and tTherefore prospective maternal reports will not lead to a spurious association, but rather bias 309

the estimate toward the null as cases and controls are subject to similar misclassification. 310

Retrospective maternal reports (e.g. interviews after birth), which are frequently utilized by case-311

control studies, are susceptible to recall bias where mothers of cases have a different motivation to 312

recall early pregnancy exposure than mothers of non-cases. Mothers may differ not only in their 313

tendency to remember the infection, but in their tendency to misinterpret an illness as 314

influenza.(MacKenzie and Houghton 1974) Some of the studies included in the systematic review give 315

an estimate of the differential recall bias’ effect. For example, in the study by Lynberg, the OR for 316

anencephaly after flu with fever decreased from 3.1 (95% CI: 1.6-6.1) when compared with non-317

malformed controls to 1.4 (95% CI: 0.7-2.6) when compared with malformed controls.(Lynberg et al. 318

1994) Part of this decrease might also have been related to CA in the control group being related to 319

influenza, thus biasing the OR towards 1. In our meta-analysis, the pooled estimate of OR for studies 320

susceptible to recall bias (2.421.92) was slightly higher lower than the pooled estimate for other 321

studies (2.122.00), contrary to expectation. This is related to the low overall OR (1.11) reported by 322

the large Baltimore-Washington Infant Study (Oster et al. 2011), which was susceptible to differential 323

recall bias and contributed over 20% to the pooled estimate of susceptible studies. 324

Case-control, cohort and ecological study designs 325

Cohort studies failed to enrol large numbers of CA and this should not be surprising considering CA 326

only make up for 2-3% of births in a general population and short follow-up time after birth could 327

have led to underascertainment. Investigation of specific CA requires even higher numbers. Due to 328

the possibility of enrolling larger numbers, case-control studies seem better suited for addressing the 329

Page 85 of 134




15

hypothesis of teratogenicity of influenza, although this comes at a cost as most case-control studies 330

ascertained exposure by retrospective maternal reports. 331

Ecological studies are generally considered a weaker study design than case-control or cohort studies 332

(Evans 2003) due to lack of individual exposure information. It cannot be verified that theany excess 333

congenital anomaliesCA occurred among infected individuals. Correlation between influenza and 334

confounding risk factors at group level may lead to ecological fallacy, for example if influenza and 335

nutritional deficiencies co-occur in winter. Ecological studies base exposure status on timing of 336

pregnancy relative to influenza season (or a proxy thereof) and therefore, the cohort defined as 337

“exposed” is diluted by pregnancies that did not have influenza. Population influenza exposure in the 338

eight enrolled ecological studies were derived from influenza incidences, counts or deaths (n=5) and 339

sickness absenteeism rates or claims (n=3). For this reason, the distinguishing power of ecological 340

studies is highly dependent on influenza attack rates and precision used to define influenza 341

seasonexposure. 342

The advantages of ecological studies are that large numbers of patients are enrolled easily and 343

ecological studies are not susceptible to the exposure misclassification or recall bias inherent in 344

individual level studies. Furthermore, they can be free of individual level confounding e.g. if those 345

most susceptible to influenza in the population have other risk factors for congenital anomaliesCA. 346

Ecological studies therefore offer great value for addressing the hypothesis of teratogenicity of 347

infectious diseases and should not be discounted at the bottom of the evidence hierarchy in this area 348

of research. Consistency between study designs lends strength to a causal interpretation.(Hofler 349

2005) 350

Associations between 1st trimester influenza exposure and CA 351

One of the most striking results of the meta-analysis is the association between 1st

trimester 352

influenza exposure and neural tube defectsNTD. NTD Neural tube defects are easily recognized at 353

birth eliminating susceptibility to underascertainment, while a previous meta-analysis reported an 354

Page 86 of 134




16

association between neural tube defectsNTD and hyperthermia (OR 1.92, 95% CI: 1.62-355

2.29).(Moretti et al. 2005) There also is evidence for neural tube defectsNTD following hyperthermia 356

exposure in guinea pigs.(Smith et al. 1992) In more recent human studies, underascertainment of 357

neural tube defectsNTD could have been a problem due to terminations of pregnancy for fetal 358

anomaly. According to a 2004 EUROCAT analysis, 88% of neural tube defectsNTD are detected 359

prenatally of which 88% are aborted.(Boyd et al. 2008) Neural tube defectsNTD were more 360

frequently studied than other congenital anomaliesCA, possibly a result of interest in the 1950s 361

studies by Coffey, suggesting an alarmingly strong link between maternal influenza exposure and 362

neural tube defects NTD (data corresponds to OR 10.58, 4.30-26.02 in the 1963 follow-up).(Coffey 363

and Jessop 1963) This study utilized standardized questionnaires for maternal interview after 364

delivery and therefore was susceptible to differential recall bias, but this limitation would not explain 365

such a high OR. One of the ecological studies found an increase in neural tube defectsNTD during the 366

1957 Asian influenza outbreak in Finland and concluded this might have been caused either by 367

influenza or influenza-related pharmaceuticals.(Hakosalo and Saxen 1971) The study by Li et al had 368

data available both on antiviral and antipyretic use (and other potential confounders, see Table IV) 369

and after adjustmenting for these co-exposures, OR for neural tube defectsNTD following maternal 370

influenza exposure dropped slightly but remained statistically significant (OR 3.93, 95% CI: 2.48-6.23). 371

Hydrocephaly can sometimes be caused by spina bifida, and we could derive estimates for 372

hydrocephaly not associated with neural tube defectsNTD for two studies of the five studies on 373

hydrocephaly.(Coffey and Jessop 1959, Coffey and Jessop 1963, Hirvensalo and Kinnunen 1962) It is a 374

general problem in congenital anomalyCA research that one baby may have more than one anomaly, 375

and some of these multiple malformed babies may have “sequences” (EUROCAT Central Registry 376

2009) that follow from a primary anomaly. The data available for review cannot distinguish different 377

types of diagnoses. 378

Page 87 of 134




17

The evidence for an association between 1st

trimester influenza exposure and cleft lip with or 379

without palate is strong due to the lack of heterogeneity in the pooled data and consistent positive 380

associations detected between across different study designs. It is well known that cleft lip +/- palate 381

differs aetiologically from cleft palate, so this difference is not surprising.(Mossey et al. 2009) The 382

associations detected for congenital heart defectsCHD and VSD ventricular septal defects are 383

somewhat contradicted by the results from the BWIS.(Oster et al. 2011) On the other hand 384

congenital heart defectsCHD have been associated with hyperthermia in rats.(Cockroft and New 385

1975, Cockroft and New 1978) Club foot has been associated with hyperthermia in guinea pigs 386

(Edwards 1971), but our meta-analysis did not find evidence for an association between club foot 387

and influenza. A large number of additional associations for other congenital anomalyCA types were 388

detected with more limited underlying evidence. 389

Pathways for mediation of hypothetical teratogenic effect of influenza 390

Influenza can mediate a possible teratogenic effect via multiple pathways and there is a risk of 391

confounding due to the intimate linkage between a disease and its cure. As well as antivirals, 392

antipyretics are also often used during influenza infection and the case-control study by Li et al 393

reported an AOR for antipyretic drugs and neural tube defectsNTD of 4.86 (95% CI: 1.33-17.78).(Li et 394

al. 2007) Associations not adjusted for antivirals or antipyretics remain of importance since from a 395

vaccination policy perspective, it is less relevant whether the influenza virus or the antivirals are 396

causing any possible anomalies as vaccination will prevent both exposures. We recognize this puts 397

limits on generalizability of study findings between populations with different use of 398

antipyretics/antivirals. 399

Direct pathways via which influenza infection can possibly lead to CA are toxic metabolites caused by 400

fever, hyperthermia and the influenza virus crossing the placenta. Hyperthermia has been associated 401

with causing neural tube defectsNTD as discussed above.(Moretti et al. 2005) It should be noted this 402

meta-analysis involved a large number of possible causes of hyperthermia, while influenza causes 403

Page 88 of 134




18

high fever which might be different from general hyperthermia. This could explain the higher OR 404

reported in this meta-analysis for influenza exposure and neural tube defects NTD (Table I). Several 405

of the included studies distinguished between 1st

trimester influenza and 1st

trimester fever.(Aro 406

1983, Botto et al. 2001, Klemetti 1977, Lynberg et al. 1994, Oster et al. 2011, Saxen 1975, Saxen 1975) 407

