Secondary Logo

Journal Logo

Vaccine Reports

Applicability of the GAIA Maternal and Neonatal Outcome Case Definitions for the Evaluation of Adverse Events Following Vaccination in Pregnancy in High-income Countries

Watson, Gabriella MRCPCH*; Dodd, Caitlin PhD; Munoz, Flor M. MD, MSc; Eckert, Linda O. MD§; Jones, Christine E. MD, PhD¶,∥; Buttery, Jim P. MD**; Yildirim, Inci B. MD, PhD, MSc††,‡‡; Kachikis, Alisa MD, MSc§§; Heath, Paul T. FRCPCH¶¶; Schlaudecker, Elizabeth P. MD, MPH∥∥; Bond, Nanette H. PA-C***; Santarcangelo, Patricia L. RN***; Wilcox, Christopher R. MBBCh†††; Bellamy, Karen BSc‡‡‡; Elmontser, Mohnd MPH§§§; Sienas, Laura MD¶¶¶; Simon, Rebecca MD¶¶¶; Khalil, Asma MRCOG∥∥∥; Townsend, Rosemary MBChB****; Sturkenboom, Miriam PhD††††; Black, Steve MD‡‡‡‡

Author Information
The Pediatric Infectious Disease Journal: December 2021 - Volume 40 - Issue 12 - p 1127-1134
doi: 10.1097/INF.0000000000003261

Abstract

Despite significant reductions in under-five mortality, neonatal mortality rates have not decreased at the same rate, thus reducing neonatal mortality is an important target of the 2030 Sustainable Development Goals.1,2 Vaccination in pregnancy is a strategy that has been shown to reduce infection in both pregnant women and neonates and is seen as a priority by the World Health Organization (WHO) to reduce the global burden of infection in these populations.3 Maternal immunization programs for tetanus are well established and have proven successful in reducing the burden of disease in mothers and neonates, with the benefits of influenza and pertussis immunization programs being demonstrated more recently.4–6 Vaccines in development, such as respiratory syncytial virus and group B Streptococcus, have specific indications for use in pregnancy and show promise for reducing the burden of these infections.7,8 Among pregnant women with SARS-CoV-2, there is an associated risk of hospitalization, intensive care admissions, preterm delivery and maternal death.9 The potential risk posed by SARS-CoV-2 to pregnant women indicates that clinical trials and observational studies of SARS-CoV-2 vaccines in pregnant women will be necessary to demonstrate safety and efficacy in this population.

Standardized case definitions to evaluate adverse events following immunizations (AEFIs) during pregnancy are essential for a globally harmonized approach to the monitoring of vaccine safety, both for vaccines progressing through clinical trials and those implemented in routine care.10,11 During the current COVID-19 pandemic, there is a pressing need for rapid up-scaled implementation of SARS-CoV-2 vaccine trials and safety assessment studies in pregnant women. This will require global collaboration, data pooling and sharing, as well as high-quality comparable data on AEFIs to ensure protection of pregnant women and their infants in a compressed time frame.

Standardized case definitions allow for comparability of data across studies and countries. Improved recording and detection of AEFI during pregnancy globally will increase vaccine confidence. The Global Alignment of Immunization safety Assessment in pregnancy (GAIA) project developed 25 case definitions for assessing AEFIs during pregnancy using the Brighton Collaboration template and levels of diagnostic certainty.10,12–22 The GAIA project was designed in response to the WHO call for global efforts to monitor the safety of vaccines in pregnancy for use in both high- and low-resource settings, in line with the WHO Global Vaccine Safety Blueprint 1.0.23

It is important to evaluate GAIA definitions in practice to test their applicability and feasibility in different contexts and understand their limitations. The GAIA definitions were designed for use in clinical trials of vaccines in pregnancy; however, investigators may apply these definitions in retrospective or observational studies. Therefore, it is important to determine their utility in these settings as well.

The objective of this study was to evaluate the applicability of ten GAIA case definitions and one enabling term (a term upon which other case definitions rely) retrospectively to data collected in routine clinical care or in research trials across 7 sites in high-resource settings: the United States, the United Kingdom and Australia.

METHODS

Study Setting

Study sites in 3 different countries in high-resource settings, 4 in the United States, 2 in the United Kingdom and 1 in Australia, were included in the study.

