Vaccine 39 (2021) B12–B26 Contents lists available at ScienceDirect Vaccine journal homepage: www.elsevier.com/locate/vaccine Review Adverse events associated with the use of recommended vaccines during pregnancy: An overview of systematic reviews q Diego Macias Saint-Gerons a,b, Iván Solà Arnau c, Bremen De Mucio d, Ingrid Arévalo-Rodríguez e, Alicia Alemán d, José Luis Castro b, Alba María Ropero Álvarez f,⇑ a Department of Medicine, University of Valencia/INCLIVA Health Research Institute and CIBERSAM, Valencia, Spain b Department of Health Systems and Services (HSS)/Unit of Medicines and Health Technologies (MT), Pan American Health Organization PAHO/WHO, Washington DC, USA c Iberoamerican Cochrane Centre, Biomedical Research Institute Sant Pau (IIB Sant Pau), CIBER of Epidemiology and Public Health, Spain d Department of Preventive Medicine, School of Medicine, University of the Republic, Montevideo, Uruguay e Clinical Biostatistics Unit, Hospital Universitario Ramón y Cajal, IRYCIS, CIBER of Epidemiology and Public Health, Madrid, Spain f Department of Family, Gender and Life Course, Immunization Unit, Pan American Health Organization PAHO/WHO, Washington DC, USA a r t i c l e i n f o a b s t r a c t Article history: Introduction: Maternal immunization is aimed at reducing morbidity and mortality in pregnant women Available online 21 September 2020 and their newborns. Updated evidence synthesis of maternal-fetal outcomes is constantly needed to ensure that the risk-benefit of vaccination during pregnancy remains positive. Keywords: Methods: An overview of systematic reviews (OoSRs) was performed. We searched The Cochrane Library, Pregnancy MEDLINE and EMBASE for SRs including recommended vaccines for maternal immunization reporting the Vaccination following: abortion, stillbirth, chorioamnionitis, congenital anomalies, microcephaly, neonatal death, Adverse effects neonatal infection, preterm birth (PTB), low birth weight (LBW), maternal death and small for gestational Epidemiological study Meta-analysis age (SGA) from 2010 to April 2019. Quality and overlap of SRs was assessed. Systematic review Results: Seventeen SRs were identified, eight of them included meta-analysis; quality was high in three SRs, moderate in six SRs, low in two SRs, and critically low in six SRs. Stillbirth and PTB were the most frequently reported outcomes by 15 and 13 SRs, respectively, followed by abortion (9 SRs), congenital anomalies (9 SRs), SGA (8 SRs), neonatal death (8 SRs), LBW (4 SRs), chorioamnionitis (3 SRs), maternal death (1 SR). SRs included mainly observational evidence for influenza and Tdap vaccines (11 SRs and 4 SRs, respec- tively); limited evidence was found for hepatitis (1 SR), yellow fever (1 SR), and meningococcal (1 SR) vaccines. Most of the SRs found no effect. Eight SRs found benefit/protection of influenza vaccine (for stillbirth, neonatal death, preterm birth, LBW), or Tdap vaccine (for preterm birth and SGA); one found a probable risk (chorioam- nionitis/Tdap). The SRs for Hepatitis B, meningococcal and yellow fever vaccines were inconclusive. Conclusions: Definite risks were not identified for any vaccine and outcome; however better evidence is needed for all outcomes and vaccines. The available evidence in the SRs to support vaccine safety was based mainly on observational data. More RCTs with adequate reporting of maternal-fetal outcomes and larger high- quality observational studies are needed. Ó 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Contents 1. Introduction . . . . . .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B13 2. Methods . . . . . . . . .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B13 3. Results. . . . . . . . . . .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B13 4. Discussion . . . . . . . .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B15 CRediT authorship contribution statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B25 q Publication of this supplement was supported by the Pan American Health Organization (PAHO). The views expressed herein are those of the authors and do not necessarily reflect the official policy or position of PAHO. ⇑ Corresponding author. E-mail address: roperoal@paho.org (A.M. Ropero Álvarez). https://doi.org/10.1016/j.vaccine.2020.07.048 0264-410X/Ó 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 B13 Declaration of Competing Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B25 Appendix A. Supplementary material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B25 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B25 1. Introduction We searched the Cochrane Database of Systematic Reviews and the Database of Abstracts of Reviews of Effect through The Maternal immunization is a promising strategy to reduce mor- Cochrane Library. Additionally, we searched MEDLINE and EMBASE bidity and mortality in pregnant women and their newborns [1]. up to April 2019. We designed search strategies (Appendix, However, many countries, mostly low- or middle-income coun- Table S1.) adapted to the requirements of each database, which tries, have not yet introduced this strategy into their national included a combination of controlled vocabulary, search terms, immunization programmes [2]. and filters to retrieve systematic reviews. In addition, we searched Vaccines have been associated with fewer concerns of pharma- the lists of references from eligible reviews for additional ones. We covigilance than other biological medicines [3,4]. Moreover, live included reviews published from 2010 onwards to prioritize the vaccines are not routinely given to pregnant women, and immu- applicability of findings [12]. For eligible Cochrane reviews, we nization with inactivated vaccines or toxoids during pregnancy is included the most recently updated version. not expected to be associated with an increased risk to the fetus Two researchers independently screened results from search [5,6]. However, additional surveillance activities (including stud- strategies, based on references, title and abstract, and confirmed ies) are recommended to gain information on vaccine safety with the studies’ eligibility by obtaining a full copy of relevant articles. regard to adverse events following immunization (AEFI) in the We solved disagreements by discussion until consensus was mother or the baby, to ensure that the risk-benefit balance remains reached or by involving a third researcher. For each included positive in pregnant women and their newborns [1,7]. review, one researcher obtained data, in a pre-piloted form, on Systematic reviews and meta-analyses of Randomized Clinical review characteristics [study design (included), and population, Trials can also be used to confirm the safety of a drug or group of vaccine, and outcomes assessed], an appraisal of its quality, and drugs when a signal arises from other methods, and their use is its findings and effect estimate for the main outcomes. A second currently considered to be third generation pharmacovigilance, researcher cross-checked the data extracted for accuracy. We sum- over spontaneous reports, and observational studies, which are marized data from this extraction process into descriptive tables. considered first and second pharmacovigilance generation respec- We assessed the methodological quality from the included reviews tively [8]. The updated evidence from systematic reviews [or over- with the AMSTAR-II checklist, rating the overall quality according views] of reviews can provide support for designing and planning to the limitations identified in the different domains [13]. We new recommendations, studies, and/or meta-analyses [9]. quantified the overlap of primary studies included in eligible SRs We conducted an overview of systematic reviews reporting by calculating the corrected covered area [CCA] from the number data on safety after immunization with recommended vaccines of SRs considered eligible and all the primary studies included in during pregnancy in the Maternal and Neonatal Immunization those reviews [14]. We classified the degree of overlap according Field Guide for Latin America and the Caribbean [6]. to the percentage overlap, indicating slight [0–5%], moderate [6– 10%], high [11–15%], and very high [>15%]. As we anticipated that the reviews would report their findings 2. Methods using different effect estimates on multiple outcome measures, we limited the synthesis of results to a narrative summary of find- We conducted an overview of reviews according to a protocol ings grouped by type of vaccine and outcomes of interest. We cat- that was registered in PROSPERO [registration number: egorized the conclusions provided by the review authors by each CRD42018091216]. We adhered to the guidance provided by pre-defined outcome according to the amount and robustness of Cochrane [10] and reported our findings according to the PRISMA available evidence into categories: inconclusive; no effect; risk; statement [11]. We included published systematic reviews that and benefit/risk protection (Table S6). We synthetized the findings explicitly stated pre-defined eligibility criteria for studies, reported from the overview, providing narrative summaries of the results a search in at least two databases, and included controlled studies. obtained from the included SRs with the highest quality according Reviews that provided an assessment or discussion of the risk of to the outcomes of interest. bias in identified studies were analyzed separately from reviews not assessing the risk of bias in included primary studies. 