One study found an association for influenza, but not for fever (Klemetti 1977), another found an 408

association for influenza with fever and neural tube defects NTD (OR 1.7, 1.1-2.5), which was lowered 409

for influenza without fever (OR 1.3, 0.7-2.5).(Lynberg et al. 1994) A study on congenital heart 410

defectsCHD reported associations for fever (OR 1.8, 1.4-2.4) and influenza (OR 2.1, 0.8-5.5).(Botto et 411

al. 2001) while a second study on CHDcongenital heart defects reported very similar low and non-412

significant excesses for fever (OR 1.14, 0.89-1.46) and influenza (OR 1.11, 0.91-1.35).(Oster et al. 413

2011) The two studies by Saxen on orofacial clefts reported comparable associations for influenza 414

(RR 2.00 for the 1st study) and fever (RR 1.96 for the 1

st study) (Saxen 1975, Saxen 1975), and the 415

study by Aro on limb reduction defects reported an OR of 1.6 for fever and 1.9 for influenza.(Aro 416

1983) It can be concluded that included studies generally reported equivalent or stronger 417

associations for influenza than for fever with respect to CA following 1st

trimester exposure. 418

Another possible pathway by which the influenza virus can mediate a teratogenic effect is placental 419

transmission. Placental transmission has been documented but appears to be rare.(Gu et al. 2007, 420

McGregor et al. 1984) Toxic metabolites associated with fever as a cause of congenital anomalies 421

have also been suggested.(Edwards 2006) 422

Limitations of the study 423

The systematic review results should be interpreted in the light of the findings that all of the included 424

studies are observational studies and many were susceptible to several types of bias and 425

ascertainment of exposure might not have been reliable. Although the funnel plots did not provide 426

evidence for publication bias, it should be noted that most of the included studies reported a wide 427

range of positive associations raising the question of whether studies reporting negative results 428

Page 89 of 134




19

remained unpublished. Reference tracking identified 6 new studies and these studies were missed 429

because they were not indexed as influenza studies. This leaves the possibility open that some, 430

particularly negative, studies were missed due to poorly indexed terms. A possible reason for this is 431

that some case-control studies investigate a wide range of possible causes of CA and may tend to be 432

selectively indexed for the positive associations. For older studies, we could not be sure whether 433

chromosomal CA were excluded from analysis. 434

A weakness of the meta-analysis was that adjustment for confounders was not possible since some 435

studies did not adjust for confounders and confounder data could not be extracted from papers that 436

did adjustnot performed in most included studies. Adjustment for confounders showed a moderate 437

effect on OR within studies that did report both crude OR and adjusted OR.(Acs et al. 2005, Granroth 438

et al. 1978, Li et al. 2007) Subgroup analysis comparing adjusted versus crude estimates did not 439

detect differences, but this could be related to the fact that 50% of the adjusted estimate was 440

composed of neural tube defects data since studies generally reported higher estimates for neural 441

tube defects than for other CA. Due to the limited amount of studies reporting adjusted OR, 442

comparison of crude and adjusted OR for subgroups of the same CA was not possible. Some very 443

large datasets involved matched and/or stratified controls (Botto et al. 2001, Czeizel et al. 2008), and 444

most other datasets were matched by one or more variables (Aro 1983, Granroth 1978, Granroth et 445

al. 1978, Karkinen-Jaaskelainen and Saxen 1974, Laurence et al. 1968, Li et al. 2007, Lynberg et al. 446

1994, Saxen 1975, Saxen 1975, Warrell et al. 1981) lowering the impact of confounding on the meta-447

analysis. 448

Statistical limitations 449

The study was susceptible to statistical limitations unique to meta-analysis of scarce events. In the 450

context of scarce events like CA and complicated exposure like 1st

trimester influenza, cohort studies 451

can enrol large numbers of exposed and unexposed, but have very few exposed congenital 452

anomalyCA outcomes. Current statistical methods will assign these studies a lot of weight compared 453

Page 90 of 134




20

to case-control studies with many cases and greater numbers of exposed cases. An example from this 454

systematic review is the cohort study by Pleydell, which reported 1 case of hydrocephaly among 12 455

1st trimester influenza exposed pregnancies and 1 case among 1071 unexposed pregnancies leading 456

to an OR of 97.27. The heterogeneity model favours outliers and smaller studies and provided this 457

study with 20% weight in the overall hydrocephaly estimate, compared to 43% weight for the case-458

control study by Czeizel which enrolled 314 cases of hydrocephaly (16 exposed). The weight allocated 459

by the heterogeneity model to the Pleydell study is clearly disproportionate. 460

A second problem lies in the continuity correction which is used to address zero events in one of 461

both arms of a study when pooling odds ratios. For rare events like CA, zero events in one or both 462

arms are not uncommon. The standard value for continuity correction is 0.5, but this arbitrary value 463

causes problems in studies with uneven arms and can even dominate the other arm. We addressed 464

this problem as proposed by Sweeting (Sweeting et al. 2004) by letting the continuity correction 465

depend on the OR of (other) studies with >0 events in both arms and group ratio imbalance. 466

However, this still led to OR > 1 for studies reporting 0 events in the exposed arm such as the 467

hydrocephaly data by Hirvensalo. 468

Conclusions and implications for CA prevention 469

Given the risk of congenital anomaly associated with influenza we show here, prevention of influenza 470

by vaccinating women who are planning to get pregnant may reduce congenital anomaly risk. 471

However, before evidence based policy can be implemented, further safety data on use of influenza 472

vaccines in pregnancy with respect to congenital anomalies is required (Kallen and Olausson 2012, 473

Pasternak et al. 2012). Other methods for preventing influenza in early pregnancy include improving 474

nutritional and general health status and adopting behaviours which prevent interpersonal spread. 475

In conclusion, prevention of influenza in pregnant women may reduce congenital anomaly risk, and 476

would be relevant to more than just neural tube defects. More research is needed to determine 477

whether influenza and/or its related medication is teratogenic, to determine the role of 478

Page 91 of 134




21

hyperthermia in teratogenicity and the role of other environmental factors such as nutritional status 479

in determining susceptibility. 480

Prevention of 1st

trimester pregnant women and women planning to get pregnant in influenza season 481

can possibly prevent occurrence of CA, most notably NTD and cleft lip, induced either by influenza or 482

related medication. Before evidence based policy can be implemented, safety data on use of 483

influenza vaccines in pregnancy with respect to congenital anomalies is required and as argued by 484

Parisi et al a new attitude towards research, medication and pregnancy might be required. 485

Declaration of authors roles 486

J.M.L. (corresponding author, first author) was involved in study design, data collection and synthesis, 487

manuscript preparation and revision, construction of tables and figures, statistical analysis and 488

submission of the manuscript. M.J.B. (co-author) was involved in data collection and manuscript 489

revision. H.D. (co-author) was involved in study design, writing and revising the manuscript. 490

Acknowledgements 491

We would like to thank Dr. I. Barisic for assistance with case classification and we would like to thank 492

Dr. Julia Métneki, Dr. Erzsébet Puhó, Professor Andrew Czeizel and Dr. Annukka Ritvanen for sending 493

additional data. We would like to thank Professor Lolkje de Jong-van den Berg enand Dr. Gordon 494

Marnoch for their helpful comments. 495

Funding 496

Funded by the EC, under the framework of the EU Health Programme 2008-2013, Grant Agreement 497

2010 22 04 (Executive Agency for Health & Consumers). 498

Conflict of interest 499

No competing interests 500

Page 92 of 134




22

Table I: Heterogeneity and number of cases for 1st

First trimester maternal influenza exposure and 501

risk of congenital anomalies.: studies, total number of cases of congenital anomaly, pooled odds ratio 502

and heterogeneity. 503

Group Participating

studies (n)

I2

Statistic for

heterogeneity (%)

Pooled OR (95%

CI)

Total number

of CA (n)

Any congenital anomaly 2122 7064 2.25 (1.77-

2.85)2.00 (1.62-

2.48)

27,58429,945bc

-Susceptible to


89 6565 2.421.92

(1.711.35-

3.412.72)

