Abstract
Background
COVID-19 vaccines are protective against disease. Pregnant women benefit from vaccination as they are at higher risk of poor maternal and neonatal outcomes following infection.
Methods
Following regulatory approval of two COVID-19 vaccines in the United Kingdom, a rapid national study of vaccination in pregnancy was instituted using three existing safety surveillance platforms: UKOSS, UKTIS and VIP. This preliminary report describes the data collected up to the 15th June 2021.
Results
There were 971 reports of COVID-19 vaccination in the UKOSS/UKTIS (n = 493) and VIP (n = 478) monitoring systems describing 908 individual pregnancies. Pfizer-BioNTech mRNA vaccination was most common (n = 501, 55.2%), most women were vaccinated in their second or third trimester (n = 566, 62.3%), and were mainly vaccinated due to occupational infection risk (n = 577, 63.5%).
Conclusion
Obstetric outcome data will be obtained by December 2021. However, women should not delay vaccination whilst awaiting further safety data to emerge.
Keywords
Background
Pregnant women, particularly those in their third trimester, are known to be at risk of severe adverse outcomes following SARS-CoV-2 infection.1,2 Around 1% of pregnant women admitted to hospital with symptomatic SARS-CoV-2 infection (COVID-19) in the UK will die, 10% require critical care, and 20% have a preterm birth. 2 Women from Black, Asian and other ethnic minority backgrounds, older women, women who are overweight or obese, and women who have medical co-morbidities such as hypertension or diabetes are at increased risk of hospital admission with COVID-19 during pregnancy. 2 Similar findings have been reported from pregnancy cohorts worldwide 3 and on this basis, vaccination of pregnant women against COVID-19 has been widely recommended.4–6 However, despite these risks of COVID-19 infection in pregnancy being evident from early in the pandemic7,8 pregnant women only began to be recruited into trials of COVID-19 vaccines in mid-2021.
Data from observational human studies currently consists of more than 34,000 completed pregnancies which were mainly exposed to mRNA COVID-19 vaccines (Pfizer-BioNTech and Moderna).9–24 Although these data provide no evidence that mRNA COVID-19 vaccine exposure in the third trimester increases the risk of adverse pregnancy, fetal or neonatal outcomes, the available data are too limited to assess malformation risks following vaccine exposure in early pregnancy. However, given the mechanism of action of these vaccines, which do not contain live SARS-CoV-2 virus, and that there is currently no known fetal risk with using vaccines that are inactivated, replication-deficient, or contain only structural components of the viral pathogen, it is unlikely that these vaccines would be harmful if administered at any stage of pregnancy.
In December 2020, the Medicines and Healthcare products Regulatory Agency (MHRA) approved both the Pfizer/BioNTech and the Oxford University/AstraZeneca COVID-19 vaccines for use in the UK, in adults aged 16 years and over. It was initially advised that the Pfizer-BioNTech or AstraZeneca COVID-19 vaccines should be considered in pregnant women who were at high risk of exposure to SARS-CoV-2, or where the woman had underlying conditions that put them at high risk of serious complications of COVID-19. 5 Whilst electronic data systems were being established, it was important to rapidly set up a study to describe the characteristics of pregnant women undergoing COVID-19 vaccination in the UK and their subsequent maternal/fetal outcomes. Surveys conducted by the Royal College of Obstetricians and Gynaecologists together with the Royal College of Midwives in February 2021 indicated widespread concerns amongst pregnant women considering vaccination about the lack of such evidence. 25
Therefore, a rapid national study of vaccination in pregnancy in the UK was instituted using three existing platforms for medicine safety or vaccine surveillance: the UK Obstetric Surveillance System (UKOSS) combined with the UK Teratology Information Service (UKTIS), and the UK Vaccination In Pregnancy (VIP) surveillance system. This analysis describes the initial findings from this study.