3. Results We considered eligible those reviews assessing the safety of the following vaccines administered during pregnancy [6]: a) Hepatitis We described the complete eligibility process in a PRISMA flow- A/B; b) Influenza (including H1N1); c) Meningococcal disease; d) chart (Appendix, Fig. S1). After removing duplicates, we screened Poliomyelitis; e) Rabies; f) Tetanus/Diphtheria/Pertussis [including 218 unique references. We obtained 34 full-text studies for a Tdap); g) Yellow fever. Our pre-defined primary outcomes detailed assessment, but excluded 17 of them because they did included: a) Neonatal death (fetal death, newborn death); b) not assess safety outcomes or used any of the outcomes of interest Neonatal infection (unrelated to the infection that prevents the for this overview as an effectiveness measure (6 studies), and the vaccine given); c) Chorioamnionitis; d) Congenital anomalies; e) remaining (11 studies) for other reasons. The complete list of Microcephaly; f) Preterm birth (<37 weeks); g) Abortion excluded studies is available in the Appendix, (Table S2). We (<24 weeks); h) Stillbirth (24 weeks). As a secondary outcome, included 17 systematic reviews [15–31]. Table 1 outlines the main we also assessed: a) LBW (<2500 g); b) Maternal death during characteristics of the included reviews. A detailed summary of each the follow-up period; c) SGA. We included reviews regardless of included SR is available in the Appendix (Table S3.). Most of them the geographic area covered. Reviews of effectiveness were included data from observational studies. Influenza vaccination, included if they also reported disaggregated data for safety including the H1N1 vaccine, was assessed in two thirds of reviews outcomes. and the most common outcomes assessed by reviews were: pre- B14 D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 Table 1 small proportion of reviews were conducted according to a pre- Characteristics from included systematic reviews. established protocol [5/17]. According to the corrected covered Review type area for those vaccines that were assessed in more than one Narrative SR 9 review, the overlap was high [influenza 33%, pertussis 42%]. SR+ MA 8 Tables 2–6 summarize the main results from included reviews Cochrane reviews 2 Study designs included according to our pre-defined outcomes. We summarize the main Randomized clinical trials (RCT) 2 conclusions for the primary outcomes from the nine reviews with Observational 7 moderate or high quality [17,21,23,24,26,28–31], which varied RCT + observational studies 8 slightly depending on the outcome and the vaccines assessed. AMSTAR scores According to six reviews [17,21,24,29–31] the influenza vaccine High 3 Moderate 6 does not increase the risk of abortion or stillbirth, and could even Low 2 have a moderate protective effect for stillbirth [17,29]. Specifically, Critically low 6 one review reported a 20% reduction in stillbirth after H1N1 vacci- Type of vaccine nation, pooling data from 11 observational studies [29]. Two Hepatitis B 1 Influenza 11 reviews reported no effects from the Tdap vaccine for the risk of Meningococcal polysacharide 1 stillbirth [26,28]. Two reviews did not find an increase of neonatal Tetanus-diphtheria/pertussis-containing [24,30] or maternal [24] deaths as a result of influenza vaccination. Yellow fever 4 The two reviews of Tdap vaccines reached similar conclusions 1 [26,28]. Outcomes assessed Primary outcomes Although two reviews found an inconclusive impact of influ- Abortion/Miscarriage 9 enza vaccines on congenital malformations and called for more rig- Stillbirth 13 orous studies in women vaccinated during their first trimester Chorioamnionitis 3 [28,29], three additional reviews concluded that the seasonal inac- Congenital anomalies 9 Microcephaly 0 tivated or monovalent H1N1 pandemic vaccines during pregnancy Neonatal Death 8 do not increase the risk of congenital defects [23,30,31]. The con- Neonatal Infection 0 clusions from the two reviews of the Tdap vaccine were dis- Preterm Birth 15 crepant; one found that the risk of malformations associated Secondary outcomes with the vaccine is very uncertain due to the scarce data available Low birth weight 4 Maternal death 1 [26], but the other concluded that this immunization does not Small for gestational age 8 likely increase the risk of congenital anomalies [28]. Five reviews reached heterogeneous conclusions regarding the impact of the influenza vaccine on pre-term birth [17,24,29–31], most of them reporting an inconclusive effect, with different esti- term birth (15/17) and stillbirth (13/17), but some pre-specified mates depending on the trimester in which women were vacci- outcomes of interest (microcephaly, neonatal infection and mater- nated [31]. On the other hand, two reviews on the Tdap vaccine nal death) were rarely assessed. The reviews used different defini- reported an inconclusive effect for this outcome [26,28]. tions for the outcomes of interest, which are compiled in the The risk of chorioamnionitis was only discussed in the reviews Appendix (Table S4). In addition, we did not find reviews about that assessed Tdap vaccines. One review concluded that this vacci- the safety of poliomyelitis and rabies vaccines. One review nation during pregnancy may increase the risk of chorioamnionitis assessed hepatitis B and meningococcal polysaccharide vaccines according to the results of one observational study, which showed [15]. The Appendix, Table S5, shows the assessment of the method- an absolute difference of 1% in the rate of this outcome between ological quality of included reviews. The reviews varied hugely in vaccinated (6%) and non-vaccinated women (5.5%) [26]. Another terms of their quality. Although nine reviews have a high review found no effect [27] and a third review concluded that [24,30,31] or moderate quality [17,21,23,26,28,29], six were con- the risk for this outcome is inconclusive due to inconsistent effects sidered to have critically low quality, because they did not perform estimates from available studies which used different study defini- an assessment of the risk of bias in the included primary studies tions of chorioamnionitis including clinical cases, histologically [15,16,18,22,25,27] and had incomplete searches [22,25,27]. A confirmed cases, or both [28]. We only included two reviews for Table 2 Hepatitis B vaccination safety: findings from included reviews. Outcomes Review/ Studies Vaccine/ Findings reported in the review Conclusions Quality included/ Comparator Sample Abortion Makris Randomized HB recombinant Review authors did not conduct a meta-analysis Inconclusive effect Preterm 2012 [15] controlled vaccine and described findings narratively. HB vaccine during pregnancy may have no effect Birth Critically trials (1) HB vaccine Major events after recombinant HB vaccine were on major safety outcomes. Event rates from Stillbirth low qual- 100 (52 derived from rare. One trial did not report side effects. A cohort included studies can be interpreted as close to ity vaccinated, 48 plasma of chronic in 168 pregnant women reported a similar those expected within general population. controls) HB virus carriers preterm birth rate to those in the general population. In a cohort in 16 women vaccinated Observational Control after an in vitro fertilization, one miscarriage was studies (5) Recombinant HB reported. 281 vaccine (different Only one cohort study in 72 women that received vaccinated doses) the HB vaccine derived from plasma of virus Placebo, no carriers reported a 1% stillbirth rate. comparator HB: hepatitis B. D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 B15 Table 3 Meningococcal disease vaccination safety: findings from included reviews. Outcomes Review/ Studies Vaccine/ Findings reported in the review Conclusions Quality included/ Comparator Sample Preterm Makris Randomized Meningococcal Review authors did not conduct a meta-analysis Inconclusive effect Birth 2012 [15] controlled polysaccharide and described findings narratively. Meningococcal polysaccharide vaccine during Stillbirth Critically trials (3) vaccine (MPV) One trial reported a single preterm delivery pregnancy may have no effect on major safety Neonatal low qual- 420 (207 (1.3%) and two low birth weight deliveries outcomes. Event rates from included studies can be Death ity vaccinated, Control (2.6%) in a group of 75 MPV vaccinated women. interpreted as close to those expected within 213 controls) Pneumococcal Two additional trials reported stillbirth (range general population. Polysaccharide 4–8%) and neonatal death rates (range 0–1.3%) Observational Vaccine or unclear that were within expected rates, according to studies (3) comparator researchers. 