3,0655,426

-Not susceptible to


13 64 2.12 (1.54-2.91) 24,519bc

-Case-control studies 1213 7560 2.251.84

(1.721.49-

2.932.27)

27,18129,542bc

-Cohort studies 9 62 2.12 (1.21-3.72) 403

-Any type, published

1955-1969

11 58 2.47 (1.50-4.70) 1,171

-Any type, Published

19745-200711

1011 7655 2.111.71

(1.621.41-

2.762.08)

26,41328,774

-Adjusted estimates only a 4 87 2.15 (1.05-4.42) 3,865

-Crude estimates only a 21 61 2.22 (1.78-2.77) 27,584

Neural Tube Defects 11 6050 3.613.33

(2.232.05-

2,500

Page 93 of 134




23

5.825.40)

-Anencephaly 10 4744 4.083.52

(2.141.69-

7.807.32)

608

-Encephalocele 4 6463 3.022.95

(0.790.78-

11.4811.13)

225

-Spina Bifida 7 190 2.682.20

(1.791.48-

4.023.28)

1,093

Hydrocephaly 5 45 5.775.74 (1.10-

30.2930.00)

323

Congenital heart

defectsCHD

910 541 1.841.56

(1.501.13-

2.262.14)

5,3547,715

-Aortic Valve

Atresia/Stenosis

3 31 2.59 (1.21-5.54) 167

-Atrial Septal Defect 3 0 0.82 (0.45-1.51) 429

-Hypoplastic Left Heart 3 0 1.58 (0.94-2.64) 203

-Transposition of the

Great Vessels

3 0 1.40 (0.90-2.17) 321

-Ventricular Septal Defect 34 0 2.061.59

(1.451.24-

2.922.04)

9551,434

Orofacial Clefts 10 037 2.151.96

(1.711.33-

2,773cd

Page 94 of 134




24

2.702.91)

-Cleft Lip +- Palate ab

7 0 3.003.12

(2.122.20-

4.254.42)

1,404

-Cleft Palate ab

3 0 1.101.05

(0.630.60-

1.921.84)

584

Digestive System ab

4 0 1.731.71

(1.111.09-2.69)

1,195

Urinary 5 30 1.441.45

(0.910.90-

2.292.34)

48

Hypospadias ab

4 0 0.951.02

(0.700.75-

1.281.39)

3,041

Limb Reduction 3 0 2.062.03

(1.291.27-

3.293.27)

1,002

Club Foot ab

4 0 1.031.11

(0.830.93-

1.271.34)

2,430

Hip Dislocation/Dysplasia 3 0 0.300.31 (0.00-

37.0837.62)

37

Polydactyly ab

4 0 1.681.72

(0.840.85-

3.383.48)

1,094

Page 95 of 134




25

a Three studies (Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994) were able to provide estimates 504

both for adjusted and crude OR.

505

ab Note that a single study contributed over 90% of the total weight to this pooled estimate

506

bc Note that for some studies such as Czeizel et al 2008, the “any congenital anomaly” group also 507

included CA which were not included in any other analysis.

508

cd Note that 591 orofacial clefts from the Saxen 1975a study (Saxen 1975) and 194 orofacial clefts 509

from the Saxen 1975b study (Saxen 1975) were not specified and therefore solely included in the 510

overall orofacial clefts analysis. 511

Syndactyly 3 6871 1.961.98

(0.180.19-

21.0320.563)

662

Musculo-Skeletal 3 0 1.031.05

(0.130.16-

8.106.97)

776

Page 96 of 134




26

Table II: Evidence included in the systematic review relating to 1st trimester influenza exposure and 512

CA not included in meta-analysis. 513

Anomaly Evidence reported with respect to 1st trimester influenza

exposure

Any anomaly Visual inspection: no convincing evidence that infant death rates

from CA are associated with 1st trimester influenza

exposure(Buck 1955); RR 1.10(Leck 1963); single defect RR 1.00,

multiple defects RR 0.9(Leck 1964); single defect RR 1.03,

multiple defects RR 1.07(Leck et al. 1969)

CNS Central Nervous System

defects

0.81% in influenza exposed group vs. 0.30% and 0.32% in control

groups(Hakosalo and Saxen 1971)

NTDNeural Tube Defects p=0.03, no data(Choi and Klaponski 1970)

Anencephaly RR 0.82(Leck 1963); sRR 1.32(Leck et al. 1969); RR 1.0, 0.8-

1.3(Saxen et al. 1990) No rise in anencephaly rates following

influenza epidemics detected(Record 1961)

Spina Bifida + Encephalocele RR 0.99(Leck 1963), sRR 1.13(Leck et al. 1969)

Hydrocephaly sRR 0.93(Leck et al. 1969)

Microcephaly OR 1.1, 0.3-4.1(Czeizel et al. 2008); sRR 0.24(Leck et al. 1969)

Anophthalmos/Microphthalmos OR 1.26, 1.02-1.57(Busby et al. 2005)

Congenital Cataract OR 0.8, 0.3-2.2(Czeizel et al. 2008)

Congenital Glucaoma OR 2.0, 0.4-10.9(Czeizel et al. 2008)

Congenital heart defectsCHD (suggestive but not significant; 0.93% in influenza exposed group

vs. 0.47% and 0.73% in control groups)(Hakosalo and Saxen

1971); AOR 1.11, 0.91-1.35(Oster et al. 2011)

Transposition of great vessels AOR 2.1, 0.8-5.5(Botto et al. 2001); OR 2.2, 0.3-15.8(Czeizel et al.

2008); AOR 1.21, 0.73-2.02(Oster et al. 2011)

Page 97 of 134




27

Ventricular Septal Defect AOR 2.0, 1.1-3.6(Botto et al. 2001); AOR 1.30, 0.92-1.83(Oster et

al. 2011)

Atrial Septal Defect OR 1.0, 0.1-7.4(Botto et al. 2001); OR 3.0, 0.3-35.8(Czeizel et al.

2008); AOR 0.73, 0.38-1.42(Oster et al. 2011)

Atrioventricular Septal Defect OR 2.0, 0.3-15.3(Botto et al. 2001); AOR 1.29, 0.82-2.01(Oster et

al. 2011)

Tetralogy of Fallot OR 0.5, 0.1-3.6(Botto et al. 2001); AOR 0.78, 0.37-1.62(Oster et

al. 2011)

Triscuspid atresia and stenosis AOR 7.9, 0.3-29.6(Botto et al. 2001); AOR 6.04, 2.36-15.42(Oster

et al. 2011)

Ebstein's anomaly OR 3.0, 0.4-23.9(Botto et al. 2001)

Pulmonary valve stenosis AOR 1.21, 0.71-2.04(Oster et al. 2011)

Pulmonary valve atresia AOR 2.71, 1.16-6.32(Oster et al. 2011)

Aortic valve atresia/stenosis OR 4.0, 0.9-17.9(Botto et al. 2001); OR 4.6, 1.4-14.9(Czeizel et al.

2008); AOR 1.53, 0.70-3.34(Oster et al. 2011)

Hypoplastic left heart OR 1.6, 0.4-6.7(Botto et al. 2001);OR 1.6, 0.4-6.1(Czeizel et al.