Methods
Study setting and data collection
This national study collected data through three distinct systems; UKOSS combined with UKTIS, and the VIP surveillance system. The UKOSS/UKTIS COVID-19 Vaccination in Pregnancy Monitoring System collected exposure notification data via two routes. UKOSS is an established national research platform which collects reports of severe pregnancy complications from nominated reporting clinicians based at each of the 194 consultant-led hospital maternity units in the UK. 26 For the UKOSS/UKTIS monitoring system, UKOSS requested notifications for all pregnant women who had received any dose of any COVID-19 vaccine at any stage during pregnancy, where vaccination occurred up to 31st March 2021. Requests were made on monthly reporting emails during March and April 2021.
When notified of a vaccination exposure, UKOSS generated an electronic data collection form requesting maternal identifiers, basic demographic details (maternal age and ethnicity), the estimated due date for the pregnancy, and vaccination exposure details (including vaccination date, vaccine manufacturer, batch number, vaccination intent and indication). All notification forms were sent directly from hospitals to the UK Teratology Information Service (UKTIS) via secure National Health Service (NHS) email, and processed for follow-up of maternal and pregnancy outcome. The second route into the UKOSS/UKTIS monitoring system involved direct self-reporting to UKTIS by vaccinated women via telephone or email (active since January 2021). Women who self-reported into UKTIS were asked the same baseline details as collected via the UKOSS data collection form. All reports of COVID-19 vaccination exposure in pregnancy are being followed-up to collect pregnancy outcome information using UKTIS standardised procedures. 27 Details regarding additional medical and obstetric history and confirmation of the original pregnancy details described via UKOSS are also collected at the time of pregnancy outcome collection.
The Immunisation Department at the UK Health Security Agency (UKHSA; formerly Public Health England) operates the UK VIP surveillance system. VIP is an established UK-wide vaccine safety initiative which follows-up pregnancies inadvertently exposed to vaccination either during pregnancy, or shortly before conception. Exposure notification reports are submitted directly to VIP by healthcare professional via telephone, post or email, and vaccinated pregnant women can also report directly. VIP surveillance was extended to include all reports of COVID-19 vaccination in pregnancy up to 15 March 2021.
Following exposure notification, PHE issues a VIP surveillance form to the pregnant woman's General Practitioner (GP) for completion of additional medical and obstetric history and confirmation of pregnancy details. Pregnancy and fetal/neonatal outcome information is subsequently requested from the GP approximately ten weeks after the pregnancy Estimated Date of Delivery (EDD). Infant health and development outcome data are also requested from the primary care team, twelve months after birth for all pregnancies which result in a live birth. Reports of COVID-19 vaccination in pregnancy have been received by VIP since 21 December 2020.
UKOSS/UKTIS and VIP surveillance systems were promoted on UKTIS, Royal College of Obstetricians and Gynaecologists and VIP websites. The VIP surveillance system was additionally promoted in the COVID-19 chapter of the Immunisation Against Infectious Disease handbook 28 published jointly by the Department of Health and Social Care and PHE.
Study sample
This preliminary report consists of all (both prospective and retrospective) reports of COVID-19 vaccination in pregnancy, as submitted to the UKOSS/UKTIS (relating to vaccination up to 31 March 2021) or VIP monitoring systems up to 15 June 2021. Any reports of preconception exposure only and duplicate reports within each of the individual systems were omitted to describe the total number of reports received by the two systems. Duplicate reports received by both the UKOSS/UKTIS and VIP data collection systems were further excluded to describe vaccination exposure details and when the prospectively reported pregnancies are expected to complete.
The exposed study group included pregnancies where one or more vaccination had occurred at any time between the date of the first day of the last menstrual period (LMP) prior to conception, and the pregnancy completion date.
