Guidance for the collection of case report form variables to assess safety in clinical trials of vaccines in pregnancy☆

Vaccination in pregnancy is an effective strategy to prevent serious infections in mothers and their infants. Safety of this strategy is of principal importance to all stakeholders. As the number of studies assessing safety of vaccines in pregnancy increases, the need to ensure consistent collection and reporting of critical data to allow comparisons and data pooling becomes more important. The Global Alignment of Immunization Safety Assessment in Pregnancy (GAIA) project aims to improve data collection and create a shared understanding of maternal, fetal and neonatal outcomes in order to progress the global agenda for vaccination in pregnancy. The guidance in this document has been developed to harmonize the data collected in case report forms used for safety monitoring in clinical trials of vaccination in pregnant women. Data to be collected is prioritized to allow applicability in diverse research settings, including low and middle-income countries. Standardized data will enable the research community to have a common base upon which to conduct meta-analyses, strengthening the applicability of outcomes to different settings.


a b s t r a c t
Vaccination in pregnancy is an effective strategy to prevent serious infections in mothers and their infants. Safety of this strategy is of principal importance to all stakeholders. As the number of studies assessing safety of vaccines in pregnancy increases, the need to ensure consistent collection and reporting of critical data to allow comparisons and data pooling becomes more important. The Global Alignment of Immunization Safety Assessment in Pregnancy (GAIA) project aims to improve data collection and create a shared understanding of maternal, fetal and neonatal outcomes in order to progress the global agenda for vaccination in pregnancy.
The guidance in this document has been developed to harmonize the data collected in case report forms used for safety monitoring in clinical trials of vaccination in pregnant women. Data to be collected is prioritized to allow applicability in diverse research settings, including low and middle-income countries. Standardized data will enable the research community to have a common base upon which to conduct meta-analyses, strengthening the applicability of outcomes to different settings.
Ó 2016 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Background and need for this guidance
Vaccination in pregnancy is an effective strategy to prevent serious infections in mothers and their infants [1][2][3][4][5]. Recommendations exist for use of tetanus and influenza vaccines in many countries and the number of countries recommending pertussis vaccination continues to increase. Other vaccines are recommended in pregnant women where there is perceived benefit, such as hepatitis A, hepatitis B and meningococcal (serogroups A,C,W,Y). Novel vaccines targeting group B streptococcus (GBS) and respira-tory syncytial virus (RSV) are in various stages of development [6,7].
Safety of vaccination in pregnancy is a key consideration for pregnant women, healthcare providers, vaccine manufacturers, regulators, sponsors and ethics committees. The number of studies assessing safety of vaccination in pregnancy continues to increase; however, inter-study variability makes comparisons and pooling of data challenging [8]. The failure to collect and consistently report critical data and the absence of guidance for data collection were identified at two international conferences, which concluded that data collection and presentation should be harmonized across different studies and settings [9,10].
The Global Alignment of Immunization Safety Assessment in Pregnancy (GAIA) project (http://gaia-consortium.net), coordinated by the Brighton Collaboration Foundation (https://brightoncollaboration.org), aims to improve data collection and create a shared understanding of maternal, fetal and neonatal outcomes in order to progress the global agenda for vaccination in pregnancy. The guidance proposed in this document have therefore been developed to harmonize the data collected in case report forms (CRFs) used for safety monitoring in clinical trials of vaccination in pregnant women. Guidance on the prioritization of the data to be collected is also provided to promote collection of at least a minimal set of high priority parameters in various settings, including low and middle-income countries (LMIC).