2008); AOR 1.57, 0.86-2.88(Oster et al. 2011)

Coarctation of aorta AOR 3.8, 1.6-8.8(Botto et al. 2001); AOR 0.41, 0.13-1.33(Oster et

al. 2011)

Total anomalous pulm venous OR 2.2, 0.3-16.9(Botto et al. 2001)

Cleft Lip RR 1.55, p<0.05(Leck 1963); isolated RR 1.4, with other defects

RR 6.3(Leck 1964); cleft lip without cleft palate sRR 1.47,

p<0.05(Leck et al. 1969); cleft lip without cleft palate RR 1.64,

p<0.05(Leck 1971); cleft lip without cleft palate sRR 1.08, cleft lip

with cleft palate sRR 1.01(Leck et al. 1969); cleft lip with cleft

palate sRR 1.10(Leck et al. 1969), cleft lip with cleft palate RR

Page 98 of 134




28

1.35(Leck 1971)

Cleft Palate RR 0.81(Leck 1963); sRR 0.94(Leck et al. 1969); sRR 0.93(Leck et

al. 1969)

Digestive System 1 in 63 exposed during 1st


non-exposed, RR 4.4(Coffey and Jessop 1959, Coffey and Jessop

1963); 1 in 171 exposed during 1st

trimester of pregnancy and 13

in 6720 non-exposed, RR 3.0(Hirvensalo and Kinnunen 1962)

Oesophageal atresia

with/without

tracheo-oesophageal fistula

OR 1.6, 0.2-11.7(Czeizel et al. 2008); RR 2.41, p<0.01(Leck 1963);


1.12(Leck et al. 1969)

Atresia/Stenosis of the small

intestine

OR 2.1, 0.5-8.0(Czeizel et al. 2008)

Ano-rectal atresia and stenosis OR 1.1, 0.4-3.0(Czeizel et al. 2008); RR 2.12, p<0.05(Leck 1963);


0.80(Leck et al. 1969)

Hirschsprung's disease OR 0.7, 0.2-3.0(Czeizel et al. 2008)

Diaphragmatic Hernia OR 3.2, 1.2-8.8(Czeizel et al. 2008); 0 in 171 exposed during 1st

trimester of pregnancy and 3 in 6720 non-exposed(Hirvensalo

and Kinnunen 1962); RR 1.05(Leck 1963); sRR 1.01(Leck et al.

1969)

Abdominal Wall Defects OR 2.8, 1.1-6.9(Czeizel et al. 2008)

Omphalocele RR 1.92, <0.05(Leck 1963); isolated RR 1.3, with other defects RR

2.3(Leck 1964); sRR 1.16(Leck et al. 1969)

Urinary 1 in 171 exposed during 1st

trimester of pregnancy and 42 in

6720 non-exposed(Hirvensalo and Kinnunen 1962)

Bilateral renal agenesis including OR 0.8, 0.2-2.8(Czeizel et al. 2008); RR 1.38(Leck 1963)

Page 99 of 134




29

Potter syndrome

Congenital Hydronephrosis RR 1.63, p=0.05(Leck 1963)

Hypospadias RR 1.51(Leck 1963); sRR 0.93(Leck et al. 1969)

Limb sRR 0.75(Leck et al. 1969)

Limb Reduction Limited to thumbs or radii RR 2.20(Leck 1963); both arms or legs

RR 1.30(Leck 1963); one arm or leg RR 1.37(Leck 1963); sRR

1.32(Leck et al. 1969); RR 1.22(Leck 1971)

Upper Limb Reductions sRR 1.91, p<0.01(Leck et al. 1969)

Lower Limb Reductions sRR 1.23(Leck et al. 1969)

Club Foot sRR 0.89(Leck et al. 1969)

Hip Dislocation and/or Dysplasia sRR 0.77(Leck et al. 1969)

Disorders of skin 0 in 171 exposed during 1st


non-exposed(Hirvensalo and Kinnunen 1962)

Down’s Syndrome sRR 1.16(Leck et al. 1969)

Solely data involving >2 CA cases was included and some reported CA could not be translated to 514

EUROCAT defined subgroups. We did not distinguish between “positive” and “negative” findings 515

since some of the studies were severely underpowered. Note for some CA, ≥3 estimates were 516

available from studies that could not be included in the meta-analysis. 517

Page 100 of 134




30

Figure 1: Flow chart of systematic review 518

[Figure 1] 519

Figure 2: Forest plot of non-chromosomal CA following 1st


[Figure 2] 521

Figure 3: Forest plot of neural tube defects following 1st trimester influenza exposure 522

[Figure 3] 523

Figure 4: Forest plot of anencephaly following 1st


[Figure 4] 525

Figure 5: Forest plot of encephalocele following 1st


[Figure 5] 527

Figure 6: Forest plot of spina bifida following 1st


[Figure 6] 529

Figure 7: Forest plot of hydrocephaly following 1st


[Figure 7] 531

Figure 8: Forest plot of congenital heart defects following 1st


[Figure 8] 533

Figure 9: Forest plot of orofacial clefts following 1st


[Figure 9] 535

Page 101 of 134




31

eAppendix 536

Quality assessment of case-control and cohort studies enrolled in meta-analysis 537

For quality assessment of studies included in the meta-analysis, we used a slightly modified version 538

of the Newcastle-Ottawa scale (NOS).(Wells et al. 2004) We did not distinguish between community-539

based or hospital-based controls since most studies would recruit from antenatal clinics it is safe to 540

assume that this population is similar to a community-recruited population. In addition, the risk of 541

referrals would not have been present in studies from before the late 1980s before prenatal 542

screening was widely available. There is some risk of bias introduced from high risk pregnancies 543

being more likely to book at specific hospitals. The following NOS items were not assessed: due to 544

the unique nature of CA we did not require case-control studies to demonstrate that controls did not 545

have a history of the outcome of interest and we did not require cohort studies to provide 546

demonstration that the outcome of interest was not present at start of study. Quality scores were 547

not used since weighting and calculating overall quality scores is arbitrary and the importance of 548

individual items and direction of bias depends on the context in which they are applied. 549

Case-control studies 550

Case definition was considered low risk of bias if the source reported diagnosis by healthcare 551

professional, hospital or primary records, or utilized independent validation. Representativeness of 552

cases were considered low risk of bias if cases were drawn from a clearly defined period of time and 553

catchment area, or consisted of an appropriate random sample of cases. We considered selection of 554

controls low risk of bias if controls could have been in the case group had the outcome of interest 555

been present as opposed to controls from different time periods or locations than cases. For 556

assessing risk of confounding (comparability in NOS) we considered any comparative study which did 557

match or adjust for confounding by at least maternal age and socioeconomic class low risk of bias. 558

For socioeconomic class, several proxies such as malnutrition for older studies and race for American 559

studies were also considered satisfactory. Fever and antiviral use were not considered confounders 560

Page 102 of 134




32

for reasons stated in the introduction. For ascertainment of exposure, we considered serological 561

confirmation, diagnosis by healthcare professionals and prospective interviews low risk of bias. Post 562

delivery maternal interviews were considered high risk of bias since mothers of infants born with 563

anomalies might make a more determined effort to recall illnesses during pregnancy than other 564

mothers leading to differential recall bias. In case-control designs, recall bias has been estimated to 565

bias OR up to a factor of 1.9.(Rockenbauer et al. 2001) We considered attrition or non-response rates 566

≥20% for cases or controls high risk of bias. 567

Cohort studies 568

Representativeness of the exposed cohort was considered low risk of bias if the exposed cohort was 569

drawn from a clearly defined period of time and catchment area, or consisted of an appropriate 570

random sample of eligible cohort subjects. If ≤80% of the identified individuals in this place and time 571

failed to enrol, representativeness was also considered high risk of bias. We considered selection of 572

the unexposed cohort low risk of bias if individuals in the unexposed cohort could have been in the 573

exposed cohort had they been exposed to 1st

trimester influenza infection. Ascertainment of 574

exposure was handled the same way as for the case-control studies (see above). Length of follow-up 575

is only relevant for certain congenital anomalies such as congenital heart defects and we considered 576

a length of follow up of 1 year satisfactory. With respect to adequacy of follow up, we considered 577

≤20% attrition low risk of bias. 578

Table III: Overview of 9 papers from 8 ecological studies included in the systematic review. 579

Authors Publication

year (study

years) and

location

Study outcome Influenza

exposure and

critical period

No. of cases and

methodology

Buck

(Buck

1955

(Jan 1944-Jul

Infant death due to CA Influenza

incidence per

Unclear population

size, infant deaths only.

Page 103 of 134




33

1955) 1951), Canada calendar month

coinciding with

1st trimester of

pregnancy

Analysis by visual

inspection.

Busby et

al (Busby

et al.

2005)

2005 (1988-

1994), England

Anophthalmos and

microphthalmos

Reported

weekly

influenza

infection counts

coinciding with

6th-10th

week

of gestation.


live births and


Poisson regression

comparing cases to all

births.

Hakosalo

and Saxen

(Hakosalo

and Saxen

1971)

1971 (Jan 15th

-

Oct 31st

, 1958),

Finland

Anencephaly and CA of

the CNS, circulatory

system and urogenital

system.