128 Three additional cohort studies did not report vaccinated major events. Table 4 Yellow fever vaccination safety: findings from included reviews. Outcomes Review/ Studies Vaccine/ Findings reported in the review Conclusions Quality included/ Comparator Sample Abortion Thomas Observational Yellow Review authors did not conduct a meta-analysis Inconclusive effect 2012 [16] studies (8) fever and described findings narratively. Yellow fever vaccination during pregnancy may Critically vaccination Primary data studies: One cohort reported 11 have no effect on major safety outcomes. low Primary data (17D or miscarriages from 304 pregnancies. Review authors interpreted the rates from studies: 1381 17DD) Secondary data studies: One case-control study included studies as close to those expected Secondary reported that vaccination did not increase the within general population. data studies Control risk for spontaneous abortion (OR: 2.29; 95% CI: 138 Not defined 0.65–8.03). One cohort including 58 pregnancies reported 7 events. Congenital Primary data studies: One study reported 10 anomalies events of malformation in new born exposed in utero to Yellow fever vaccination (3.3% from the total sample; 95% CI: 1.7–14.6%). In addition, one cohort reported no differences in malformations from 304 pregnancies. Secondary data studies. One cohort including 58 pregnancies reported around 4% major malformations. Neonatal death Primary data studies: One cohort reported 2 early neonatal deaths from 304 pregnancies. Preterm birth Primary data studies: One cohort reported 7.8% of premature delivery rate from 304 pregnancies. OR: Odds Ratio. yellow fever [16], and hepatitis B and meningococcal polysaccha- nonexistent for those vaccines indicated in exceptional situations ride vaccines [15] but they had major methodological limitations, (i.e.: Hepatitis B or yellow fever vaccines). We also found differ- as results were based on studies using external comparators with ences in the available evidence according to the type of limited capacity to establish causality effects and thus conclusions maternal-fetal and neonatal outcome. may not be reliable. These two reviews concluded that the vaccines Preterm birth [PTB] is defined by WHO as all births before 37 assessed have an inconclusive effect for abortion and stillbirth, completed weeks of gestation [32]. PTB is a frequent neonatal out- neonatal deaths and preterm births [15,16], and reported that come with an estimated global prevalence ranging from 5 to 18% event rates collected from original studies can be extrapolated to [33] and it was the most frequently reported outcome in the SRs. those expected within general population. Congenital abnormali- Seven SRs with the influenza vaccine found no difference in risk. ties were only assessed in the yellow fever review [16]. Two SRs reported statistically significant protection associated with both seasonal and monovalent pandemic vaccines. On the other hand, four SRs of the Tdap vaccine reporting PTB found no 4. Discussion differences in risk, or were inconclusive [28,26–28]. One of the causes of preterm birth is chorioamnionitis [33], an In this OoSRs we found 17 systematic reviews reporting infection of the amniotic fluid, fetal membranes, placenta, and/or maternal-fetal and neonatal outcomes of interest after immuniza- uterus. Chorioamnionitis can affect 1–2% of full term and 5–10% tion during pregnancy. No major safety concerns with regard to of preterm pregnancies, but is rarely reported to be associated with maternal immunization have been identified. However, differences vaccines to passive pharmacovigilance systems [34]. In one SR, a in the available evidence and its quality have been found. Consis- small but statistically significant increased risk of chorioamnionitis tently with the current recommendations for maternal immuniza- associated with the Tdap vaccine was reported, based on data from tion, influenza and pertussis vaccines were the most studied a single study [risk ratio, 1.19, CI 95%: 1.13–1.26] [26]. However, vaccines [6]. On the other hand, evidence from SRs was almost two SRs were inconclusive or found no difference in the risk of Table 5 B16 Influenza vaccination safety: findings from included reviews. Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Abortion Bratton 2015 [19] Observational studies (4) seasonal or H1N1pdm09 influenza RR: 0.91; 95% CI: 0.68–1.22 (I2 = 39.9%). No effect Low quality vaccination The pooled effect estimate for spontaneous abortion was non 40,746 significant in relation to non vaccinated women. (4336 vaccinated, 36,410 Control No vaccination non-vaccinated) Demichelli 2018 Observational studies (8) 2009/2010 H1N1 monovalent OR adjusted data: 0.75; 95% CI: 0.62–0.90 No effect [30] HR adjusted data: 127,103 pandemic vaccine (5 studies) Seasonal inactivated vaccine or monovalent H1N1 pandemic High quality (27944 vaccinated, 99,159 HR adjusted data : 0.81; 95% CI: 0.63–1.04 vaccines during pregnancy do not increase risk of abortion. non-vaccinated) Control Placebo (3 studies). Observational studies (2) Seasonal influenza vaccine OR unadjusted data: 0.60; 95% CI: 0.41–0.86 OR adjusted data: 486 (243 Control Placebo OR adjusted: 0.80; 95% CI: 0.36–1.78. vaccinated, 243 non- vaccinated) McMillan 2015 Observational studies (10) Monovalent influenza A (H1N1) 2009 Review authors did not conduct a meta-analysis Inconclusive effect [21] vaccine, or (H1N1) 2009 antigen and described findings narratively. Maternal influenza vaccination may not increase risk of Moderate quality 79,796 (16187 vaccinated, containing trivalent vaccine Spontaneous abortion (defined as foetal death prior spontaneous abortion. As effect estimates showed imprecision D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 63,609 controls) to 24 weeks gestation) ranged from HR: 0.45 to OR: and important heterogeneity it is not possible to exclude the Control No vaccination 1.23, with 95% confidence intervals crossing or risk of adverse effects. Studies on vaccination during their first below the null value. trimester should be prioritized. Zhang 2017 [29] Observational studies (3) H1N1 vaccination RR adjusted: 1.04; 95% CI: 0.72–1.52. No effect Moderate quality 63,868 (8025 vaccinated, H1N1 vaccination during pregnancy does not increase the 55,843 controls) Control No vaccination risk of spontaneous abortion. Salam 2015 [7] Randomized controlled trial Trivalent influenza vaccination Miscarriage (24–28 weeks) No effect High quality (1) RR: 0.60; 95% CI: 0.14–2.49. Immunisations during pregnancy with viral influenza vaccines 2116 (1062 vaccinated, Control Placebo likely do not increase risk of foetal death. 1054 controls) Fell 2015 [8] Observational studies (15) Seasonal trivalent inactivated Review authors did not conduct a meta-analysis No effect Low quality influenza (TIV) or monovalent H1N1 and described findings narratively. Influenza immunization during pregnancy likely not 268,777 (69592 vaccinated, influenza vaccine Early foetal death (prior to 20 weeks): Only one out increases risk of foetal death. Available studies had serious 199,185 controls) of eight studies reported a reduced risk of foetal limitations but results within studies were consistent. Control No vaccination death between 9 and 12 weeks of gestation (adjusted HR: 0.74; 95% CI: 0.62–0.88). Outcomes Review/Quality Studies included/ Vaccine/Comparator Findings reported in the review Conclusions Sample Stillbirth Salam 2015 [7] Randomized Trivalent influenza vaccination Stillbirth No effect High quality controlled trial (1) RR: 1.65; 95% CI: 0.73–3.76. Immunisations during pregnancy with viral influenza Control Placebo vaccines likely do not increase risk of foetal death. 2116 (1062 vaccinated, 1054 controls) McMillan 2015 (20) Observational Monovalent influenza A (H1N1) Review authors did not conduct a meta-analysis and described Inconclusive effect Moderate quality studies (13) 2009 vaccine, or (H1N1) 2009 findings narratively. Maternal influenza vaccination may not increase risk of antigen containing trivalent Effect estimates for the risk of foetal death (in the review, those foetal death. As effect estimates showed imprecision and 544,581 (105895 vaccine that occurred after 24 weeks gestation) in women vaccinated an important heterogeneity it is not possible to exclude vaccinated, 438,686 during the first trimester ranged from OR: 0.23 to 2.95, with 95% the risk of adverse effects. controls) Control No vaccination confidence intervals crossing or below the null value. Fell 2015 [8] Observational Seasonal trivalent inactivated Review authors did not conduct a meta-analysis and described No effect Low quality studies (15) influenza (TIV) or monovalent findings narratively. Influenza immunization during pregnancy likely not H1N1 influenza vaccine Foetal death occurring at any time: two cohort studies that increases risk of foetal death. Available studies had 268,777 (69592 assessed the adjuvant monovalent H1N1 vaccine did not show serious limitations but results within studies were vaccinated, 199,185 Control No vaccination an increased risk of foetal loss after vaccination (adjusted HR consistent. controls) 0.79 (95%CI 0.53–1.16) and 0.88 (95%CI 0.66–1.17) respectively) Late foetal death (20 weeks): Some studies did not report adjusted effect measures. High heterogeneity in the rest (RR ranged from 0.44 to 1.44). Giles 2019 [31] Observational Inactivated influenza OR adjusted: 0.84; 95% CI 0.65–1.08. No effect High quality studies (20) vaccination Immunisation during pregnancy with inactivated influenza vaccines likely do not increase risk of foetal 152,713 women Control No vaccination death. vaccinated (145,185 included in risk adjusted analyses) Bratton 2015 [19] Observational Seasonal or H1N1pdm09 RR: 0.73; 95% CI: 0.55–0.96 (I2 = 68%). Moderate/Large benefit Low quality studies (7) influenza vaccination Vaccinated women likely reduced their likelihood of stillbirth. 