Use of this guidance
The aim of this guidance is to provide a standard for the collection of data in CRFs in clinical vaccine trials involving pregnant women where safety is an outcome. The guidance is presented as a series of tables and is referred to henceforth as a data collection matrix. It is intended as a tool to optimize data collection and to do so in a standardized manner in order to improve accuracy and comparability between clinical trials of vaccines in pregnancy. A standardized set of data will enable the research community to have a common base upon which to conduct meta-analyses, strengthening the applicability of outcomes to different settings. This data collection matrix is intended to be useful in all phases of clinical trials, from Phase I to Phase IV, from initial planning to implementation and evaluation. It is aimed at all stakeholders, from investigators, research networks, ethics committees and sponsors. It is also intended to be applicable in all resource settings; however, it is acknowledged that there are particular challenges to implementation in low and middle income countries (LMIC).
In consideration of this wide remit, data variables are prioritized into Priority 1 and Priority 2 data as follows: Priority 1: Essential: data considered essential for the understanding of the trial results and/or required by national and/or international regulatory authorities Priority 2: Complementary: data considered complementary, but not essential.
This data collection matrix is intended as a tool to assist all stakeholders; it is not regulatory or mandatory in nature. It is not intended to guide or establish criteria for clinical management. It is also not all-encompassing; it should be considered as a minimal data set in clinical trials where safety is an outcome. It is also expected that it may be adapted according to the specific aims and objectives of the individual clinical trial and that additional variables and data may also be collected.
It is intended that this guidance will be used alongside other existing guidelines from the Brighton Collaboration and the GAIA project. In particular, it is complementary to the ''Guideline for collection, analysis and presentation of safety data in clinical trials of vaccines in pregnant women" [14]. The existing case definitions of key neonatal and maternal outcomes for clinical trials of vaccines in pregnancy, produced as part of GAIA should be referred to [14] as relevant and as new case definitions are developed these should also be used for safety assessments in future clinical trials. They will become available at www.brightoncollaboration.org.

Development process of data collection matrix
This data collection matrix was constructed using an iterative process. Six CRFs from investigator initiated and industry sponsored clinical trials carried out in diverse geographical settings, including Africa, Asia, Europe and North America, and assessing different vaccines were collected and all variables were extracted into Microsoft Excel (see acknowledgments section for contributors of the CRFs used). Each variable was then coded according to whether or not it was collected in each study. This enabled a visual representation of the variables in each study to be displayed. Each member of the Data Collection Matrix Working Group (CEJ, MS, PTH, SB, UH), then independently assessed each variable as essential, important or non-essential in clinical trials assessing safety of vaccination in pregnancy. Any other variables considered essential but missing from this master list were added at this stage. Each variable was then scored according to the number of individuals who considered it as essential or important. The list of variables for inclusion or exclusion was then reviewed and agreed by all members of the working group during a series of telephone conferences. Included variables were further refined during a process of review by the Executive Committee of the GAIA consortium by telephone calls and face to face meetings. Variables were grouped into tables and harmonized with the Guideline for collection, analysis and presentation of safety data in clinical trials of vaccines in pregnant women (referred to forthwith as the Guidelines document) [14]. The data collection matrix was then refined following structured peer-review by the broad global Brighton Collaboration Reference Group and review by subject matter experts attending the Harmonized Safety Monitoring of Immunization in Pregnancy International Consensus Conference and Investigators Workshop, 28-30th March 2016, National Institutes of Health, Bethesda, USA [11,12]. This guidance should be considered as a 'living document', which will be reviewed periodically and updated to take account of emerging data and feedback from investigators implementing this guidance, these will be available at www.brightoncollaboration.org.

Rationale for overall structure of the data collection matrix
The data collection matrix is presented as a series of tables of variables to be collected in case report forms. Each table relates to a different time-point or section of the case report forms. The data collection matrix and the Guidelines document are discrete documents, which are highly inter-related. It is expected that both documents will be used in parallel. The Guidelines document provides higher level information whereas the data collection matrix provides greater granularity. For example, the Guidelines document advises that safety follow-up should include a symptom diary to record solicited and unsolicited local and systemic adverse events following immunization (AEFI). The data collection matrix provides the detail of what variables should be collected in this symptom diary and suggests how signs and symptoms should be measured or graded. Each table in the data collection matrix relates to a section in the Guidelines document. Therefore, whilst distinct documents, they are harmonized and should be used together.