Sickness

absenteeism

rates coinciding

with 5th

-11th

week of

gestation.


live births and


χ2 test, comparing high


pregnancies. High risk

pregnancies were

exposed to Pandemic

Asian Flu in 5th

-11th

week of gestation. Low

risk pregnancies were

conceived earlier

(Group 1) or conceived

later (Group 2)

Leck (Leck 1963 and 1964 Anencephaly, Spina Weekly new 939 malformed infants,

Page 104 of 134




34

1963, Leck

1964)

(1957-

1961(Leck

1963), 1957-

1963(Leck

1964)),

Birmingham,

UK

Bifida, Cleft Lip, Cleft

Palate, Oesophageal

Atresia, Anal Atresia,

Renal Agenesis or

Hypoplasia,

Hydronephrosis,

Hypospadias, Limb

Reductions (Thumb or

Radii, Bilateral arms or

legs, single arm or leg),

Exomphalos and

Diaphragmatic

Hernia.(Leck 1963)

claims to

sickness benefit

coinciding with

26 to 40 weeks

before delivery.

live births and

stillbirths.(Leck 1963)

15 additional 1957-

1961 cases and

unknown amount of

1962-1963 cases.(Leck

1964)

Analysis by incidence

ratio’s of high risk

versus low risk

pregnancies.

Leck et al

(Leck et al.

1969)

1969 (1956-

1965), Several

USA states

Orofacial Clefts (1956-

1965) and any CA (1961-

196%). Anencephalus,

Spina Bifida,

Hydrocephalus,

Microcephalus, Cleft

Palate, Cleft Lip, Cleft Lip

with Cleft Palate,

Esophageal Defects,

Anoreactal Defects,

Hypospadias, Clubfoot,

Reduction Defects

(Upper Limb, Lower

Weekly number

of deaths

attributable to

influenza

coinciding with

26 to 40 weeks

before delivery.

For 1956-1961: 1,496

isolated cleft palate,

1,418 isolated cleft lip,

2,298 cleft lip with cleft

palate. For 1962-1965:

9,980 CAs. Live births

only. Analysis by crude

and standardized (by

season and year)



pregnancies and χ2

test.

Page 105 of 134




35

Limb, Upper and Lower

Limb, Limbs

Unspecified), Congenital

Hip Dislocation,

Diaphragmatic Hernia,

Down’s Syndrome and

Exomphalos.

Leck (Leck

1971)

1971

(Birmingham:

1954-1965,

England and

Wales 1964-

1968, USA:

1955-1965).

Birmingham: Cleft Lip,

Cleft Lip with Cleft Palate

and Limb Reductions.

England and Wales:

Anencephalus, Spina

Bifida, Encephalocele,

Cleft Lip, Cleft Lip with

Cleft Palate, Esophageal

Atresia and Stenosis,

Anorectal Atresia and

Stenosis, Limb

Reduction, Exomphalos.

USA: Cleft Lip and Limb

Reductions.

Weekly new

claims to

sickness benefit

coinciding with

26 to 40 weeks

before delivery.

Birmingham: 106 cleft

lip, 200 cleft lip with

cleft palate and 136

limb reductions, live

births and stillbirths.

England and Wales:

13,707 CAs, live births

and stillbirths. USA:

Unknown amount of

CAs, live births only.

Analysis by crude and

standardized (by

season and year)



pregnancies and χ2

test.

Record

(Record

1961 (1938-

1958),

Anencephaly Winters (Oct-

Mar) with

5,039 Anencephaly

cases. Analysis by

Page 106 of 134




36

1961) Scotland. influenza

deaths >500

were

considered

epidemic

winters and

linked with

anencephaly

incidences the

next May-Oct.

visual inspection.

Saxen et

al (Saxen

et al.

1990)

1990 (1968-

1982), Finland

Anencephaly High risk

months were

identified via

the Central

Public Health

Laboratory and

coinciding

pregnancies in

the 1st

trimester

were

considered high

risk.

248 Anencephaly

cases. Analysis by rate

ratio’s between high

risk and low risk

pregnancies.

Results from these studies are included in Table II. 580

Page 107 of 134




37

Table IV: Overview of 25 papers from 15 case-control studies sorted by year. 581

1st

Author Publication

year (study

years) and

location

Exposure ascertainment Birth types Controls, matching and

adjustment (if any)

Coffey (Coffey and Jessop 1955) 1955 (1953-

1954), Ireland

Exposure was based on maternal

reports, retrospective to delivery.

Live and

stillbirths were

included.

Random sample of non-malformed

controls.

Laurence (Laurence et al. 1968) 1968 (1956-

1962), UK

The study suggests exposure was based

on maternal reports, retrospective to

delivery for 1956-1960 and mainly

prospective to delivery after 1960.

Live and

stillbirths were

included.

Controls without neural tube

defects.

Matched by date of birth, place of

residence, sex of infant.

Elizan (Elizan et al. 1969) 1969 (1959-

1964), USA

Exposure was based on complement

fixation test.

Autopsied

stillbirths were

included.

Controls without central nervous

system defects.

Matched by gravidity, institution,

last menstrual period, maternal

Page 108 of 134




38

age and race.

Choi (Choi and Klaponski 1970) 1970 (1963-

1968), Canada


interviews after birth supplemented by

medical records.

It is unknown

which birth

types were

included.

Controls without neural tube

defects.

Matched by date of birth, hospital,

sex of infant.

Saxen (Saxen 1975) 1975a (1967-

1971), Finland

Finnish registry of congenital

malformations (FRCM). Exposure was

based on maternal reports, retrospective

to delivery.

Live and

stillbirths were

included.

Controls without orofacial clefts.

Matched by place of residence and

time of birth.

Saxen (Saxen 1975) 1975b (1972-

1973), Finland




to delivery.

Live and

stillbirths were

included.

Controls without orofacial clefts.

Matched by place of residence and

time of birth.

Klemetti (Klemetti 1977) 1977 (1963-

1965), Finland



based on maternal reports, prospective

Live and

stillbirths were

included.

Non-malformed controls and cases

originated from a series of

consecutive pregnancies registered

Page 109 of 134




39

to delivery. in the same county.

Granroth (Granroth 1978, Granroth et

al. 1978) (+earlier papers by

Karkinen-Jääskeläinen (Karkinen-

Jaaskelainen and Saxen 1974) and

Saxen (Saxen et al. 1974)).

1978 (1965-

1973), Finland




to delivery and maternity centres,


Live and

stillbirths were

included.

Non-malformed controls. Matched

by date of delivery and Maternity

Welfare District.

Warrell (Warrell et al. 1981) 1981 (1972-

1976), UK

Exposure was based on fluorescent-

antibody technique.

It is unknown

which types of

births were

included.

Non-malformed controls.

Matched (as far as possible) by

data of blood serum collection,

gestational age and parity.

Aro (Aro 1983) 1983 (1964-

1977), Finland



mainly based antenatal records,

collected prospective to delivery.

Live and

stillbirths were

included.

Non-malformed control.

Matched by date of delivery and

place.

Lynberg (Lynberg et al. 1994) (+

earlier paper by Park (Park et al.

1994 (1968-

1980), USA

Atlanta Birth Defects Case Control Study

(ABDCCS). Exposure was based on

Live and

stillbirths were

Non-malformed and malformed

controls (with CA “outside the

Page 110 of 134




40

1992)) maternal reported episodes of >2 days

of flu and fever from 1 month before to

3 months after conception, retrospective

to delivery.

included. hyperthermia spectrum”).

Matched by hospital of birth, race,

year and quarter of birth.

Adjusted for alcohol use,

education, maternal age,

multivitamin use and smoking.

Botto (Botto et al. 2001) (+ earlier

paper by Adams (Adams et al. 1989))

2001 (1968-

1980), USA

Atlanta Birth Defects Case Control Study

(ABDCCS). Exposure was based on

maternal reports, retrospective to

delivery.

Live and

stillbirths were

included.


Matched by hospital, race and time

of birth.

Adjusted for alcohol use, chronic

illness, education, race,

multivitamin use, period of birth

and smoking.

Li(Li et al. 2007) 2006 (2003-

2005), China



Live and

stillbirths were


Matched by county, estimated

Page 111 of 134




41

included. date of conception, ethnic group

and sex.