145,960 Control No vaccination (41301 vaccinated, 104,659 non- vaccinated) Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Stillbirth Galvao 2013 [9] Observational studies (1) Influenza vaccination OR unadjusted data: 0.61; 95% CI: 0.42–0.88. Moderate/Large benefit (continued) Moderate quality Vaccinated women likely reduced their likelihood D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 84,843 (8690 vaccinated, 76,153 Control No vaccination of stillbirth. controls) Meijer 2015 [22] Observational studies (28) Influenza vaccination Review authors did not conduct a meta-analysis and No effect Critically low quality described findings tabulated and narratively. Influenza immunization during pregnancy likely 3876 Control Unclear None of the included studies showed an increase in the reduces risk of this outcome. Despite the stillbirth risk after influenza immunization, but larger inconsistency of results from included studies studies reported a statistically significant reduction of vaccines seem to be safe. the outcome. Zhang 2017 [29] Observational studies (10) H1N1 vaccination HR adjusted: 0.80; 95% CI: 0.69–0.92 (I2 = 7.7%) Moderate/Large benefit Moderate quality vaccination during pregnancy lowered the incidence H1N1 vaccination during pregnancy likely reduces 808,551 (171,906 vaccinated, Control No vaccination of stillbirth by 20%. the incidence of stillbirth. 636,645 controls) Neonatal death Demichelli 2018 [30] Observational studies (3) 2009/2010 H1N1 OR adjusted data: 1.09; 95% CI: 0.40–2.95 No effect High quality monovalent pandemic 2 studies. Seasonal inactivated vaccine or monovalent H1N1 OR adjusted data: 90,679 (8541 vaccine pandemic vaccines during pregnancy do not vaccinated, 82,138 non- increase risk of neonatal death. vaccinated) Control Placebo Seasonal influenza OR unadjusted data: 0.55; 95% CI: 0.35–0.88 vaccine 1 study. Control Placebo Galvao 2013 [9] Observational studies (1) Influenza vaccination OR unadjusted data: 0.55; 95% CI: 0.35–0.88. Small benefit Moderate quality Influenza vaccination during pregnancy may 84,843 (8690 vaccinated, 76,153 Control Placebo or no decrease the risk of neonatal death slightly. controls) vaccination Salam 2015 [7] Randomized controlled trial (1) Trivalent influenza Perinatal death (stillbirth and death during first week of No effect High quality vaccination life) Immunisation during pregnancy with viral 2083 (1041 vaccinated, 1042 RR: 1.32; 95% CI: 0.73–2.38. influenza vaccines likely not increases risk of controls) Control Placebo perinatal death. Outcomes Review/ Studies included/ Vaccine/Comparator Findings reported in the review Conclusions Quality Sample Congenital anomalies Demichelli 2018 [30] Observational studies (6) 2009/2010 H1N1 monovalent pandemic vaccine High quality OR adjusted data: 304,415 (35,017 vaccinated, 269,398 non- Control Placebo OR No effect vaccinated) adjusted Seasonal inactivated (continued on next page) B17 B18 data: 1.11; vaccine or 95% CI: monovalent H1N1 0.99–1.23 pandemic vaccines (6 studies). during pregnancy do Seasonal not increase risk of OR unadjusted data: 0.55; 95% influenza congenital CI: 0.08–3.73 vaccine malformation. (2 studies). Control Placebo McMillan Observational studies Monovalent influenza A Review authors did not conduct a meta-analysis and Inconclusive effect 2015 [21] (12) (H1N1) 2009 vaccine, or described findings narratively. Maternal influenza vaccination may not increase risk of congenital Moderate (H1N1) 2009 antigen Effect estimates for the risk of congenital malformations in malformations. As effect estimates showed imprecision and an quality 297,243 (46,137 containing trivalent vaccine women vaccinated during the first trimester ranged from OR important heterogeneity it is not possible to exclude the risk of vaccinated, 251,106 0.67 to 2.18, with 95% confidence intervals crossing or below adverse effects. Studies in women vaccinated during their first controls) Control No vaccination the null value. trimester should be prioritized. Polyzos Observational studies Inactivated influenza vaccine OR: 0.96; 95% CI: 0.86–1.07 (I2 = 36%). No effect 2015 [23] (15) (seasonal trivalent or Influenza immunization during pregnancy at any trimester does not Moderate monovalent H1N1, adjuvant or increase risk for congenital defects. quality 282,931 non-adjuvant) D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 Control No vaccination Zhang Observational studies H1N1 vaccination OR adjusted: 1.14; 95% CI: 1.01–1.29 Inconclusive effect 2017 [29] (9) H1N1 vaccination during pregnancy may not increase the risk of Moderate Control No vaccination Sensitivity analysis of adjusted estimates after excluding congenital anomalies. Although a sensitivity analysis showed a quality 385,062 (47,483 studies of non adjuvanted H1N1 vaccine: OR adjusted: 1.15; potential increase of this outcome, studies of vaccination during early vaccinated, 337,579 95% CI: 1.02–1.30. pregnancy was not associated with congenital abnormalities. controls) Giles 2019 Observational studies Inactivated influenza OR adjusted: 1.03; 95% CI: 0.99–1.07. No effect [31] (12) vaccination Influenza immunization during pregnancy does not increase risk for High congenital defects. quality 169,829 women Control No vaccination vaccinated (157,601 included in risk adjusted analyses) Outcomes Review/ Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Quality Preterm Demichelli Observational studies (11) 2009/2010 H1N1 OR adjusted data: 0.84; 95% CI: 0.76–0.93 No effect birth 2018 [30] monovalent (7 studies). Seasonal inactivated vaccine or monovalent H1N1 pandemic High qual- OR adjusted data: 279,609 pandemic vaccine vaccines during pregnancy do not increase risk of prematurity. ity (69,190 vaccinated, 210,419 HR adjusted data: 1.11; 95% CI: 0.46 to 2.68 non-vaccinated) Control Placebo (2 studies). HR adjusted data: 16,963 (5263 vaccinated, 11,700 non- vaccinated) 124,755 (59479 vaccinated, Seasonal influenza OR adjusted data: 0.93; 95% CI: 0.82–1.06 65,276 non-vaccinated) vaccine (2 studies). Control Placebo Fell 2015 Observational studies (19) Seasonal trivalent Review authors did not conduct a meta-analysis and described No effect [8] inactivated (TIV) or findings narratively.Most studies (18/19) did not show a harmful Influenza immunization during pregnancy likely not increases Low qual- 633,932 (278,907 vaccinated, monovalent H1N1 effect with a broad range of estimates (from 0.63 to 1.20) risk of preterm birth. Available studies had serious limitations ity 355,025 controls) . Only 1 study reported an increased risk of preterm birth but with a but results within studies were consistent. Control Unclear very imprecise effect estimate. Galvao Randomized controlled trials(1) Influenza Review authors did not conduct a meta-analysis and described Inconclusive effect 2013 [9] 340 (172 vaccinated, 168 vaccination findings tabulated and narratively. The effect of vaccination during pregnancy in the risk of Moderate controls)observational studies For this outcome conflicting results were found across the studies. preterm labour is very uncertain due to the heterogeneity from quality (4) Control Placebo, no effect estimates and their precision. 92,925 (7665 vaccinated, 85,260 vaccination controls) Meijer Observational studies (28) Influenza Review authors did not conduct a meta-analysis and described No effect 2015 [22] vaccination findings tabulated and narratively. Influenza immunization during pregnancy likely not increases Critically 3876 None of the included studies showed an increase in the preterm birth risk of preterm birth. Despite the inconsistency of results from low quality Control Unclear risk after influenza immunization. included studies vaccines seem to be safe. Nunes Observational studies (5) Seasonal trivalent OR: 0.87; 95% CI: 0.77–0.98. Small benefit 2016 [25] 145,638 (62829 vaccinated, inactivated Influenza vaccination during pregnancy likely decreased risk Critically 62,809 controls) Control No of preterm births slightly. low quality vaccination Observational studies (11) A/H1N1pdm09 OR: 0.92; 95% CI: 0.85 to 0.99. 606,644 (270,418 vaccinated, monovalent 336,226 controls) Control No vaccination Outcomes Review/ Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 Quality Preterm birth Salam Randomized controlled trial Haemophilus influenza type RR: 1.28; 95% CI: 0.12–13.86. Inconclusive effect (continued) 2015 [7] (2) B (Hib) vaccination The effect of Hib vaccination during pregnancy in the risk of High 213 (130 vaccinated, 83 preterm labour is very uncertain due to the limitations from the quality controls) Control Placebo evidence and the imprecision of its estimates. Randomized controlled trial Trivalent influenza RR: 0.92; 95% CI: 0.53–1.59. No effect (1) vaccination Immunisation during pregnancy with viral influenza vaccines 2118 (1062 vaccinated, likely not increases risk of preterm labour. 