Case Definitions Evaluated

The GAIA case definitions evaluated comprised 5 neonatal outcomes: preterm birth, low birth weight, small for gestational age, respiratory distress and microcephaly and 5 maternal outcomes: preterm labor, fetal growth restriction, preeclampsia, nonreassuring fetal status and dysfunctional labor. Gestational age is required for most case definitions, as such this enabling term was assessed in all pertinent cases. ICD-9 and ICD-10-CM codes were created using the Codemapper tool,24 and reviewed by medical experts (S.B. and F.M.M.).

Data Collection

Clinical cases were identified through individual hospital coding departments according to ICD-9 and 10 codes. Research cases were identified differently depending on the study site, either using MedDRA codes or by hand searching research records from relevant clinical trials.

Data collection forms were developed for each outcome and used uniformly across all study sites (see Appendix, Supplemental Digital Content 11, https://links.lww.com/INF/E482, for data collection forms and guidance for use). All investigators were from clinical backgrounds, either pediatric or obstetric, and underwent training to abstract data from test cases before using the data collection forms to abstract data from clinical or research records. Inter and intrarater comparisons were made by conducting a review of the responses from an exercise of adjudication or mock cases and scenarios.

Abstracted data was recorded on paper data collection forms. Fully anonymized data were then entered into password-protected REDcap database developed for this project.25,26 A Brighton Collaboration level of diagnostic certainty was assigned to each abstracted case according to the GAIA case definitions; the site principal investigator verified the level of diagnostic certainty assigned to each case. Where initial data abstraction was performed by the principal investigator, another investigator verified the level of diagnostic certainty. Where there were discordances, a third investigator reviewed the case.

Brighton Collaboration levels of diagnostic certainty were determined in two ways: at abstraction by the investigator, and at the analysis stage by applying an automated series of decision rules based upon the Brighton definitions and associated case logic. This was programmed using SAS (Version 9.4; SAS Institute Inc, Cary, NC; 2014), which was based on the rules from the Automated Brighton Case (ABC tool) classification (see Appendix, Supplemental Digital Content 11, https://links.lww.com/INF/E482, for logic of ABC tool). The ABC tool was in development at the time, and not fully functional; therefore, the modified tool programmed in SAS is described as “ABC case logic” here.

Box 1. - The Brighton Collaboration Levels of Diagnostic Certainty
Level 1 Definite case
Level 2 Probable case based on resources
Level 3 Possible case based on resources
Level 4 Insufficient evidence to confirm
Level 5 Not a case of the outcome event

Analysis

Based on the abstracted data in REDCap database, the following parameters were assessed across all sites for each of the case definitions, (analysis by site was also performed and is described in Supplemental Digital Content 1–10, https://links.lww.com/INF/E472; https://links.lww.com/INF/E473; https://links.lww.com/INF/E474; https://links.lww.com/INF/E475; https://links.lww.com/INF/E476; https://links.lww.com/INF/E477; https://links.lww.com/INF/E478; https://links.lww.com/INF/E479; https://links.lww.com/INF/E480; https://links.lww.com/INF/E481):

  1. Ability to assign a level of diagnostic certainty up to level 3—both by abstractor and using ABC logic. Where the level of certainty was not assessable, the missing components required to achieve a level 1 diagnostic certainty were identified.
  2. Quality of case description available in records, what level of diagnostic certainty was ascertainable for each outcome—both by abstractor and using ABC logic.
  3. Performance of the case definitions—the positive predictive value (PPV) for ICD-10 codes (for clinical cases only, as research cases were not selected based on ICD-10 codes).

Ethics

The study protocol was reviewed and approved by the institutional review board at Cincinnati Children’s Hospital Medical Center (Ref: 00002988) and at Baylor College of Medicine and Affiliated Hospitals (H-42922). Monash Health Human Research Ethics Committee Low Risk Panel approved (NMA HREC Reference Number: LNR/18/MonH/405 and Monash Health Ref: RES-18-0000-280L).

RESULTS

A total of 1246 cases were identified across the 7 study sites: 624 neonatal and 622 maternal. Of the neonatal records, 578 were from clinical case records and 46 from research case records. Of the maternal records, 583 were from clinical case records and 39 from research case records. Clinical case records were available at all participating sites, but research records were only available at 3 sites, all from interventional or observational studies of vaccines in pregnancy.