Adjusted for antibiotics,

antipyretics, diet, folic acid, history

of major congenital anomaly

affected pregnancy, maternal

education and infectious disease

other than influenza.

Czeizel (+ earlier papers by Acs,

Medveczky and Metneki) (Acs et al.

2005, Czeizel et al. 2008, Medveczky

et al. 2004, Metneki et al. 2005) (+

later paper by Csáky-Szunyogh

(Csaky-Szunyogh et al. 2013))

2007 (1980-

1996), Hungary

Hungarian Case-Control Surveillance of

Congenital Anomalies (HCCSCA).



Live and

stillbirths were

included.


Matched by district of residence,

sex, year and week of birth.

Oster (Oster et al. 2011) (+ later paper

by Kelly (Kelly et al. 2012))

2011 (1981-

1989), USA

Baltimore-Washington Infant Study

(BWIS). Exposure was based on maternal

Live births

were included.

Random sample of live-bortn

controls without congenital heart

Page 112 of 134




42

reports, retrospective to delivery. disease.

Frequency matched on age at

interview, month, year and

hospital of birth.

Adjusted for alcohol use, BMI,

family history of congenital heart

disease, gestational diabetes, race,

sex and smoking.


Page 113 of 134




43

Table V: Overview of 12 papers from 10 cohort studies sorted by year. 583

1st

Author Publication

year (study

years) and

location

Recruitment Exposure ascertainment Matching

additional to

location and

adjustment (if

any)

Campbell

(Campbell

1953)

1953 (1950-

1951), UK

Mothers attending

antenatal clinics

were enrolled.


maternal reports,


None

Abramowitz

(Abramowitz

1958)

1958 (1957),

South Africa

Women attending

gynaecology wards

were enrolled.

Exposure ascertainment

was not described.

None

Walker (Walker

and McKee

1959)

1959 (1957),

USA

Ward patients who

delivered at

University hospitals

were enrolled.


hemagglutination

inhibition and maternal

reports, retrospective to

delivery.

None

Doll (Doll et al.

1960)

1960 (1957-

1958), UK

Mothers attending

an antenatal clinic

for the first time

during the study

period were

enrolled.


maternal reports,

prospective to delivery

and some of these were

confirmed with

professional diagnosis.

None

Pleydell

(Pleydell 1960)

1960 (1957),

UK

Expecting mothers

reported to suffer

from ILI were


maternal reports,


None

Page 114 of 134




44

included and

exposed and non-

exposed were

recruited via local

midwives.

Saxen (Saxen et

al. 1960)

1960 (1958),

Finland

Women delivering in

the University

Central Hospital of

Helsinki were

enrolled.


maternal reports,


None

Hardy (Hardy et

al. 1961)

1961 (1957-

1958), USA

Patients visiting an

obstetrical prenatal

clinic were enrolled.


complement fixation,

hemagglutinin inhibition

and maternal reports,


None

Hirvensalo

(Hirvensalo and

Kinnunen 1962)

1962 (1957-

1959),

Finland

Women attending

the Maternity

Health Centres were

enrolled.


maternal reports,


None

Coffey (Coffey

and Jessop

1959, Coffey

and Jessop

1963)

1959 and

1963 follow-

up (1957-

1958), Ireland

Women attending

antenatal clinics

were enrolled.


maternal reports,


Stage of

pregnancy.

Wilson (Wilson

et al. 1959,

1959 and

1969 follow-

Prenatal registrants

were enrolled.


hemagglutinin inhibition.

None

Page 115 of 134




45

Wilson and

Stein 1969)

up (1957),

USA


Figure 10: Risk of bias assessment of included case-control studies. 585

[Figure 10] 586


bias. 588

Figure 11: Risk of bias assessment of included cohort studies. 589

[Figure 11] 590


bias. 592

Supplementary Forest Plots 593

Figure 12: Forest plot of aortic valve atresia/stenosis following 1st


[Figure 12] 595

Figure 13: Forest plot of atrial septal defect following 1st trimester influenza exposure 596

[Figure 13] 597

Figure 14: Forest plot of hypoplastic left heart following 1st


[Figure 14] 599

Figure 15: Forest plot of transposition of the great vessels following 1st trimester influenza exposure 600

[Figure 15] 601

Figure 126: Forest plot of ventricular septal defects following 1st


[Figure 126] 603

Figure 137: Forest plot of cleft lip ± cleft palate following 1st


[Figure 137] 605

Figure 148: Forest plot of cleft palate following 1st


Page 116 of 134




46

[Figure 148] 607

Figure 159: Forest plot of digestive system defects following 1st


[Figure 159] 609

Figure 1620: Forest plot of urinary defects following 1st trimester influenza exposure 610

[Figure 1620] 611

Figure 1721: Forest plot of hypospadias following 1st


[Figure 1721] 613

Figure 1822 Forest plot of limb reduction defects following 1st


[Figure 1822] 615

Figure 1923: Forest plot of club foot following 1st trimester influenza exposure 616

[Figure 1923] 617

Figure 2024: Forest plot of hip dislocation and/or dysplasia following 1st


[Figure 2025] 619

Figure 2125: Forest plot of polydactyly following 1st trimester influenza exposure 620

[Figure 2125] 621

Figure 2226: Forest plot of syndactyly following 1st


[Figure 2226] 623

Figure 2327: Forest plot of musculo-skeletal defects following 1st


[Figure 2327] 625

Supplementary Funnel Plots 626

Funnel plot asymmetry is indicative for publication bias as for example small studies reporting 627

negative findings might not be published. Visual inspection of funnel plots did not provide evidence 628

for publication bias. 629

Figure 248: Funnel plot of non-chromosomal CA following 1st


[Figure 2428] 631

Page 117 of 134




47

Figure 259: Funnel plot of neural tube defectsNTD following 1st


[Figure 2529] 633

Figure 2630: Funnel plot of anencephaly following 1st trimester influenza exposure 634

[Figure 2630] 635

Figure 2731: Funnel plot of encephalocele following 1st trimester influenza exposure 636

[Figure 2731] 637

Figure 2832: Funnel plot of spina bifida following 1st


[Figure 2832] 639

Figure 2933: Funnel plot of hydrocephaly following 1st


[Figure 2933] 641

Figure 3034: Funnel plot of congenital heart defects following 1st


[Figure 3034] 643

Figure 3235: Funnel plot of orofacial clefts following 1st


[Figure35] 645

Figure 36: Funnel plot of aortic valve atresia/stenosis following 1st


[Figure 36] 647

Figure 37: Funnel plot of atrial septal defect following 1st trimester influenza exposure 648

[Figure 37] 649

Figure 38: Funnel plot of hypoplastic left heart following 1st


[Figure 38] 651

Figure 39: Funnel plot of transposition of the great vessels following 1st


[Figure 39] 653

Figure 3140: Funnel plot of ventricular septal defects following 1st


[Figure 3140] 655

Figure 32: Funnel plot of orofacial clefts following 1st trimester influenza exposure 656

[Figure 32] 657

Page 118 of 134




48

Figure 3341: Funnel plot of cleft lip ± cleft palate following 1st


[Figure 3341] 659

Figure 3442: Funnel plot of cleft palate following 1st


[Figure 3442] 661

Figure 3543: Funnel plot of digestive system defects following 1st


[Figure 3543] 663

Figure 3644: Funnel plot of urinary defects following 1st


[Figure 3644] 665

Figure 3745: Funnel plot of hypospadias following 1st


[Figure 3745] 667

Figure 3846: Funnel plot of limb reduction defects following 1st


[Figure 3846] 669

Figure 3947: Funnel plot of club foot following 1st trimester influenza exposure 670

[Figure 3947] 671

Figure 4048: Funnel plot of hip dislocation and/or dysplasia defects following 1st


exposure 673

[Figure 4048] 674

Figure 4149: Funnel plot of polydactyly following 1st


[Figure 4149] 676

Figure 4250: Funnel plot of syndactyly following 1st trimester influenza exposure 677

[Figure 4250] 678

Figure 4351: Funnel plot of musculo-skeletal defects following 1st


[Figure 4351] 680

Page 119 of 134




49

References 681

Abramowitz LJ. The effect of Asian influenza on pregnancy. S Afr Med J 1958:32:1155-1156. 682