1054 controls) Control Placebo Zhang Observational studies (12) H1N1 vaccination RR adjusted: 0.92; 95% CI: 0.84–1.01. No effect 2017 [29] 484,291 (115,980 H1N1 vaccination during pregnancy does not increase the risk of Moderate vaccinated, 368,311 Control No vaccination preterm delivery. quality controls) Giles Observational studies (26) Inactivated influenza OR adjusted: 0.86; 95% CI: 0.78–0.96 Inconclusive effect 2019 [31] 184,305 women vaccinated vaccination Inactivated influenza vaccine during pregnancy may not reduce High (173,131 included in risk Estimates differed depending on the trimester in which the incidence of preterm births. Considerable uncertainty due to quality adjusted analyses) Control No vaccination women were vaccinated (OR for vaccination during 1st the heterogeneity of effect estimates depending on the trimester trimester 1.08; 95% CI 0.92–1.28; OR for 2nd/3rd in which women were vaccinated. trimester 0.96; 95% CI 0.87–1.06). Low birth weight Nunes Observational studies (2) Seasonal trivalent OR: 0.74; 95% CI: 0.61–0.88; I2 = 0%). Small benefit 2016 [25] inactivated influenza Influenza vaccination during pregnancy likely decreased risk of Critically 20,940 (3781 vaccinated, vaccine preterm births slightly. low qual- 17,159 controls) ity Control No vaccination Observational studies (7) A/H1N1pdm09 monovalent OR = 0.88; 95% CI: 0.79–0.98 (I2 = 61.6%). 395,045 (215,684 vaccine vaccinated, 179,361 controls) Control No vaccination Giles Observational studies (12) Inactivated influenza OR adjusted: 0.82; 95% CI: 0.76–0.89 Inconclusive effect 2019 [31] 84,314 women vaccinated vaccination Estimates differed depending on the trimester in which Inactivated influenza vaccine during pregnancy may not reduce High (81,609 included in risk women were vaccinated (OR for vaccination during 1st the incidence of low birth weight births. Considerable uncertainty quality adjusted analyses) Control No vaccination trimester 1.00; 95% CI 0.80–1.24; OR for 2nd/3rd due to the heterogeneity of effect estimates depending on the trimester 0.97; 95% CI 0.71–1.32). trimester in which women were vaccinated. Maternal death Salam Randomized controlled trial trivalent influenza RR: 4.96; 95% CI: 0.24–103.24. No effect 2015 [7] (1) vaccination Vaccination during pregnancy (with viral vaccines) likely not High 2116 (1062 vaccinated, increases risk of maternal death. quality 1054 controls) Control Placebo B19 B20 Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Small for gestational age Galvao 2013 [9] Randomized controlled Influenza vaccination Review authors did not conduct a meta-analysis and Inconclusive effect Moderate quality trials (1) described findings tabulated and narratively. The effect of vaccination during pregnancy in the 340 (172 vaccinated, 168 Control Placebo, no For this outcome conflicting results were found risk of preterm labour is very uncertain due to the controls) vaccination across the studies. heterogeneity from effect estimates and their precision. Observational studies (2) 89,011 (9268 vaccinated, 79,743 controls) Nunes 2016 [25] Observational studies (3) Seasonal trivalent OR: 0.95; 95% CI: 0.86–1.06. No effect Critically low 136,048 (61,335 vaccinated, inactivated influenza Seasonal or A/H1N1pdm09 monovalent vaccines quality 74,713 controls) vaccine during pregnancy may not increase the risk of small for gestational age births. Control No vaccination Observational studies (9) A/H1N1pdm09 monovalent OR: 0.96; 95% CI: 0.90–1.03. 292,105 (74,463 vaccinated, vaccine 217,642 controls) Control No vaccination Meijer 2015 [22] Observational studies (6) Influenza vaccination Review authors did not conduct a meta-analysis and No effect Critically low 1766 described findings tabulated and narratively. Influenza immunization during pregnancy likely D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 quality Control Unclear None of the included studies showed an increase in not increases risk of this outcome. Despite the the risk of a birth small for gestational age after inconsistency of results from included studies influenza immunization. vaccines seem to be safe. Zhang 2017 [29] Observational studies (7) H1N1 vaccination OR adjusted: 0.98; 95% CI: 0.91–1.06 No effect Moderate quality 313,458 H1N1 vaccination during pregnancy does not Control No vaccination increase the risk of small for gestational age births. Giles 2019 [31] Observational studies (17) Inactivated influenza OR adjusted: 0.99; 95% CI: 0.94–1.04 (I2 = 23.2%). No effect High quality 176,486 women vaccinated vaccination Estimates differed depending on the trimester in H1N1 vaccination during pregnancy may not (164,966 included in risk which women were vaccinated (OR for vaccination increase the risk of small for gestational age births. adjusted analyses) Control No vaccination during 1st trimester 0.90; 95% CI 0.66–1.24; OR for Imprecision in estimates increases depending on the 2nd/3rd trimester 0.96; 95% CI 0.89–1.04). trimester in which women were vaccinated. HR: Hazard Ratio; OR: Odds Ratio; RR: Risk Ratio. Table 6 Tetanus-diphtheria and pertussis-containing vaccination safety: findings from included reviews. Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Neonatal death Furuta 2017 [26] Observational study (1) Tdap vaccine Low event rate in both groups No effect Moderate quality 24,708 (6185 vaccinated, 18,523 (<0.1%; RR: 1.0; 95% CI: 0.2–4.9). Tdap immunization during controls) Control No vaccination pregnancy likely not increases risk of neonatal death. McMillan 2017 [28] Observational studies (2) Tdap or Tdap-IPV Review authors did not conduct a No effect Moderate quality 32,086 (13,337 vaccinated, 18,749 Control No vaccination meta-analysis and described findings Tdap immunization during controls) tabulated and narratively. pregnancy likely not increases risk Two retrospective cohorts did not of neonatal death. report differences in the rate of neonatal death (rates lower to 0.1%). Abortion Rivero-Santana 2014 [15] Observational study (1) Tdap vaccine Review authors did not conduct a Inconclusive effect Critically low quality 132 meta-analysis and described findings The effect of Tdap vaccination during Control No comparator tabulated and narratively. pregnancy in the risk of abortion is Data obtained during 5 years from an very uncertain due to the scarce data adverse event register in 132 women available. (mostly vaccinated during the first trimester) reported a 16.3% D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 spontaneous miscarriage rate. Furuta 2017 [26] Randomized controlled trial (1) Tdap vaccine Lower rate within vaccinated women Inconclusive effect Moderate quality 48 (33 vaccinated, 15 controls) Control No vaccination (3% vs 13%; RR: 0.2; 95% CI: 0.0–2.3) The effect of Tdap vaccination during but no conclusion can be reached due pregnancy in the risk of to the small sample size. malformations is very uncertain due to the scarce data available. Gkentzi 2017 [27] Observational study (1) Tdap vaccine Low rate (1%) of malformations No effect Critically low quality 7378 (7152 vaccinated, 226 controls) between immunized women and Tdap immunization during Control No vaccination their controls. pregnancy likely not increases risk of congenital malformations. McMillan 2017 [28] Randomized controlled trials (1) 48 Tdap or Tdap-IPV Review authors did not conduct a No effect Moderate quality (33 vaccinated, 15 controls) meta-analysis and described findings Tdap immunization during Control No vaccination tabulated and narratively. pregnancy likely not increases risk Observational studies (4) One small trial did not show of congenital malformations. 9926 (8456vaccinated, 1470 controls) differences between vaccinated and unvaccinated women (OR: 0.20; 95% CI: 0.02–2.44). Three additional retrospective cohorts showed similar results (OR estimates range from 0.80 to 0.91). Outcomes Review Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Stillbirth Furuta 2017 [26] Randomized controlled trials (1) Tdap vaccine One trial did not report any case of No effect Moderate quality 103 (52 vaccinated, 51 controls) intrauterine death after 24 weeks Tdap immunization during Observational studies (2) Control No vaccination gestation. Two cohort studies did not pregnancy likely not increases risk 25,398 (6323 vaccinated, 19,075 controls) show differences in stillbirth rates of stillbirth. (pooled RR: 0.82; 95% CI: 0.44–1.54; I2:0%). Gkentzi 2017 [27] Observational study (1) Tdap vaccine Low event rate (0.2%) reported No effect Critically low quality 24,708 (6185 vaccinated, 18,523 controls) without differences between Tdap immunization during Control No vaccination vaccinated women and controls (RR pregnancy likely not increases risk 0.69; 95% CI: 0.23–1.62). These of stillbirth. results were confirmed in other cohort studies. McMillan 2017 [28] Observational studies (3) Tdap or Tdap-IPV Review authors did not conduct a No effect Moderate quality meta-analysis and described findings Tdap immunization during 32,776 (13,475 vaccinated, 19,301 controls) Control No vaccination tabulated and narratively. pregnancy likely not increases risk (continued on next page) B21 B22 Three retrospective cohorts did not of stillbirth. report differences in the rate of stillbirths. Congenital anomalies Furuta 2017 [26] Randomized controlled trial (1) Tdap vaccine Lower rate within vaccinated women Inconclusive effect Moderate quality 48 (33 vaccinated, 15 controls9 (3% vs 13%; RR: 0.2; 95% CI: 0.0–2.3) The effect of Tdap vaccination during Control No vaccination but no conclusion can be reached due pregnancy in the risk of to the small sample size. malformations is very uncertain due to the scarce data available. Gkentzi 2017 [27] Observational study (1) Tdap vaccine Low rate (1%) of malformations No effect Critically low quality 7378 (7152 vaccinated, 226 controls) between immunized women and Tdap immunization during Control No vaccination their controls. pregnancy likely not increases risk of congenital malformations. McMillan 2017 [28] Randomized controlled trials Tdap or Tdap-IPV Review authors did not conduct a No effect Moderate quality 48 (33 vaccinated, 15 controls) meta-analysis and described findings Tdap immunization during Control No vaccination tabulated and narratively. pregnancy likely not increases risk Observational studies (4) One small trial did not show of congenital malformations. 