Definitions
Prospective reports of vaccine exposure included pregnancies where the woman was still pregnant at the time of exposure notification, while retrospective reports were received after the pregnancy had completed. The reported Estimated Date of Delivery (EDD) was either projected from ultrasound dating scan results, or calculated as 280 days after the Last Menstrual Period (LMP) date. The LMP was back calculated as 280 days prior to the reported EDD to define exposure periods. The peri-conception period was defined as being from the LMP date to 1 week and six days post-LMP, first trimester from 2 weeks and 0 days to 12 weeks and 6 days, second trimester from 13 weeks and 0 days to 27 weeks and 6 days, and the third trimester as 28 weeks onwards. Exposure indications were defined according to the Joint Committee on Vaccination and Immunisation (JCVI) recommendations in the initial stages of the UK vaccine delivery programme, with women qualifying for vaccination based on either their clinical (based on the JCVI Shielded Patient List 29 ) or occupational 5 risk status (i.e. frontline health or social care workers). Details of the clinical conditions which necessitated vaccination for the clinically vulnerable group were collected. The clinical conditions on the data collection form were not always mutually exclusive, and multiple conditions could be reported. Where multiple conditions were described, a hierarchal assessment was made (by one of the clinical study authors - KKH) as to the most severe condition described and the main disease category that the study participant should be categorised into. Finally, we use the term ‘women’ throughout this manuscript to refer to those who are pregnant and give birth. We acknowledge that not all people who are pregnant and give birth identify as women, and it is important that evidence-based care for maternity, perinatal and postnatal health is inclusive.
Vaccine intent (inadvertent or active choice) was categorised for at least the first vaccine exposure occurring during pregnancy. Intent was only categorised as inadvertent when administration occurred prior to maternal recognition of pregnancy.
Statistical analysis
Continuous variables are described using the median (with interquartile range). Categorical variables are expressed as counts and percentages. Data visualisations were produced using Microsoft Excel.
Case review
Outcome follow-up data from completed pregnancies were individually reviewed after collection. Gestational age at vaccination, concomitant risk factors for adverse pregnancy/fetal/neonatal outcomes, obstetric history, gestational age at delivery and mode of delivery were all carefully considered when assessing any potential aetiological role of vaccination for all adverse outcomes described. Due to the limited data currently available, outcomes are not described in this manuscript, but will be published separately once data are available.
Ethical considerations
In order to rapidly implement data collection existing medicine safety/teratogen surveillance systems in the UK were utilised as the required infrastructure and ethical/information governance approvals were already in place. UKOSS case notification to the vaccine monitoring system was approved by the North London REC1 (Ref. Number: 10/H0717/20). UKTIS national surveillance is provided by section 251 of the NHS Act 2006 and regulation 3 of The Health Service (Control of Patient Information) Regulations 2002 (PHE Caldicott Advisory Panel Approval Number: 13091). VIP surveillance is also conducted under section 251 of the NHS Act 2006 of The Health Service Regulation 2002. The analysis of surveillance data collected by UKTIS and VIP does not require separate approval by a UK Research Ethics Committee.
Results
There were 971 reports of COVID-19 vaccination in the UKOSS/UKTIS (n = 493) and PHE VIP (n = 478) monitoring systems, with exposure notification ongoing. The demographic characteristics for these women are described in Table 1. The majority of the reports described prospective cases of vaccinated women (n = 928, 95.6%), living in England (n = 909, 93.6%), of white ethnicity (n = 583, 60.0%), and with a median age of 33 years (IQR: 30 to 36). The demographics of the vaccinated women were similar between the UKOSS/UKTIS and VIP monitoring systems, with the exception of missing data on ethnicity in the VIP cohort, which will be collected at later follow up.
Demographic characteristics described on the vaccination in pregnancy reports.
Key: LMP = last menstrual period, IQR = Interquartile range.
Includes duplicate pregnancies that were independently reported to both systems (n = 63).
Excludes unknown ethnicity from the denominator to allow easier comparison of maternal ethnicity between the surveillance systems.