Gestational Age

Gestational age was nonassessable by the abstractor in 18.3% (114/624) of neonatal cases reviewed, and 2.1% (13/622) of maternal cases. In neonatal cases, of those assessable, 48% (298/624) had a level 1 level of diagnostic certainty, which was highest in sites where obstetric and neonatal records were linked. At sites where there was no maternal information in the infant records, in particular research records, assessability was low. From maternal cases, gestational age was assessable for 78% (484/622) at level 1 level of diagnostic certainty, which was consistent across all sites. Figure 1 illustrates the level of diagnostic certainty for gestational age, by study site and by abstractor or ABC case logic. Inability to assess gestational age was most often due to lack of information on “certain” last menstrual period and missing information on first trimester ultrasound.

F1
FIGURE 1.:
Stacked bar graphs to illustrate level of diagnostic certainty by gestational age by abstractor and ABC case logic across all study sites. AU_1, Australian study site 1; LOC, level of diagnostic certainty; UK_1, UK study site 1; UK_2, UK study site 2; US_1, USA study site 1; US_2, USA study site 2; US_3, USA study site 3.

Neonatal Case Definitions

Preterm Birth

Preterm birth was nonassessable by the abstractor in 17.6% (25/142) of cases, and 27.5% (39/142) by the ABC case logic. Where it was assessable, the majority of cases were level 1 level of diagnostic certainty (Fig. 2). Where there was difficulty in assigning level 1, this was due to missing data on last menstrual periods and first trimester ultrasound. The PPV was very high in most sites (Table 1), except 2 sites where neonatal records often did not include information on timing of first ultrasound or “certain” last menstrual period.

TABLE 1. - Summary of Results; Total Numbers of Medical and Research Records Reviewed, Percentage of Outcomes Nonassessable, Mean PPVs for Neonatal and Maternal Outcomes for ICD-10 Codes and Ranges Across All Sites
Outcomes Medical and Research Records Records Reviewed Nonassessable by Abstractor, n (%) Nonassessable by ABC case logic, n (%) Mean % PPV for ICD-10 Code Across All Sites, From Medical Records (95% CI) Range in % PPV Estimates Across Sites
Gestational age Medical Records 1161 113 (9.7) 243 (20.9)
Research records 85 14 (16.5) 17 (20.0)
Total 1246 127 (10.2) 260 (20.9)
Neonatal outcomes
Preterm birth Medical records 123 25 (20.3) 38 (30.9) P07: PPV 76.4 (67.6–85.2) 52.9%–100%
Research records 19 0 (0) 1 (5.3)
Total 142 25 (17.6) 39 (27.5)
Low birth weight Medical records 110 31 (28.2) 36 (32.7) P07: PPV 88.0 (80.6–95.4) 0%–100%
Research records 13 1 (7.7) 1 (7.7)
Total 123 32 (26.0) 37 (30.1)
Small for gestational age Medical Records 118 32 (27.1) 65 (55.1) P05: PPV 70.4 (62.1–78.8) 20%–100%
Research records 7 2 (28.6) 7 (100)
Total 125 34 (27.2) 72 (57.6)
Respiratory Distress Medical Records 119 4 (3.4) 40 (33.6) P22: PPV 76.7 (67.8–85.7) 30%–100%
Research records 7 4. (57.1) 5 (71.4)
Total 126 8 (6.3) 45 (35.7)
Microcephaly Medical Records 108 33 (30.6) 72 (66.7) Q02: PPV 40.0 (29.0–50.0) 0%–100%
Research records
Total 108 33 (30.6) 72 (66.7)
Maternal outcomes
Preterm Labor Medical records 121 60 (49.6) 58 (47.9) O60. PPV: 56.8 (47.7–65.8) 0%–100%
Research records 5 5 (100) 3 (60.0)
Total 126 65 (51.6) 61 (48.4)
Fetal growth restriction Medical Records 120 32 (26.7) 43 (5.0) O36. PPV: 80.9 (72.9–88.8) 55%–100%
Research records 12 1 (8.3) 1 (8.3)
Total 132 33 (30.5) 44 (33.3)
Preeclampsia Medical Records 111 28 (25.2) 42 (37.8) O14: PPV 81.2 (72.9–89.5) 33%–100%
Research records 14 5 (35.7) 9 (64.3)
Total 125 33 (26.4) 51 (40.8)
Nonreassuring fetal status Medical records 108 75 (64.4) 74 (68.5) O68: PPV 38.3 (24.4–52.2) O76: PPV 15.0 (0–30.7) O77: PPV 57.9 (35.7–80.1) 0%–64%
Research records 5 4 (80.0) 4 (80.0)
Total 113 79 (69.9) 78 (69.0)
Dysfunctional labor Medical records 123 50 (40.7) 123 (100) O62: PPV 34.0 (21.1–46.7) O66: PPV 55.0 (33.2–76.8) 11.1%–50%
Research records 3 2 (66.7) 3 (100)
Total 126 52 (41.2) 126 (100)