9926 (8456 vaccinated, 1470 controls) differences between vaccinated and unvaccinated women (OR 0.20; 95% CI: 0.02–2.44). Three additional retrospective cohorts showed similar results (OR estimates range from 0.80 D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 to 0.91). Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Chorio-amnionitis Furuta 2017 [26] Observational study (1) Tdap vaccine One observational study showed a 6% Small impact Moderate quality 123,494 (26,229 vaccinated, 972,653 controls) rate in vaccinated women compared Tdap vaccination during pregnancy Control No vaccination to a 5.5% rate in controls (adjusted may increase the risk of RR: 1.19; 95% CI: 1.13–1.26). Review chorioamnionitis slightly. authors highlighted that the study did not linked this highest rate with an increase in the risk of preterm birth. Gkentzi 2017 [27] Observational study (1) Tdap vaccine Low and comparable rate of No effect Critically low 7378 (7152 vaccinated, 226 controls) chorioamnionitis between Tdap immunization during quality Control No vaccination immunized women and their pregnancy likely not increases risk controls (6% vs 4%, p = 0.3). of chorioamnionitis. McMillan 2017 Observational studies (3) Tdap or Tdap-IPV Review authors did not conduct a Inconclusive effect [28] 132,631 (34,490 vaccinated, 98,141 controls) meta-analysis and described findings The effect of Tdap vaccination during Moderate quality Control No vaccination tabulated and narratively. pregnancy in the risk of Although one observational study chorioamnionitis is uncertain due to showed a greater rate in vaccinated the inconsistency in effect estimates. women (6% in vaccinated vs 5.5% in controls; adjusted RR: 1.19; 95% CI: 1.13–1.26), two additional cohorts did not show differences (OR around 1.50 in both studies). Low birth weight Furuta 2017 [26] Observational study (1) Tdap vaccine Similar event rate between No effect Moderate quality vaccinated and non-vaccinated Tdap immunization during 24,708 (6185 vaccinated, 18,523 controls) Control No vaccination women (2% vs 1.7%; RR: 1.2; 95% CI: pregnancy likely not increases risk 1.0 to 1.5). of low birth weight labours. McMillan 2017 Observational studies (2) Tdap or Tdap-IPV Review authors did not conduct a No effect [28] meta-analysis and described findings Tdap immunization during Moderate quality 26,467 (7294 vaccinated, 19,173 controls) Control No vaccination tabulated and narratively. pregnancy likely not increases risk Two retrospective cohorts did not of low birth weight labours. report differences in the rate of low birth weight (<2500 g) labours (range of rates from 2% to 9%). Outcomes Review/Quality Studies included/Sample Vaccine/Comparator Findings reported in the review Conclusions Preterm birth Furuta 2017 [26] Randomized controlled trials (2) Tdap Results differed according the study Inconclusive effect Moderate quality 151 (85 vaccinated, 66 controls) design: two trials showed a 3% rate The effect of Tdap vaccination during Control No both in vaccinated and unvaccinated pregnancy in the risk of preterm birth Observational studies (2) vaccination women (pooled RR: 0.86; 95% CI: is uncertain due to the scarce data 124,133 (26,363 vaccinated, 97,770 controls) 0.104–5.21; I2 = 0%), but two available and the discordant results additional cohort studies showed a from trials and observational studies. lower rate within immunized women (5.8% vs 7.8%; pooled RR: 0.75; 95% CI: 0.75–0.79; I2 = 0%). Gkentzi 2017 [27] Observational study (1) Tdap vaccine Lower rate of births before 37 Moderate benefit Critically low quality 7378 gestation weeks in vaccinated Vaccinated women may reduce their (7152 vaccinated, 226 controls) Control No women (6% vs 12%, p < 0.001). These likelihood of preterm birth vaccination results were confirmed in other cohort studies. McMillan 2017 [28] Randomized controlled trials (1) 48 (33 vaccinated, 15 Tdap or Tdap-IPV Review authors did not conduct a Inconclusive effect Moderate quality controls) meta-analysis and described findings The effect of Tdap vaccination during Control No tabulated and narratively. pregnancy in the risk of preterm birth Observational studies (4) vaccination With the exception of one is uncertain due to the discordant 133,270 (34,624 vaccinated, 98,646 controls) retrospective cohort that showed a results from available studies. lower rate in vaccinated women (6% vs 12%; OR: 0.47; 95% CI: 0.31–0.71), D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 three additional cohorts and one small trial did not show differences. Rivero-Santana 2014 Observational study (1) Tdap vaccine Review authors did not conduct a Inconclusive effect [15] 132 meta-analysis. The effect of Tdap vaccination during Critically low quality Control No Data obtained during 5 years from an pregnancy in the risk of preterm birth comparator adverse event register in 132 women is very uncertain due to the scarce (mostly vaccinated during the first data available. trimester) reported a 1.5% premature birth rate. Small for gestational Furuta 2017 [26] Observational study (1) Tdap vaccine The study showed an 8% event rate in No effect age Moderate quality 123,494 (2214 vaccinated, 8086 controls) both groups (RR: 1.0, 95% CI: 1.0–1.1). Tdap immunization during Control No pregnancy likely not increases risk vaccination of small for gestational age births. Gkentzi 2017 [27] Observational study (1) Tdap vaccine Lower rate of small for gestational Small benefit Critically low quality 7378 (7152 vaccinated, 226 controls) age births in vaccinated women (10% Tdap vaccination during pregnancy Control No vs 15%, p = 0.032). These results were likely decreased risk of small for vaccination confirmed in other cohort studies. gestational age births slightly. McMillan 2017 [28] Observational studies (3) Tdap or Tdap-IPV Review authors did not conduct a No effect Moderate quality meta-analysis and described findings Tdap immunization during 132,631 (34,490 vaccinated, 98,141 controls) Control No tabulated and narratively. pregnancy likely not increases risk vaccination Three retrospective cohorts did not of small for gestational age labours. report differences in the rate of small for gestational age labours. Tdap: tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis; Tdap-IPV: Tdap with inactivated poliomyelitis antigens; OR: Odds Ratio; RR: Risk Ratio. B23 B24 D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 chorioamnionitis over the background rate, respectively [27,28]. sents less than a half of the cumulative overall evidence available Further studies have reported inconsistent findings: in later retro- for stillbirth and influenza immunization. Moreover, the only SR spective cohort studies no risk differences were found between the that reported hepatitis vaccination safety regarding stillbirth was group that received Tdap, and those that did not, [35,36], whereas inconclusive due to the insufficient evidence provided by compar- in a larger retrospective cohort study a slightly higher risk of ative cohorts with less than 100 patients that did not find maternal chorioamnionitis associated with the Tdap vaccine was increased risk associated with vaccination [15]. However, much found [37]. Nevertheless, the excess of chorioamnionitis was not more data is needed and vaccination should be limited to pregnant accompanied by an increase of adverse infant clinical outcomes women who are identified as being at risk for HBV infection during such as preterm birth, a major sequela of chorioamnionitis [37]. pregnancy, according to current recommendations [6]. The fetus is considered to be most at risk for congenital anoma- A reduction in the infant (<1 year) case fatality rate has been lies during embryogenesis and the first trimester of pregnancy observed in the ecological analyses of LAC countries that intro- [28]. However, further homogenous studies reporting congenital duced maternal immunization for pertussis during pregnancy anomaly outcomes after first-trimester vaccination have been sug- [40]. However, no difference in risk was found for neonatal death gested. [28]. Current guidelines recommend pregnant women to be and Tdap vaccination in two SRs [26,28]. Similarly, two SRs of vaccinated against influenza at any stage of gestation, including the influenza vaccine found no risk differences in the rate of neona- the first trimester [6]. From the five SRs that reported the outcome tal death [20,30]. A protective effect was reported for seasonal with influenza vaccination, three SRs found no differences in risk influenza vaccines in one SR [17]; however, conclusions were and the two other SRs were inconclusive. The marginally signifi- drawn from data of one single study. In a recently published large cant increase in risk reported with the influenza vaccine in the cohort study from administrative Canadian databases, no differ- pooled cohort studies of H1N1 vaccine was considered inconsistent ences in under-5 mortality were found in vaccinated women [30]. who received pandemic H1N1 influenza vaccination during preg- One high quality SR supported that both seasonal and 2009 nancy (vs non vaccinated) [41]. pandemic vaccines