Acs N, Banhidy F, Puho E and Czeizel AE. Maternal influenza during pregnancy and risk of congenital 683

abnormalities in offspring. Birth Defects Res A Clin Mol Teratol 2005:73:989-996. 684

Adams MM, Mulinare J and Dooley K. Risk factors for conotruncal cardiac defects in Atlanta. J Am 685

Coll Cardiol 1989:14:432-442. 686

Aro T. Maternal diseases, alcohol consumption and smoking during pregnancy associated with 687

reduction limb defects. Early Hum Dev 1983:9:49-57. 688



Boyd PA, Devigan C, Khoshnood B, Loane M, Garne E, Dolk H and EUROCAT Working Group. Survey of 691

prenatal screening policies in Europe for structural malformations and chromosome anomalies, and 692

their impact on detection and termination rates for neural tube defects and Down's syndrome. BJOG 693

2008:115:689-696. 694

Buck C. Exposure to virus diseases in early pregnancy and congenital malformations. Can Med Assoc J 695

1955:72:744-746. 696

Busby A, Dolk H and Armstrong B. Eye anomalies: seasonal variation and maternal viral infections. 697

Epidemiology 2005:16:317-322. 698

Campbell WA. Influenza in early pregnancy; effects on the foetus. Lancet 1953:1:173-174. 699

Choi NW and Klaponski FA. On neural-tube defects: an epidemiological elicitation of etiological 700

factors. Neurology 1970:20:399-400. 701

Page 120 of 134




50

Cockroft DL and New DA. Abnormalities induced in cultured rat embryos by hyperthermia. 702

Teratology 1978:17:277-283. 703

Cockroft DL and New DA. Effects of hyperthermia on rat embryos in culture. Nature 1975:258:604-704

606. 705

Coffey VP and Jessop WJ. Maternal influenza and congenital deformities. A follow-up study. Lancet 706

1963:1:748-751. 707

Coffey VP and Jessop WJ. Maternal influenza and congenital deformities: a prospective study. Lancet 708

1959:2:935-938. 709

Coffey VP and Jessop WJ. Congenital abnormalities. Ir J Med Sci 1955:6:30-48. 710

Csaky-Szunyogh M, Vereczkey A, Kosa Z, Urban R and Czeizel AE. Association of maternal diseases 711

during pregnancy with the risk of single ventricular septal defects in the offspring--a population-712

based case-control study. J Matern Fetal Neonatal Med 2013:26:738-747. 713

Czeizel AE, Puho EH, Acs N and Banhidy F. Use of specified critical periods of different congenital 714

abnormalities instead of the first trimester concept. Birth Defects Res A Clin Mol Teratol 2008:82:139-715

146. 716

Davies HT, Crombie IK and Tavakoli M. When can odds ratios mislead?. BMJ 1998:316:989-991. 717

DerSimonian R and Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986:7:177-188. 718

Dodds L, McNeil SA, Fell DB, Allen VM, Coombs A, Scott J and MacDonald N. Impact of influenza 719

exposure on rates of hospital admissions and physician visits because of respiratory illness among 720

pregnant women. CMAJ 2007:176:463-468. 721

Page 121 of 134




51

Doll R, Hill AB and Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 722

1960:14:167-172. 723

Edwards MJ. Review: Hyperthermia and fever during pregnancy. Birth Defects Res A Clin Mol Teratol 724

2006:76:507-516. 725

Edwards MJ. The experimental production of Arthrogryposis multiplex congenita in guinea-pigs by 726

maternal hyperthermia during gestation. J Pathol 1971:104:221-229. 727

Edwards MJ. The experimental production of clubfoot in guinea-pigs by maternal hyperthermia 728

during gestation. J Pathol 1971:103:49-53. 729

Edwards MJ. Congenital defects in guinea pigs: fetal resorptions, abortions, and malformations 730

following induced hyperthermia during early gestation. Teratology 1969:2:313-328. 731

Edwards MJ. Congenital defects in guinea pigs: prenatal retardation of brain growth of guinea pigs 732

following hyperthermia during gestation. Teratology 1969:2:329-336. 733

Elizan TS, Ajero-Froehlich L, Fabiyi A, Ley A and Sever JL. Viral infection in pregnancy and congenital 734

CNS malformations in man. Arch Neurol 1969:20:115-119. 735

EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013:2013. 736

EUROCAT Central Registry. EUROCAT Guide 1.3 and reference documents. Instructions for the 737

Registration and Surveillance of Congenital Anomalities. Revised: May 2009 2009:. 738

Evans D. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare 739

interventions. J Clin Nurs 2003:12:77-84. 740

Germain MA, Webster WS and Edwards MJ. Hyperthermia as a teratogen: parameters determining 741

hyperthermia-induced head defects in the rat. Teratology 1985:31:265-272. 742

Page 122 of 134




52

Granroth G. Defects of the central nervous system in Finland: III. Disease and drugs in pregnancy. 743

Early Hum Dev 1978:2:147-162. 744



Gu J, Xie Z, Gao Z, Liu J, Korteweg C, Ye J, Lau LT, Lu J, Gao Z, Zhang B et al. H5N1 infection of the 747

respiratory tract and beyond: a molecular pathology study. Lancet 2007:370:1137-1145. 748

Hakosalo J and Saxen L. Influenza epidemic and congenital defects. Lancet 1971:2:1346-1347. 749

Hardy JM, Azarowicz EN, Mannini A, Medearis DN,Jr and Cooke RE. The effect of Asian influenza on 750

the outcome of pregnancy, Baltimore, 1957-1958. Am J Public Health Nations Health 1961:51:1182-751

1188. 752

Harris J. Influenza occurring in pregnant women: a statistical study of thirteen hundred and fifty 753

cases. JAMA 1919;72:978–80 1919:. 754

Hirvensalo M and Kinnunen O. Influenza and pregnancy. Duodecim 1962:78:740-747. 755

Hofler M. The Bradford Hill considerations on causality: a counterfactual perspective. Emerg Themes 756

Epidemiol 2005:2:11. 757

Jamieson DJ, Theiler RN and Rasmussen SA. Emerging infections and pregnancy. Emerg Infect Dis 758

2006:12:1638-1643. 759

Kallen B and Olausson PO. Vaccination against H1N1 influenza with Pandemrix((R)) during pregnancy 760

and delivery outcome: a Swedish register study. BJOG 2012:119:1583-1590. 761

Karkinen-Jaaskelainen and Saxen L. Maternal influenza, drug consumption, and congenital defects of 762

the central nervous system. Am J Obstet Gynecol 1974:118:815-818. 763

Page 123 of 134




53

Kelly S, Keuhl K and Leffredo C. Tricupsid Atresia: an epidemiologic investigation in the Baltimore-764

Washington area. American Journal of Epidemiology 2012:175:S59. 765

Klemetti A. Definition of congenital malformations and detection of associations with maternal 766

factors. Early Hum Dev 1977:1:117-123. 767

Laurence KM, Carter CO and David PA. Major central nervous system malformations in South Wales. 768

II. Pregnancy factors, seasonal variation, and social class effects. Br J Prev Soc Med 1968:22:212-222. 769

Leck I. Further tests of the hypothesis that influenza in pregnancy causes malformations. HSMHA 770

Health Rep 1971:86:265-269. 771

Leck I. Examination of the Incidence of Malformations for Evidence of Drug Teratogenesis. Br J Prev 772

Soc Med 1964:18:196-201. 773

Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963:17:70-80. 774

Leck I, Hay S, Witte JJ and Greene JC. Malformations recorded on birth certificates following A2 775

influenza epidemics. Public Health Rep 1969:84:971-979. 776

Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural tube 777

defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 778

Teratol 2007:79:295-300. 779

Luteijn JM, Dolk H and Marnoch GJ. Differences in pandemic influenza vaccination policies for 780

pregnant women in Europe. BMC Public Health 2011:11:819. 781

Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: a 782

population-based case-control study. Am J Epidemiol 1994:140:244-255. 783

Page 124 of 134




54

MacKenzie JS and Houghton M. Influenza infections during pregnancy: association with congenital 784

malformations and with subsequent neoplasms in children, and potential hazards of live virus 785

vaccines. Bacteriol Rev 1974:38:356-370. 786

Mak TK, Mangtani P, Leese J, Watson JM and Pfeifer D. Influenza vaccination in pregnancy: current 787

evidence and selected national policies. Lancet Infect Dis 2008:8:44-52. 788

McGregor JA, Burns JC, Levin MJ, Burlington B and Meiklejohn G. Transplacental passage of influenza 789