F2
FIGURE 2.:
Stacked bar graphs to illustrate level of diagnostic certainty for neonatal outcomes by abstractor and ABC case logic across all study sites. AU_1, Australian study site 1; LOC, level of diagnostic certainty; UK_1, UK study site 1; UK_2, UK study site 2; US_1, USA study site 1; US_2, USA study site 2; US_3, USA study site 3.

Low Birth Weight

Low birth weight was nonassessable by the abstractor in 26.0% (32/123) of cases and 30.1% (37/123) by the ABC case logic, and the most frequent reason for this was missing information on calibration of scales. Three sites classified all cases as level 1 level of diagnostic certainty (Fig. 2). The mean PPV across all sites was high (Table 1).

Small for Gestational Age

Small for gestational age was mostly assigned level 3B level of diagnostic certainty by abstractors (Fig. 2), except 3 sites where mothers’ last menstrual period was recorded in the neonatal records and level 1 diagnostic level of certainty was high. A total of 27.2% (34/125) were nonassessable by the abstractor and 57.6% (72/125) by the ABC case logic. The main difficulty for assessing small for gestational age related to the definition depending on a gestational age assessment and proper weight measurement, one hospital was not able to ascertain standard scale calibration required for the case definition. The mean PPV across all sites was good (Table 1).

Respiratory Distress

Almost all cases were assessable, with only 6.3% (8/126) of cases nonassessable by the abstractor, and 33.6% not assessable using the ABC case logic. The majority of cases were classified as level 1 level of diagnostic certainty (Fig. 2). The mean PPV was high (Table 1).

Microcephaly

Among cases of microcephaly, 30.6% (33/108) were nonassessable by abstractor and 66.7% (72/108) by the ABC case logic. Where cases were identified, it was based on postnatal diagnosis from neonatal records. The PPV was very low (Table 1), and varied across sites, dependent on whether maternal or neonatal records were used. The difficulties in classifying microcephaly were due to missing data, as it was often made as a post neonatal period diagnosis and making it impossible to look back to neonatal case records. The difficulties in assigning a level of diagnostic certainty were due to poor documentation of head circumference centile.

Maternal Case Definitions

Preterm Labor

There was significant variability across sites for preterm labor, with almost half nonassessable, 51.6% (65/126) were nonassessable by the abstractor and 48.4% (61/126) by the ABC case logic. One site was able to classify all cases, across other sites the majority were nonassessable due to missing recorded information on the number of contractions and change in the cervix (Fig. 3).

F3
FIGURE 3.:
Stacked bar graphs to illustrate level of diagnostic certainty for maternal outcomes by abstractor and ABC case logic across all study sites. AU_1, Australian study site 1; LOC, level of diagnostic certainty; UK_1, UK study site 1; UK_2, UK study site 2; US_1, USA study site 1; US_2, USA study site 2; US_3, USA study site 3.

Fetal Growth Restriction

Fetal growth restriction was mainly assessable at level 1 level of diagnostic certainty (Fig. 3). Only 30.5% (33/132) were nonassessable by the abstractor and 33.3% (44/132) by the ABC case logic. When cases were nonassessable, this was due to missing information on weight. The PPV was good (Table 1).

Preeclampsia

Only 26.4% (33/125) were nonassessable by the abstractor, and 37.8% were nonassessable by the ABC tool, with most cases from the United Kingdom classified as level 1 level of diagnostic certainty (Fig. 3). The PPV was high across all sites (Table 1).

Nonreassuring Fetal Status

The majority of these cases were nonassessable, 69.9% (79/113) by the abstractor and 69% (78/113) by the ABC case logic, and this was due to fetal heart rate not being captured on the data collection forms, as likely not recorded in the medical or research records. As a result, the PPV was very low (Table 1).

Dysfunctional Labor

On abstraction, 41.2% (52/126) cases were nonassessable, and on the ABC case logic, all cases were nonassessable as details on cervical dilation were missing in all cases (Fig. 3). The PPV was very low (Table 1).