during pregnancy had no significant effect on Although it is estimated that 15% to 20% of all births worldwide abortion, based on observational data from more than 250,000 are LBW, representing more than 20 million births a year [42], only women [30]. A similar conclusion was found in five other SRs four SRs reported this adverse outcome [25,26,28,31]. No effect [19,20,24,29]. However, one SR was inconclusive and, although was found in two SRs of the Tdap vaccination [26,28]. A beneficial considered unlikely, the risk of abortion cannot be excluded, due effect of influenza vaccination was reported by a SR of critically to the observational nature of the evidence [21]. Much less evi- low quality [25], but another SR was inconclusive [31]. Moreover, dence is available for the SRs of Tdap vaccines. Two SRs of Tdap SGA babies are at higher risk of early mortality, especially in vaccines found no effect [27,28], and two others were inconclusive low-and middle-income countries [43]. No differences in risk were [15,26]. More studies investigating women vaccinated during their found in four SRs with influenza vaccine [25,29], whereas one SR first trimester, rather than combining all trimesters, are needed to was inconclusive [17,22]. Similarly, no effect was observed in obtain more precise estimates on the risk of spontaneous abortion two SRs with Tdap vaccination. Moreover, one low quality SR [21]. found a small benefit of the vaccination [26–28]. Stillbirth incidence varies notably among countries, ranging Limitations from 1.2 per 1000 [1.0–1.5] in Iceland to 56.3 per 1000 [32.3– There are some limitations in our study which must be men- 98.2] in South Sudan [38]. Stillbirth was analyzed in eight SRs of tioned. As in any SR publication, bias cannot be ruled out. More- influenza and three SRs of Tetanus, Diphtheria, and Pertussis Vac- over, a considerable overlap in evidence was found in the cination safety. Influenza immunization was associated with a primary studies included in the SRs of influenza and pertussis decreased risk of stillbirth in three SRs [17,19,29] in which still- immunization. Although we searched for routine maternal-fetal birth was reported. In one SR, the pooled estimate of seven obser- and neonatal outcomes, none of the SRs examined and reported vational studies showed a 27 to 31% decrease in stillbirths (defined all (of the) outcomes of interest. In addition, the sample population as fetal loss after 20 or 22 weeks of gestational age) associated with considered in some SRs could be insufficient to address the risk for influenza immunization and H1N1pdm09 immunization respec- the less frequent outcomes. Furthermore, the main source of evi- tively, versus no vaccination. Notwithstanding, no studies of triva- dence in the SRs was observational studies, which are prone to bias lent seasonal vaccines met inclusion criteria to be considered in the and confounding, a limitation also recognized by the WHO [44]. SR. Consistently, other SRs found that influenza immunization low- Residual confounding can play a major role in the small size effects ered the incidence of stillbirth by 20% [29]. Studies finding protec- suggested by the SRs [45]. On the other hand, pregnant women can tive associations between stillbirth and vaccination have been be at higher risk of infections, such as severe influenza [46], and large retrospective cohort studies including more than 50,000 pregnancies complicated by infection are also at higher risk of women [19]. Conclusions were more conservative in other SRs adverse pregnancy outcomes, including congenital anomalies, due to the lack of RCTs statistically powered enough to address this spontaneous abortion and stillbirth, preterm birth and LBW [47]. outcome [22]. More RCTs reporting stillbirth are needed to draw More evident benefits of vaccination would have been expected firm conclusions. In addition, more homogeneous outcome defini- in low- and middle-income countries in which related morbidity tions of stillbirth would also be necessary to avoid misclassification and mortality is higher [48,49]. However, since most of the evi- with abortion [31,39]. Further studies should also consider the evi- dence included in the SRs comes from studies in high-income dence gaps identified to date, including the timing of vaccination, countries, generalization of the results is limited. Lastly, pooled data on exposure to the trivalent seasonal influenza vaccine, and efficacy results from clinical trials have supported the label exten- larger sample sizes [19]. On the other hand, three SRs found no sion of the indication of influenza vaccine to pregnant women effect of Tetanus, Diphtheria, and Pertussis vaccination on still- [50,51], however the latest safety evidence from these trials has birth. Pooled estimates from 25,398 babies (6323 from vaccinated not yet been considered in (the) updated MA. mothers) found no differences in risk versus non-vaccinated In conclusion, the PAHO/WHO recommendations for Maternal women (28). Although the estimates for this outcome do not indi- Immunization are supported by current evidence. Definite risks cate an association with vaccine administration, it has to be noted were not identified for any vaccine or outcome of interest. In spite that the available evidence in the SRs comes from only three retro- of this, better evidence is needed for all outcomes and vaccines that spective cohort studies plus one small RCT (n = 151) which repre- are broadly administered. The findings from our overview suggest D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 B25 that influenza vaccines do not increase the risk for the outcomes of [11] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ (Clin Res interest. Pertussis vaccines likely do not increase the risk of Ed) 2009;339:b2535. adverse outcomes, but more uncertainties exist on their net effect. [12] Garner P, Hopewell S, Chandler J, MacLehose H, Schünemann HJ, Akl EA, et al. The available cumulative evidence in the SRs to support vaccine Panel for updating guidance for systematic reviews (PUGs). When and how to safety was based mostly on observational data. More RCT with ade- update systematic reviews: consensus and checklist. BMJ. 2016;354:i3507. [13] Shea BJ, Reeves BC, Wells G, Thuku M, Hamel C, Moran J, et al. AMSTAR 2: a quate reporting of maternal-fetal outcomes and larger high-quality critical appraisal tool for systematic reviews that include randomised or non- observational studies are needed, especially in low- and middle- randomised studies of healthcare interventions, or both. BMJ (Clinical Res Ed) income countries. This could be addressed improving the efficiency 2017;358:j4008. [14] Pieper D, Antoine SL, Mathes T, Neugebauer EA, Eikermann M. Systematic of randomised trials adopting innovative methodological designs review finds overlapping reviews were not mentioned in every other (ie, adaptive design trials, basket trials, registry trials, umbrella overview. J Clin Epidemiol 2014;67(4):368–75. protocols) and reducing the administrative complexity among [15] Makris MC, Polyzos KA, Mavros MN, Athanasiou S, Rafailidis PI, Falagas ME. Safety of hepatitis B, pneumococcal polysaccharide and meningococcal other potential measures [52]. polysaccharide vaccines in pregnancy: a systematic review. Drug Saf 2012;35(1):1–14. [16] Thomas RE, Lorenzetti DL, Spragins W, Jackson D, Williamson T. The safety of CRediT authorship contribution statement yellow fever vaccine 17D or 17DD in children, pregnant women, HIV+ individuals, and older persons: systematic review. Am J Trop Med Hygiene Diego Macias Saint-Gerons: Conceptualization, Writing - orig- 2012;86(2):359–72. [17] Galvao TF, Silva MT, Zimmermann IR, Lopes LA, Bernardo EF, Pereira MG. inal draft, Writing - review & editing. Iván Solá Arnau: Conceptu- Influenza vaccination in pregnant women: a systematic review. ISRN Prevent alization, Methodology, Writing - review & editing. Bremen De Med 2013;2013:879493. Mucio: Methodology, Writing - original draft. Ingrid Arévalo-Ro [18] Rivero-Santana A, Cuellar-Pompa L, Sanchez-Gomez LM, Perestelo-Perez L, Serrano-Aguilar P. Effectiveness and cost-effectiveness of different dríguez: Methodology. Alicia Alemán: Writing - original draft. immunization strategies against whooping cough to reduce child morbidity José Luis Castro: Conceptualization, Supervision. Alba María and mortality. Health Policy (Amsterdam, Netherlands) 2014;115(1):82–91. Ropero Álvarez: Conceptualization, Writing - original draft, Super- [19] Bratton KN, Wardle MT, Orenstein WA, Omer SB. Maternal influenza immunization and birth outcomes of stillbirth and spontaneous abortion: a vision, Writing - review & editing. systematic review and meta-analysis. Clin Infect Dis: Off Publ Infect Dis Soc Am 2015;60(5):e11–9. Declaration of Competing Interest [20] Fell DB, Platt RW, Lanes A, Wilson K, Kaufman JS, Basso O, et al. Fetal death and preterm birth associated with maternal influenza vaccination: systematic review. BJOG: Int J Obstet Gynaecol 2015;122(1):17–26. The authors declare that they have no known competing finan- [21] McMillan M, Porritt K, Kralik D, Costi L, Marshall H. Influenza vaccination cial interests or personal relationships that could have appeared during pregnancy: a systematic review of fetal death, spontaneous abortion, and congenital malformation safety outcomes. Vaccine 2015;33(18):2108–17. to influence the work reported in this paper. The opinions