Guidance for the collection of case report form variables
The tables define data that should be collected prior to vaccination (Table 1), at the time of vaccination (Table 3) and in the follow-up period (Table 4). Early phase clinical trials may select pregnant women at low risk of obstetric complications, whereas in other studies, particularly late phase clinical trials, it may be preferable to enroll pregnant women regardless of their obstetric risk. In recognition of this, an example of variables that may be collected to assess obstetric risk is provided ( Table 2).
In order to assess safety of vaccination in pregnancy, it is recommended that the minimum follow-up period for women is 6 months post-delivery or the early termination of pregnancy; the minimum recommended follow-up period for infants is until 1 year of age (Table 4). It is acknowledged that there are significant logistical challenges with extended follow-up periods; where a shorter follow-up period is pre-defined in the study protocol, adequate justification should be given, for example, based on biological characteristics of the vaccine, the vaccine-targeted disease or of the adverse event of specific interest, including patterns identified f Guidance on the assessment of gestational age can be found in the GAIA case definition of preterm birth at brightoncollaboration.org [14].

Table 2
Example table of data that may be collected to inform obstetric risk assessment. org [14]. b Case definition of stillbirth is available at brightoncollaboration.org [14]. c Case definition of neonatal death is available at brightoncollaboration.org [14]. d Case definition of congenital anomaly is available at brightoncollaboration.org [14].

Table 3
Vaccine and immunization related data.

Priority
Investigational vaccine and immunization procedures Participant baseline vital signs (including blood pressure, heart rate, respiratory rate, and temperature) Need for analgesia or antipyretic 2 Table 4 Follow-up monitoring data (including data pertaining to the mother pre-delivery, fetus, the delivery and the neonate, mother and infant until completion of follow-up). 1 Infant ethnicity/race 1 Infant APGAR score (or other immediate assessment for signs of life) at 1, 5 and 10 min 1 Need for infant resuscitation with details 1 Specify any abnormalities on newborn examination 1 Infant feeding: breast milk (mother/donor), formula feeding, parenteral nutrition, mixed feeding, other (specify nutrient given), with start dates 2 Results of any neonatal laboratory tests specified in the protocol (minimum of full blood count, differential, transaminases, bilirubin, glucose, blood urea nitrogen and creatinine)   [14]. b Case definition of gestational hypertension is available at brightoncollaboration.org [14]. c Case definition of postpartum haemorrhage is available at brightoncollaboration.org [14]. d Case definition of stillbirth is available at brightoncollaboration.org [14]. e Case definition of neonatal death is available at brightoncollaboration.org [14]. f Guidance on the assessment of gestational age can be found in the GAIA case definition of preterm birth at brightoncollaboration.org [14]. g Case definition of congenital anomaly is available at brightoncollaboration.org [14]. h Case definition of neonatal infection is available at brightoncollaboration.org [14].
in previous trials. There may also be reasons for extending safety follow up further based on the above factors, or the characteristics of the vaccine recipient (e.g. nutrition, underlying diseases such as immune-depressing illnesses and other pre-existing conditions), or the intention to assess child development and late onset outcomes as part of the Risk Management Plan (which may require follow-up periods of 5 years or more). Table 5 provides guidance for the collection of high quality data in order to allow interpretation of adverse events following immunization (AEFI); the purpose is not to provide guidance on assess-ment of a causal relationship. More detailed guidance on protocol development for clinical trials assessing vaccines is available elsewhere [13] (see Table 6).

Conclusions
This guidance is intended as a tool to optimize data collection in all phases of clinical trials of vaccination in pregnancy where safety is an outcome. The aim of standardizing data collection is to improve accuracy and comparability and to allow pooling of data a Case definition of congenital anomaly is available at brightoncollaboration.org [14]. b Guidance of causality assessment can be found at: https://brightoncollaboration.org/public/resources/standards/guidelines.html. c Case definition of maternal death is available at brightoncollaboration.org [14]. d Case definition of neonatal death is available at brightoncollaboration.org [14].
between studies. The remit of this guidance is wide and therefore data variables are prioritized into essential data (priority 1) and data that are useful but less essential (priority 2). Variables collected in CRFs are expected to be dependent on the pre-specified aims and objectives, study setting and resources in the clinical trial. It is acknowledged that there are particular challenges of collecting all data variables in LMICs. Evaluation of this tool will be the subject of future work in order to continue to improve the utility of the data collection matrix to all stakeholders in clinical trials of vaccines in pregnancy.

Disclaimer
The findings, opinions and assertions contained in this consensus document are those of the authors. They do not necessarily represent the official positions or views of their respective institution or organization (e.g., government, university, or corporation).