A/Bangkok (H3N2) mimicking amniotic fluid infection syndrome. Am J Obstet Gynecol 1984:149:856-790

859. 791

Medveczky E, Puho E and Czeizel AE. An evaluation of maternal illnesses in the origin of neural-tube 792

defects. Arch Gynecol Obstet 2004:270:244-251. 793

Mereckiene J, Cotter S, D'Ancona F, Giambi C, Nicoll A, Levy-Bruhl D, Lopalco PL, Weber JT, Johansen 794

K, Dematte L et al. Differences in national influenza vaccination policies across the European Union, 795

Norway and Iceland 2008-2009. Euro Surveill 2010:15:19700. 796

Metneki J, Puho E and Czeizel AE. Maternal diseases and isolated orofacial clefts in Hungary. Birth 797

Defects Res A Clin Mol Teratol 2005:73:617-623. 798

Moher D, Liberati A, Tetzlaff J, Altman DG and PRISMA Group. Preferred reporting items for 799

systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009:339:b2535. 800

Monto AS, Gravenstein S, Elliott M, Colopy M and Schweinle J. Clinical signs and symptoms predicting 801

influenza infection. Arch Intern Med 2000:160:3243-3247. 802

Moretti ME, Bar-Oz B, Fried S and Koren G. Maternal hyperthermia and the risk for neural tube 803

defects in offspring: systematic review and meta-analysis. Epidemiology 2005:16:216-219. 804

Page 125 of 134




55

Mossey PA, Little J, Munger RG, Dixon MJ and Shaw WC. Cleft lip and palate. Lancet 2009:374:1773-805

1785. 806

Neuzil KM, Reed GW, Mitchel EF, Simonsen L and Griffin MR. Impact of influenza on acute 807

cardiopulmonary hospitalizations in pregnant women. Am J Epidemiol 1998:148:1094-1102. 808



Park CH, Stewart W, Khoury MJ and Mulinare J. Is there etiologic heterogeneity between upper and 811

lower neural tube defects?. Am J Epidemiol 1992:136:1493-1501. 812

Pasternak B, Svanstrom H, Molgaard-Nielsen D, Krause TG, Emborg HD, Melbye M and Hviid A. Risk 813

of adverse fetal outcomes following administration of a pandemic influenza A(H1N1) vaccine during 814

pregnancy. JAMA 2012:308:165-174. 815

Pleydell MJ. Anencephaly and other congenital abnormalities. An epidemiological study in 816

Northamptonshire. Br Med J 1960:1:309-315. 817

Record RG. Anencephalus in Scotland. Br J Prev Soc Med 1961:15:93-105. 818

Rockenbauer M, Olsen J, Czeizel AE, Pedersen L, Sorensen HT and EuroMAP Group. Recall bias in a 819

case-control surveillance system on the use of medicine during pregnancy. Epidemiology 820

2001:12:461-466. 821

Saxen I. The association between maternal influenza, drug consumption and oral clefts. Acta Odontol 822

Scand 1975:33:259-267. 823

Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev Soc 824

Med 1975:29:103-110. 825

Page 126 of 134




56

Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev Soc 826

Med 1975:29:103-110. 827

Saxen L, Hjelt L, Sjostedt JE, Hakosalo J and Hakosalo H. Asian influenza during pregnancy and 828

congenital malformations. Acta Pathol Microbiol Scand 1960:49:114-126. 829

Saxen L, Holmberg PC, Kurppa K, Kuosma E and Pyhala R. Influenza epidemics and anencephaly. Am J 830

Public Health 1990:80:473-475. 831

Saxen L, Klemetti A and Haro AS. A matched-pair register for studies of selected congenital defects. 832

Am J Epidemiol 1974:100:297-306. 833

Siston AM, Rasmussen SA, Honein MA, Fry AM, Seib K, Callaghan WM, Louie J, Doyle TJ, Crockett M, 834

Lynfield R et al. Pandemic 2009 influenza A(H1N1) virus illness among pregnant women in the United 835

States. JAMA 2010:303:1517-1525. 836

Smith MS, Upfold JB, Edwards MJ, Shiota K and Cawdell-Smith J. The induction of neural tube defects 837

by maternal hyperthermia: a comparison of the guinea-pig and human. Neuropathol Appl Neurobiol 838

1992:18:71-80. 839

Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA and 840

Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-841

analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000:283:2008-2012. 842

Sweeting MJ, Sutton AJ and Lambert PC. What to add to nothing? Use and avoidance of continuity 843

corrections in meta-analysis of sparse data. Stat Med 2004:23:1351-1375. 844

Walker WM and McKee AP. Asian influenza in pregnancy; relationship to fetal anomalies. Obstet 845

Gynecol 1959:13:394-398. 846

Page 127 of 134




57

Warrell MJ, Tobin JO and Wald NJ. Examination for influenza IgA and IgM antibodies in pregnancies 847

associated with fetal neural-tube defects. J Med Microbiol 1981:14:159-162. 848

Wells GA, Shea B and O'Connell D. The Newcastle-Ottawa Scale (NOS) for assessing the quality of 849

nonrandomised studies in meta-analyses 2004:. 850

Wilson MG, Heins HL, Imagawa DT and Adams JM. Teratogenic effects of Asian influenza. J Am Med 851

Assoc 1959:171:638-641. 852

Wilson MG and Stein AM. Teratogenic effects of asian influenza. A n extended study. JAMA 853

1969:210:336-337. 854

Zambon M, Hays J, Webster A, Newman R and Keene O. Diagnosis of influenza in the community: 855

relationship of clinical diagnosis to confirmed virological, serologic, or molecular detection of 856

influenza. Arch Intern Med 2001:161:2116-2122. 857

Page 128 of 134




PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist

Section/topic # Checklist item Reported

on page #

TITLE

Title 1 Identify the report as a systematic review, meta-analysis, or both. 8 (of

133).

ABSTRACT

Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria,

participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and

implications of key findings; systematic review registration number.

9-10

INTRODUCTION

Rationale 3 Describe the rationale for the review in the context of what is already known. 11

Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons,

outcomes, and study design (PICOS).

12

METHODS

Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide

registration information including registration number.

NA

Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered,

language, publication status) used as criteria for eligibility, giving rationale.

12

Page 129 of 134





Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify

additional studies) in the search and date last searched.

12

Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be

repeated.

12

Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable,

included in the meta-analysis).

12

Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes

for obtaining and confirming data from investigators.

13

Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and

simplifications made.

13

Risk of bias in individual

studies

12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was

done at the study or outcome level), and how this information is to be used in any data synthesis.

13-14

(full

summary

100-101)

Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 14

Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency

(e.g., I2) for each meta-analysis.

14

Page 1 of 2

Page 130 of 134





Section/topic # Checklist item Reported

on page #

Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective

reporting within studies).

Publication

Bias page

14.

Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done,

indicating which were pre-specified.

Subgroup

analysis

page 14.

RESULTS

Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions

at each stage, ideally with a flow diagram.

15 (+Flow

Chart page

41).

Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period)

and provide the citations.

Ecological

Studies 56-

60.

Case-

Control

Page 131 of 134





studies 61-

66.

Cohort

studies 67-

69.

Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). 15-17, 70-

71.

Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each

intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

37-40 for

studies not

enrolled in

meta-

analysis.

42-49 (plus

figures 12-

27 of

eAppendix)

for studies

Page 132 of 134





enrolled in

meta-

analysis.

Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. 34-36

Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). 34

Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]). 34

DISCUSSION

Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to

key groups (e.g., healthcare providers, users, and policy makers).

15-17.

Relevance

key groups

26-27.

Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of

identified research, reporting bias).

18-19.

(limitations

and

statistical

limitations

discussed

Page 133 of 134





in separate

paragraphs)

Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research. 19-20

FUNDING

Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for

the systematic review.

20

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097

For more information, visit: www.prisma-statement.org.

Page 2 of 2

Page 134 of 134




influenza and congenital anomalies: a systematic review ... · please describe and define eurocat....

Documents