expressed [22] Meijer WJ, van Noortwijk AG, Bruinse HW, Wensing AM. Influenza virus in this article are those of the authors and do not necessarily reflect infection in pregnancy: a review. Acta Obstet Gynecol Scand 2015;94 the views of the Pan American Health Organisation (PAHO), its (8):797–819. [23] Polyzos KA, Konstantelias AA, Pitsa CE, Falagas ME. Maternal influenza Board of Directors, or the countries they represent. vaccination and risk for congenital malformations: a systematic review and meta-analysis. Obstet Gynecol 2015;126(5):1075–84. [24] Salam RA, Das JK, Dojo Soeandy C, Lassi ZS, Bhutta ZA. Impact of Haemophilus Appendix A. Supplementary material influenzae type B (Hib) and viral influenza vaccinations in pregnancy for improving maternal, neonatal and infant health outcomes. Cochrane Database Supplementary data to this article can be found online at System Rev 2015(6):Cd009982. https://doi.org/10.1016/j.vaccine.2020.07.048. [25] Nunes MC, Aqil AR, Omer SB, Madhi SA. The effects of influenza vaccination during pregnancy on birth outcomes: a systematic review and meta-analysis. Am J Perinatol 2016;33(11):1104–14. References [26] Furuta M, Sin J, Ng ESW, Wang K. Efficacy and safety of pertussis vaccination for pregnant women – a systematic review of randomised controlled trials and observational studies. BMC Pregnancy Childbirth 2017;17(1):390. [1] Lacriktz E, Stergachis A, Stepanchack M. Maternal immunization safety [27] Gkentzi D, Katsakiori P, Marangos M, Hsia Y, Amirthalingam G, Heath PT, et al. monitoring in low- and middle-income countries: a roadmap for program Maternal vaccination against pertussis: a systematic review of the recent development. Building an approach that is practical, affordable, and literature. Arch Dis Child Fetal Neonatal Ed 2017;102(5):F456–63. sustainable Bill and Melinda Gates Foundation & Global Alliance to Prevent [28] McMillan M, Clarke M, Parrella A, Fell DB, Amirthalingam G, Marshall HS. Prematurity and Stillbirth; 2017. Available from: http://apps.who.int/ Safety of tetanus, diphtheria, and pertussis vaccination during pregnancy: a medicinedocs/en/m/abstract/Js23275en/. systematic review. Obstet Gynecol 2017;129(3):560–73. [2] Lambach P, Hombach J, Ortiz JR. A global perspective of maternal influenza [29] Zhang C, Wang X, Liu D, Zhang L, Sun X. A systematic review and meta-analysis immunization. Vaccine 2015;33(47):6376–9. of fetal outcomes following the administration of influenza A/H1N1 [3] Ingrasciotta Y, Cutroneo PM, Marciano I, Giezen T, Atzeni F, Trifiro G. Safety of vaccination during pregnancy. Int J Gynaecol Obstet 2017. biologics, including biosimilars: perspectives on current status and future [30] Demicheli V, Jefferson T, Ferroni E, Rivetti A. Di Pietrantonj C. Vaccines for direction. Drug safety; 2018. preventing influenza in healthy adults. Cochrane Database System Rev 2018. [4] Onakpoya IJ, Heneghan CJ, Aronson JK. Post-marketing withdrawal of 462 [31] Giles ML, Krishnaswamy S, Macartney K, Cheng A. The safety of inactivated medicinal products because of adverse drug reactions: a systematic review of influenza vaccines in pregnancy for birth outcomes: a systematic review. Hum the world literature. BMC Med 2016;14:10. Vaccin Immunother 2019;15(3):687–99. [5] Global Advisory Committee on Vaccine Safety. Safety of Immunization during [32] Vogel JP, Chawanpaiboon S, Watananirun K, Lumbiganon P, Petzold M, Moller Pregnancy A review of the evidence: World Health Organization; 2014. AB, et al. Global, regional and national levels and trends of preterm birth rates Available from: http://www.who.int/vaccine_safety/publications/safety_ for 1990 to 2014: protocol for development of World Health Organization pregnancy_nov2014.pdf. estimates. Reprod Health 2016;13:76. [6] Pan American Health Organization. The Maternal and Neonatal Immunization [33] Lee SM, Park JW, Kim BJ, Park CW, Park JS, Jun JK, et al. Acute histologic Field Guide for Latin America and the Caribbean recommends the chorioamnionitis is a risk factor for adverse neonatal outcome in late preterm administration of Influenza (inactivated) and Tetanus/diphtheria vaccines. birth after preterm premature rupture of membranes. PLoS ONE 2013;8(12): Washington, D.C.; 2017. Available from: http://iris.paho.org/xmlui/bitstream/ e79941. handle/123456789/34150/9789275119501-eng.pdf. [34] Datwani H, Moro PL, Harrington T, Broder KR. Chorioamnionitis following [7] Chandler RE. Safety concerns with HPV vaccines continue to linger: are current vaccination in the vaccine adverse event reporting system. Vaccine 2015;33 vaccine pharmacovigilance practices sufficient?. Drug Saf 2017;40 (27):3110–3. (12):1167–70. [35] Morgan JL, Baggari SR, McIntire DD, Sheffield JS. Pregnancy outcomes after [8] Laporte JR. Fifty years of pharmacovigilance - medicines safety and public antepartum tetanus, diphtheria, and acellular pertussis vaccination. Obstet health. Pharmacoepidemiol Drug Saf 2016;25(6):725–32. Gynecol 2015;125(6):1433–8. [9] Smith V, Devane D, Begley CM, Clarke M. Methodology in conducting a [36] Griffin JB, Yu L, Watson D, Turner N, Walls T, Howe AS, et al. Pertussis systematic review of systematic reviews of healthcare interventions. BMC Med Immunisation in Pregnancy Safety (PIPS) Study: a retrospective cohort study Res Method 2011;11(1):15. of safety outcomes in pregnant women vaccinated with the Tdap vaccine. [10] Higgins J, Green S, editors. Cochrane handbook for systematic reviews of Vaccine 2018;36(34):5173–9. interventions. Version 5.1.0 ed: The Cochrane Collaboration; 2011. B26 D. Macias Saint-Gerons et al. / Vaccine 39 (2021) B12–B26 [37] DeSilva M, Vazquez-Benitez G, Nordin JD, Lipkind HS, Klein NP, Cheetham TC, pregnancy: Final report of a World Health Organization working group. et al. Maternal Tdap vaccination and risk of infant morbidity. Vaccine 2017;35 Vaccine 2017;35(43):5738–50. (29):3655–60. [45] Haneuse S, VanderWeele TJ, Arterburn D. Using the E-value to assess the [38] GBD 2015 Child Mortality Collaborators. Global, regional, national, and potential effect of unmeasured confounding in observational studies. JAMA selected subnational levels of stillbirths, neonatal, infant, and under-5 2019(Jan). https://doi.org/10.1001/jama.2018.21554. mortality, 1980-2015: a systematic analysis for the Global Burden of Disease [46] Mertz D, Lo CK, Lytvyn L, Ortiz JR, Loeb M. Pregnancy as a risk factor for severe Study 2015. Lancet 2016;388(10053):1725–74. doi: 10.1016/S0140-6736(16) influenza infection: an individual participant data meta-analysis. BMC Infect 31575-6. Dis 2019;19(1):683. https://doi.org/10.1186/s12879-019-4318-3. [39] Stuurman AL, Riera M, Lamprianou S, Perez-Vilar S, Anderson SA, Mangtani P, [47] Kochhar S, Bonhoeffer J, Jones CE, Muñoz FM, Honrado A, Bauwens JS, et al. et al. Vaccine safety surveillance in pregnancy in low- and middle-income Immunization in pregnancy clinical research in low- and middle-income countries using GAIA case definitions: a feasibility assessment. Vaccine countries - study design, regulatory and safety considerations. Vaccine 2018;36(45):6736–43. 2017;35(48 Pt A):6575–81. [40] Hozbor D, Ulloa-Gutierrez R, Marino C, Wirsing von König CH, Tan T, Forsyth K. [48] Seale AC, Blencowe H, Manu AA, Nair H, Bahl R, Qazi SA, et al. Estimates of Pertussis in Latin America: Recent epidemiological data presented at the 2017 possible severe bacterial infection in neonates in sub-Saharan Africa, south Global Pertussis Initiative meeting. Vaccine 2019. pii: S0264-410X(19)30885- Asia, and Latin America for 2012: a systematic review and meta-analysis. 0. Lancet Infect Dis 2014;14(8):731–41. [41] Walsh LK, Donelle J, Dodds L, Hawken S, Wilson K, Benchimol EI, et al. Health [49] World Health Organization, 2006. Neonatal and perinatal mortality: country, outcomes of young children born to mothers who received 2009 pandemic regional and global estimates. World Health Organization, Geneva. Available H1N1 influenza vaccination during pregnancy: retrospective cohort study. BMJ from: http://www.who.int/iris/handle/10665/43444. 2019;366:l4151. [50] Omer SB, Clark DR, Aqil AR, Tapia MD, Nunes MC, Kozuki N, et al. Maternal [42] World Healh Organization. Low Birth Weight Policy Brief. Global Nutrition influenza immunization and prevention of severe clinical pneumonia in young Targets 2025. WHO/NMH/NHD/14.5. Available from: https://www.who.int/ infants: analysis of randomized controlled trials conducted in Nepal, Mali and nutrition/publications/globaltargets2025_policybrief_lbw/en/. South Africa. Pediatr Infect Dis J 2018;37(5):436–40. [43] Katz J, Lee AC, Kozuki N, Lawn JE, Cousens S, Blencowe H, et al. Mortality risk in [51] The electronic medicines compendium. Summary of Product Characteristics preterm and small-for-gestational-age infants in low-income and middle- (SmPC) of quadrivalent Influenza vaccine (Split virion, inactivated). Available income countries: a pooled country analysis. Lancet 2013;382(9890):417–25. on line: https://www.medicines.org.uk/emc/product/666/smpc [Last accessed [44] Fell DB, Azziz-Baumgartner E, Baker MG, Batra M, Beauté J, Beutels P, et al. Sept 18, 2019]. WHO taskforce to evaluate influenza data to inform vaccine impact and [52] Cipriani A, Ioannidis JPA, Rothwell PM, Glasziou P, Li T, Hernandez AF, et al. economic modelling. Influenza epidemiology and immunization during Generating comparative evidence on new drugs and devices after approval. Lancet 2020;395(10228):998–1010.