DISCUSSION

Neonatal outcomes were most likely to be assessable and able to be assigned a level of diagnostic certainty. PPVs for preterm birth, low birth weight, small for gestational age and respiratory distress were all above 75%. Maternal outcomes for preeclampsia and fetal growth restriction also showed a high assessability with PPV over 80%. However, neonatal outcomes for microcephaly were often nonassessable, with a very low PPV. Maternal outcomes for preterm labor, nonreassuring fetal status and dysfunctional labor were also often nonassessable with poor PPV. The range of PPV was large for all definitions across sites and could not be extrapolated from one site to another, indicating the utility of some GAIA case definitions in this setting and the limitations of others, depending on the data recorded in clinical or research records. Missing data were one of the most important reasons a case could not be assigned a GAIA level of diagnostic certainty. A key observation derived from this study is that the quality of documentation in clinical and research records directly impacts the applicability of the GAIA case definitions, even in high-resource settings.

Correct identification of gestational age is fundamental for maternal immunization programs. Timing of immunization during pregnancy is an important factor in vaccine immunogenicity and a consideration in vaccine safety. Many other outcomes also rely on accurate gestational age identification. As such, accurate identification of this enabling factor is essential for maternal vaccine studies. Gestational age had excellent assessability from maternal records and was good from neonatal records. This was best where maternal and neonatal records were linked; however, assessability was poor where data were missing or incomplete and maternal and neonatal records were not linked. This highlights the importance of linking maternal and neonatal records within health facilities, and during clinical trials documenting key maternal information in neonatal records. Gestational age assessment should also include flexibility of options, for example, certain or uncertain last menstrual period with third trimester ultrasound.

Neonatal outcomes were most likely to be assessed and classified to a level of diagnostic certainty. Low birth weight and small for gestational age were not classified as frequently; this was due to weight not being recorded or information on the calibration of weighing machines not being specified or available. Some flexibility on requirements for machine calibration could be considered in the GAIA case definitions. Microcephaly was also less likely to be classified, with a very low PPV. The difficulties in classifying microcephaly were due to missing data, as it was often made as a post neonatal period diagnosis and making it impossible to look back to neonatal case records. The difficulties in assigning a level of diagnostic certainty were due to poor documentation of head circumference centile.

Maternal outcomes varied in their assessability. Preeclampsia and fetal growth restriction had good assessability; however, nonreassuring fetal status and dysfunctional labor had low assessability due to discrepancies in the data collection forms and what is documented in the medical or research records, making it difficult to classify the outcome by level of diagnostic certainty. Preterm labor had low assessability due to missing records. The data collection forms for nonreassuring fetal status and dysfunction labor should be reviewed and revised. Additionally, it could be relevant to assess whether abstractors who provide obstetric care were more frequently able to complete these forms for the obstetric definitions than abstractors who do not provide obstetrics care. Again, missing or incomplete data present a large problem.

The previous evaluation of GAIA case definitions in low-resource settings showed outcomes for preterm birth and hypertension were sensitive in both retrospective and prospective studies and reliable and feasible to use; however, the stillbirth definition was not as sensitive and would need further modification of gestational age assessment parameters to be useful in the setting.27 A recent study in the United States demonstrated successful application of GAIA case definition in retrospectively collected electronic medical records for pregnancy outcomes (Moll et al, abstract and presentation at ICPE conference, 2020).

Case identification in retrospective studies is usually based on ICD codes; however, coding alone does not allow for verification of cases, and codes do not allow for the case classification into level of diagnostic certainty. Research documentation has changed over time with different documentation requirements, and MedDRA codes are not used globally and were not necessarily reported appropriately. Therefore, a retrospective review presented challenges to determine applicability of the GAIA case definitions, or to validate them against MedDRA codes. It is important to understand the GAIA case definitions were primarily designed for prospective research data collection, rather than retrospective. Applicability of GAIA case definitions to retrospectively classify outcomes varied across sites and countries. For use in retrospective studies, the GAIA case definitions would need to be reviewed and adapted. Study findings highlight the priority for adapting and revising some of these definitions.

The investigators noted that review of both clinical and research records was labor-intensive, with between 1 and 2 hours spent on each record. To use the GAIA case definitions properly, investigators need to fully understand the background and rationale for each case definition, be familiar with the terminology and documentation in clinical or research situations, and utilize data collection forms and guidance documents specifically developed for the project.

Investigator bias in assigning the level of diagnostic certainty and classification needs to be considered; it is expertise-dependent, and there will likely be intersite and interuser variability. This topic will be addressed further in a complimentary paper on abstractors’ variability.