After excluding pregnancies that had been reported to both systems (n = 63) there were 908 unique pregnancies across the total dataset. Table 2 describes the vaccination exposure details stratified by vaccine type for these non-duplicate reports. Exposure to the Pfizer-BioNTech mRNA vaccine was most commonly described (n = 501, 55.2%), whilst 356 women (39.2%) received the Oxford University/AstraZeneca COVID-19 viral vector vaccine. Most women received at least one dose of the vaccine in either the second or third trimester (n = 566, 62.3%), with 405 receiving at least one dose in the first trimester (44.6%); the majority of these were exposed in the critical period of organogenesis (2 weeks and 0 days to 10 weeks and 6 days post-LMP, n = 357). This was similar for both the mRNA and viral vector vaccines, with higher proportions receiving at least one dose of the vaccine in the second or third trimester (59.9% and 67.4% respectively) than in the first trimester (51.3% and 36.8%).
Exposure details for the vaccination in pregnancy reports stratified by vaccination type.
Key: CEV = clinically extremely vulnerable, P-BNT = Pfizer BioNTech, AZ = AstraZeneca.
These figures represent the number of women receiving at least one dose of the vaccine during the corresponding exposure periods, and as such the categories are not mutually exclusive.
Details of the maternal conditions are provided in Figure 1.
Other reported indications for vaccination included risk from residential settings and maternal age aligning with the national programme rollout (VIP only).
At the time of reporting most women had only received one vaccination (n = 644, 70.9%), and in most cases this was their first vaccination of the two dose schedule (n = 573, 63.1%), with a smaller proportion having received both vaccinations during pregnancy (n = 234, 25.8%).
The majority of the women actively sought to be vaccinated during pregnancy (n = 545, 60.0%), with a smaller proportion inadvertently vaccinated in pregnancy (n = 307, 33.8%). Maternal occupational risk was the indication for vaccination in a higher number of women (n = 577, 63.5%) compared to clinical vulnerability (n = 228, 25.1%). The majority of those vaccinated due to occupational risk were healthcare or social care workers (n = 551, 95.5%). The various conditions reported among the clinically vulnerable sub-group are described in Figure 1.
Clinical conditions that were reported for those vaccinated due to clinical vulnerability.
Figures 2 and 3 describe the cumulative number of prospectively reported pregnancies reaching their estimated due date by month and year, stratified by vaccine type (Figure 2) and trimester of exposure (Figure 3).
Expected number of full term (40 weeks gestational age) pregnancies stratified by vaccine type.
Expected number of full term (40 weeks gestational age) pregnancies stratified by trimester at exposure.
Discussion
Data regarding pregnancy outcome following COVID-19 vaccination in the UK are accumulating, and through combination of the UKOSS/UKTIS and VIP safety monitoring systems, outcomes for a cohort of approximately 900 vaccinated women is expected to be available for analysis by December 2021. Currently, approximately 300 women have reached their estimated due date, and follow-up of these mainly third trimester exposed pregnancies is underway. Data regarding outcomes of first trimester exposed pregnancies are also beginning to accumulate. The results of this preliminary analysis describe the study population which will be included in the December 2021 analysis of pregnancy outcome data. As expected, given the national vaccination recommendations at the time the study sample was collected, the indication for vaccination was mostly due to occupational risk or clinical vulnerability. Accordingly, exposure to the Pfizer-BioNTech mRNA vaccine, which requires stringent cold-storage conditions and was predominantly used to vaccinate healthcare workers at hospital vaccination centres, was more common than the Oxford University/AstraZeneca viral vector vaccine.