Strengths and Limitations

This is the first study to evaluate GAIA case definitions in high-income settings. A large number of cases were evaluated across multiple different sites and countries ensuring their usability in different contexts and settings. The data collection forms developed can be used as a blueprint for application of the GAIA case definitions globally.

Individual abstractor expertise could have influenced decisions on the level of diagnostic certainty, with the potential for bias with interuser and intersite variability. In some cases, the design of the data collection forms presented issues with appropriate data collection to complete the level of diagnostic certainty. Data collection was often limited by incomplete or missing notes. We were able to assess few research records due to the relatively low numbers of women enrolled in intervention studies in pregnancy compared with those receiving clinical care and the low frequency of adverse events in this selected population of women, who were often at low-risk of complications.

-
Lessons Learned
• A priori development of source documents and data collection forms based on the GAIA case definitions is necessary to ensure that all elements of the definition are included • Training of personnel responsible for data abstraction/extraction is crucial to ensure consistency and comparability of data collection
• Case review can be labor-intensive
• There is difficulty in retrospectivity ascertaining cases from clinical records • Missing data and lack of clarity on data documentation led to difficulties assigning level of diagnostic certainty
• Importance of full understanding of background, methodology and rationale for all GAIA case definitions before assigning levels of diagnostic certainty
• GAIA case definitions were designed to be applied prospectively and this must be taken into consideration when applying retrospectively
• GAIA case definitions could be applied in retrospective case ascertainment by adapting them by using alternative sources of data, linkage of various data sources, and allowing flexibility in the ascertainment of the elements and levels of certainty of the case definition. • International variations in case documentation and practice make standardization challenging
• Some GAIA case definitions with low PPV might need to be updated to ensure relevance in clinical observational studies

CONCLUSION

The applicability of the GAIA case definitions to retrospectively identify and classify maternal and neonatal outcomes reported in either clinical or research records was variable in sites in high-resource settings. Even though the case definitions include various levels of diagnostic certainty to be applicable to various resource settings based on diagnostic capabilities, the implementation of the case definitions is largely dependent on the type and quality of documentation in clinical and research records in both high- and low-resource settings. Furthermore, while originally designed for use in the prospective evaluation of maternal vaccine safety, the GAIA case definitions would likely need to be specifically adapted for observational studies by using alternative sources of data, linking various data sources, and allowing flexibility in the ascertainment of the elements and levels of certainty of the case definition.

ACKNOWLEDGMENTS

We extend our thanks to the following people who assisted with data abstraction at study sites: Suzan Walker, University of Washington, United States; Felicia Scaggs Huang and Hilary Miller-Handley, Cincinnati Children’s Hospital Medical Center, United States; Saad Omer, Yale Institute for Global Health, United States; Yonatan Mesfin, Monash Centre for Health Research and Implementation, Monash University, Australia; Uzma Khan, Fran Mabesa, Yaa Acheampong, St George’s University Hospital NHS Trust, London, United Kingdom.