Since April 2021, national databases have been modified to allow the identification of women who knew they were pregnant at the point of vaccination. Data linkage studies are underway, linking vaccination to obstetric outcome data obtained through routine clinical care (e.g. Hospital Episode Statistics in England 24 and COPS (COVID-19 in Pregnancy in Scotland) study in Scotland 18 ). Automating data collection in this way provides an opportunity for improved data capture within the population; however, it is unlikely that these studies will produce data imminently. The UKOSS/UKTIS/VIP study aims to fill the immediate need for safety data, and to reassure pregnant women that the safety of these vaccines is being monitored. Although a formal statistical analysis of pregnancy outcome data will not be performed before the end of the year, preliminary safeguarding techniques are being undertaken. Any prospective follow-up outcome reports of adverse pregnancy, fetal or neonatal events are being closely reviewed by experienced clinicians and pregnancy pharmacovigilance specialists to assess any potential aetiological role of vaccination. Patterns of adverse outcomes that are unexplained by concomitant risk factors will be rapidly communicated to key public health and medicines regulatory stakeholders. Another key strength of the UKOSS/UKTIS/VIP study is that the indication for vaccination has been clarified for the majority of exposure reports. Vaccination in the UK was initially limited to healthcare and social care workers and those in a ‘Clinically Extremely Vulnerable’ (CEV) group. It is therefore anticipated that a proportion of the women initially vaccinated at the start of the programme would be at increased risk of adverse pregnancy outcome due to their pre-existing health conditions. Careful interpretation of outcome data from this group, which represents approximately 25% of the cohort, will be necessary. Comparisons of pregnancy outcomes between the CEV group and women vaccinated due to other indications will aid this interpretation. This will allow distinctions to be made between possible vaccine related adverse effects and any consequences of the underlying maternal medical condition.
There are limitations to this study. Whilst pregnancies reported through UKOSS had good ascertainment through active null reporting (i.e. obstetric units provide monthly exposure data, even if this is zero) pregnancies reported through UKTIS and VIP rely on active voluntary reporting, and as such there may be a bias towards reporting those with adverse pregnancy outcomes. This may be particularly problematic for pregnancies where the outcome was known prior to reporting (retrospective adverse outcome reporting bias). However, detailed information is being collected in both the UKOSS/UKTIS and VIP monitoring systems which will allow strict definitions of prospective and retrospective cases to be applied. Through restricting analysis to cases which meet true prospective case definition, the impact of the sampling bias in the UKTIS and VIP datasets will be limited. An additional limitation is that as time elapses and vaccination uptake increases the UKOSS/UKTIS/VIP study cohort will likely represent a small sample of the wider vaccinated pregnant population. Although ethnic diversity was represented in our sample with 7% Asian, 3% Black, 1% Chinese and 2% of mixed ethnicity, this was not quite in keeping with estimates of the national UK ethnicity demographics 30 where approximately 12% Asian, 5% Black, 5% Chinese and 2% of mixed ethnicity would be expected. Possible reasons for the variation in ethnicity could relate to the documented vaccine hesitancy of non-white ethnicity in the general population of the UK31,32. Of note, there was considerable missing data regarding maternal ethnicity (28.9%), which is mainly explained by the VIP system collecting this data on pregnancy follow-up rather than at exposure notification. Further attempts to obtain missing ethnicity data will be undertaken when collecting obstetric outcome data. Finally, although this study was national across the UK, fewer than 7% of the reports were obtained from outside England. Caution may need to be applied on interpretation of the final study results in their applicability to the wider general population of the UK.
Around the world a number of pregnancy vaccine surveillance systems have reported initial maternal and fetal outcome data.9–24 Most of these datasets consist of vaccinations which occurred during the second or third trimester, and outcome data have now been published on more than,24,000 completed pregnancies.9–24 Collectively, these data do not provide evidence indicating adverse effects in either the mother or the baby. Although the majority of these data are not currently suitable for assessing risks pertaining to early pregnancy vaccine exposure and malformation risks, four studies have provided preliminary data indicating that miscarriage is not associated with COVID-19 mRNA12,15 or viral vector13,16 vaccination. The UKOSS/UKTIS/VIP cohort includes over 350 cases of Oxford/AstraZeneca viral vector vaccine exposure. In combination with the Public Health Scotland dataset which includes more than 1000 women who have received the Oxford/AstraZeneca vaccine, these data will provide important additional information on the safety of this viral vector vaccine in pregnancy (noting that this vaccine is not yet authorised for use in US and therefore US reports have not included outcome data). Since guidelines regarding vaccine type have changed in the UK to favour mRNA COVID-19 vaccines in pregnancy, 33 substantial further UK data are unlikely to accumulate. However, vaccines based on viral vector technology are useful for other emerging pathogens of public health concern, for example Ebola virus, 34 and may be relevant for further COVID outbreaks in countries where mRNA vaccines are unavailable, and for other emerging epidemics in future.
Initial COVID-19 vaccine trials excluded pregnant women. However, COVID-19 infection is potentially harmful to both the woman and baby, and emerging variants are showing signs of increased transmissibility and pathogenicity. 35 If pregnant women had been included in initial vaccine trials, then at least some safety data would be in the public domain at the time of the initial vaccination rollout. However, this did not occur and therefore it is vital that data are routinely collected on as many pregnant women as possible to provide further reassurance. Our cohort will report approximately 900 women who received the vaccine in the early stages of the UK vaccination programme. Meanwhile, the COVID-19 pandemic has reached a critical stage, and pregnant women, a largely unvaccinated population, are at significant risk. Women should be advised not to delay vaccination whilst awaiting further safety data to emerge, since doing so puts them at risk of COVID-19 at a time when they are particularly vulnerable.
Supplemental Material
sj-pdf-1-obm-10.1177_1753495X221076713 - Supplemental material for Monitoring the safety of COVID-19 vaccination in pregnancy in the UK: A national study using the UK Obstetric Surveillance System (UKOSS), UK Teratology Information Service (UKTIS) and Vaccination in Pregnancy (VIP) safety monitoring systems
Supplemental material, sj-pdf-1-obm-10.1177_1753495X221076713 for Monitoring the safety of COVID-19 vaccination in pregnancy in the UK: A national study using the UK Obstetric Surveillance System (UKOSS), UK Teratology Information Service (UKTIS) and Vaccination in Pregnancy (VIP) safety monitoring systems by Jonathan L Richardson, Sally Stephens, Lucy C Chappell, Helen Campbell, Gayatri Amirthalingam, Shennae O’Boyle, Antoaneta Bukasa, Marian Knight and Kenneth K Hodson in Obstetric Medicine
Footnotes
Acknowledgements
All reporters to the UKOSS, UKTIS and VIP systems, staff who have supported the UKTIS phone service during the study period (Mr Nathan George, Dr Amanda Greenall, Ms Elaine Donohoe, Dr Tabassum Moonira and Ms Jane Ingram), RCOG communications team, and the UKOSS administration team (Melanie O’Connor, Hatty Goddard, Anna Balchan).
Declaration of conflicting interests
JLR, SS and KKH are employees of the Newcastle upon Tyne Hospitals NHS Foundation Trust, LCC is an employee of King's College London, HC, GA, SOB and AB are employees of UK Health Security Agency, and MK is an employee of the University of Oxford. All authors and contributors declare no conflicts of interest; no other financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no previous or ongoing relationships or activities that could appear to have influenced the submitted work.
Ethical approval
UKOSS case notification to the vaccine monitoring system was approved by the North London REC1 (Ref. Number: 10/H0717/20). Approval for UKTIS national surveillance is provided by section 251 of the NHS Act 2006 and regulation 3 of The Health Service (Control of Patient Information) Regulations 2002 (PHE Caldicott Advisory Panel Approval Number: 13091). VIP surveillance is also conducted under section 251 of the NHS Act 2006 of The Health Service Regulation 2002. The analysis of surveillance data collected by UKTIS and VIP does not require separate approval by a UK Research Ethics Committee.
Funding
The UK Teratology Information Service is commissioned by the UK Health Security Agency and hosted by Newcastle upon Tyne Hospitals NHS Foundation Trust. The Vaccination In Pregnancy surveillance system is run by the Immunisation and Countermeasures National Infection Service at the UK Health Security Agency. MK and LC are NIHR Senior Investigators. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. The funder played no role in study design; data collection and analysis, decision to publish or preparation of the manuscript.
Guarantor
Marian Knight and Kenneth K Hodson (joint senior authors)
Author Contributions
All of the authors have contributed equally to the study. Study design – MK, KKH, JLR, LC, HC; Data analysis –MK, KKH, JLR, HC, SOB; Manuscript preparation – all authors
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