REFERENCES

1. Lawn JE, Blencowe H, Oza S, et al.; Lancet Every Newborn Study Group. Every Newborn: progress, priorities, and potential beyond survival. Lancet. 2014;384:189–205.
2. Transforming our world, the 2030 agenda for sustainable development. A/RES/70/1. United Nations; 2015. Available at: https://sdgs.un.org/publications/transforming-our-world-2030-agenda-sustainable-development-17981. Accessed December 12, 2020.
3. Zaman K, Roy E, Arifeen SE, et al. Effectiveness of maternal influenza immunization in mothers and infants. N Engl J Med. 2008;359:1555–1564.
4. Amirthalingam G, Andrews N, Campbell H, et al. Effectiveness of maternal pertussis vaccination in England: an observational study. Lancet. 2014;384:1521–1528.
5. Madhi SA, Cutland CL, Kuwanda L, et al.; Maternal Flu Trial (Matflu) Team. Influenza vaccination of pregnant women and protection of their infants. N Engl J Med. 2014;371:918–931.
6. World Health Organization. Progress towards maternal and neonatal tetanus elimination worldwide, 2000-2018. Wkly Epidemiological Rec. 2020;95:173–84.
7. Madhi SA, Polack FP, Piedra PA, et al.; Prepare Study Group. Respiratory syncytial virus vaccination during pregnancy and effects in infants. N Engl J Med. 2020;383:426–439.
8. Dzanibe S, Madhi SA. Systematic review of the clinical development of group B streptococcus serotype-specific capsular polysaccharide-based vaccines. Expert Rev Vaccines. 2018;17:635–651.
9. Ellington S, Strid P, Tong VT, et al. Characteristics of women of reproductive age with laboratory-confirmed SARS-CoV-2 infection by pregnancy status - United States, January 22-June 7, 2020. MMWR Morb Mortal Wkly Rep. 2020;69:769–775.
10. Bonhoeffer J, Kochhar S, Hirschfeld S, et al.; GAIA project participants. Global alignment of immunization safety assessment in pregnancy - The GAIA project. Vaccine. 2016;34:5993–5997.
11. Kochhar S, Bauwens J, Bonhoeffer J; GAIA Project Participants. Safety assessment of immunization in pregnancy. Vaccine. 2017;35(48, pt A):6469–6471.
12. Brighton Collaboration. Available at: https://www.brightoncollaboration.org/. Accessed December 15, 2020.
13. Harrison MS, Eckert LO, Cutland C, et al.; Brighton Collaboration Pathways to Preterm Birth Working Group. Pathways to preterm birth: case definition and guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2016;34:6093–6101.
14. Cutland CL, Lackritz EM, Mallett-Moore T, et al.; Brighton Collaboration Low Birth Weight Working Group. Low birth weight: case definition & guidelines for data collection, analysis, and presentation of maternal immunization safety data. Vaccine. 2017;35(48, pt A):6492–6500.
15. Schlaudecker EP, Munoz FM, Bardají A, et al.; Brighton Collaboration Small for Gestational Age Working Group. Small for gestational age: case definition & guidelines for data collection, analysis, and presentation of maternal immunisation safety data. Vaccine. 2017;35(48, pt A):6518–6528.
16. Sweet LR, Keech C, Klein NP, et al.; Brighton Collaboration Respiratory Distress in the Neonate Working Group. Respiratory distress in the neonate: case definition & guidelines for data collection, analysis, and presentation of maternal immunization safety data. Vaccine. 2017;35(48, pt A):6506–6517.
17. DeSilva M, Munoz FM, Sell E, et al.; Brighton Collaboration Congenital Microcephaly Working Group. Congenital microcephaly: case definition & guidelines for data collection, analysis, and presentation of safety data after maternal immunisation. Vaccine. 2017;35(48, pt A):6472–6482.
18. Quinn JA, Munoz FM, Gonik B, et al.; Brighton Collaboration Preterm Birth Working Group. Preterm birth: case definition & guidelines for data collection, analysis, and presentation of immunisation safety data. Vaccine. 2016;34:6047–6056.
19. Easter SR, Eckert LO, Boghossian N, et al.; Brighton Collaboration Fetal Growth Restriction Working Group. Fetal growth restriction: case definition & guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2017;35(48, pt A):6546–6554.
20. Rouse CE, Eckert LO, Wylie BJ, et al.; Brighton Collaboration Preeclampsia Working Group. Hypertensive disorders of pregnancy: case definitions & guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2016;34:6069–6076.
21. Gravett C, Eckert LO, Gravett MG, et al.; Brighton Collaboration Non-reassuring fetal status Working Group. Non-reassuring fetal status: case definition & guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2016;34:6084–6092.
22. Boatin AA, Eckert LO, Boulvain M, et al.; Brighton Collaboration Dysfunctional Labor Working Group. Dysfunctional labor: case definition & guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2017;35(48, pt A):6538–6545.
23. World Health Organization. Global vaccine safety blueprint. World Health Organization; 2012. Available at: https://apps.who.int/iris/handle/10665/70919. Accessed December 15, 2020.
24. Becker BFH, Avillach P, Romio S, et al.; ADVANCE consortium. CodeMapper: semiautomatic coding of case definitions. A contribution from the ADVANCE project. Pharmacoepidemiol Drug Saf. 2017;26:998–1005.
25. Harris PA, Taylor R, Minor BL, et al.; REDCap Consortium. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208.
26. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–381.
27. Kochhar S, Clarke E, Izu A, et al. Immunization in pregnancy safety surveillance in low and middle-income countries- field performance and validation of novel case definitions. Vaccine. 2019;37:2967–2974.
Keywords:

adverse event; immunization; vaccines; pregnancy; safety

Supplemental Digital Content

Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.