Description of maternal and neonatal adverse events in pregnant people immunised with COVID-19 vaccines during pregnancy in the CLAP NETWORK of sentinel sites. Nested case-control analysis of the immunization-associated risk: A study protocol.

Introduction

Pregnant people do not appear to have a higher susceptibility to SARS-CoV-2 coronavirus infection than non-pregnant people.1 However, COVID-19 is associated with higher morbidity and mortality among pregnant people compared with non-pregnant people.2 3 Mortality in pregnant people with COVID-19 could be 70% higher, and the risk of admission to intensive care units (ICUs) or the need for mechanical ventilation would be three times higher compared with non-pregnant unvaccinated people of similar age.3 4 Pregnant people with COVID-19 show an increased risk of occurrence of some adverse maternal–fetal outcomes such as increased risk of preterm delivery, pre-eclampsia or miscarriage compared with pregnant people without COVID-19.2 It has been observed that, in its most complicated or critical forms, COVID-19 infection is associated with prematurity, gestational diabetes, pre-eclampsia and low birth weight.3 Neonates born to mothers with COVID-19 appear to have a higher probability of admission to a neonatal care unit, which could be attributed to a greater extent to the negative effect of COVID-19 on maternal health during gestation than to vertical transmission.4–6 In addition, other maternal, fetal and neonatal complications associated with COVID-19 have also been reported. A systematic review described adverse events such as fetal distress, premature rupture of membranes and other neonatal adverse events such as respiratory distress, gastrointestinal events and fever. Maternal death reports have been more rare.7

After a global research effort to develop effective treatments and vaccines against COVID-19, vaccines became available on January 2021. It should be noted that a characteristic of the new COVID-19 vaccines is that they were developed from a range of different platforms: viral vector (adenoviral), mRNA platform, inactivated, attenuated viruses, being pregnancy an exclusion criterion in the main pivotal clinical trials undertaken for the authorisation of these COVID-19 vaccines.8 In addition, vaccines that are currently recommended for maternal immunisation—and with which there is most experience—such as the influenza vaccine or the Tdap vaccine, do not use mRNA-based platforms as is the case with the Pfizer-BioNTech and Moderna vaccines.9 However, despite being based on a newer platform, mRNA vaccines do not entail the risk of acquiring vaccination-induced infection, as might be the case with the use of attenuated/inactivated vaccines. On the other hand, they could be useful in pregnant people, as a robust transfer of the humoral response to the fetus has been observed with mRNA vaccines.10 11 Furthermore, no remarkable adverse events have been detected in experimental genotoxicity studies conducted on rats.12

Although various clinical trials have been planned for the study of COVID-19 vaccines in pregnant people, most of the initiatives are based on the creation of registries of pregnant people exposed to the vaccines. In many cases, it is the adaptation of previous registries that had been created to study the effects of COVID-19 infection in pregnant people. In this respect, Moderna created a registry to monitor pregnancy outcomes in individuals who received their COVID-19 vaccine. PREGISTRY—for example—is a startup focused on improving maternal–fetal health, in collaboration with the Harvard School of Public Health (https://www.pregistry.com); aim to enrol 8000 pregnant people in their registry to evaluate their obstetric, neonatal and infant outcomes following COVID-19 immunisation. In the UK, the UKOSS registry (https://www.npeu.ox.ac.uk/ukoss) is collecting data to allow the parallel nesting of cohort and case–control studies. The CDC’s ‘v-safe’ registry has reported preliminary results with observed frequencies of spontaneous abortion and fetal death in vaccinated pregnant people similar to the expected frequencies in the general population of pregnant people of the same age and without known comorbidities.13

Apart from these active registries that include pregnant people, there is little evidence from published studies with complete data on the effectiveness and safety of vaccination for COVID-19 in this population and information from surveillance systems, monitoring acute adverse events following immunisation during pregnancy, did not find increased risk of clinically serious acute adverse in different populations groups.14 15 One retrospective cohort study observed that, in pregnant people, vaccination with Pfizer-BioNTech was associated with a lower risk of SARS-CoV-2 infection compared with those who did not receive vaccination (HR 0.22; 95% CI 0.11 to 0.43).16 In another prospective cohort study, short-term adverse events associated with mRNA vaccines were similar among pregnant people to those occurring in non-pregnant people of the same childbearing age.17 In another large comparative study involving 157 521 singleton pregnancies from the Swedish pregnancy and Norwegian childbirth electronic registries, no statistically significant differences were found in the risk for fetal death, small for gestational age rates, preterm and low Apgar score (Appearence, Pulse, Grimace, Activity and Respiration) value in mRNA and Pfizer-BioNTech vaccinated and unvaccinated subjects.18 Also, the risk of miscarriage does not appear to be higher than expected among people immunised before or during pregnancy.19 20 Moreover, in a retrospective cohort including more than 40 000 pregnant people, COVID-19 vaccines (mostly mRNA) were not associated with preterm birth or small-for-gestational-age at birth.21 On the other hand, a case–control study using Norwegian registry data observed no statistically significant risk difference for first-trimester miscarriage in pregnant people exposed to COVID-19 vaccines (Pfizer-BioNTech, Moderna, Astra) and those not immunised.22 A study investigating on the associations of COVID-19 vaccination throughout pregnancy with delivery and neonatal outcomes suggests that there is no associated risk between prenatal COVID-19 vaccination and adverse delivery and neonatal outcomes in a cohort sample from NYC.23 A systematic review and meta-analysis included data from 23 studies and 117 552 pregnant people immunised mostly with mRNA vaccines. In this meta-analysis, which accumulates the most evidence on the subject to date, no differences were found in vaccinated compared with unvaccinated pregnant people with regards the risk of occurrence of miscarriage, earlier gestation at birth, placental abruption, pulmonary embolism, postpartum haemorrhage, maternal death, mothers’ admission to the ICU, low birth weight or newborns’ admission to the ICU. The meta-analysis identified that immunisation conferred a 15% protective effect against fetal death, although the heterogeneity found was very high (OR 0.85; 95%CI 0.73 to 0.99; I²=93.9%).24 Some absolute estimates of the impact of vaccination with mRNA vaccines have been proposed to assess the number of maternal and neonatal adverse outcomes that could be avoided such as ‘number needed to vaccinate (NNV)’. The impact measures suggest a benefit in avoiding certain events such as COVID-19 infection (NNV:11), caesarean section delivery (NNV: 182) and prematurity (NNV: 200), while the NNV would be higher for ‘neonatal problems’ (NNV: 463), severe cases of COVID-19 (NNV: 412–2058), requirement for mechanical ventilation (NNV: 1371–6857) and small for gestational age (NNV: >2000). In all cases far from the ‘number needed to harm’ (NNH) estimated for some adverse events: myocarditis (NNH: 37 000) or thrombosis syndrome with thrombocytopenia (NNH: 50 000).25 estimates carried out based on observational data in Israel to determine the ‘number needed to boost’ and assess the impact of the third booster dose of the Pfizer-BioNTech vaccine in preventing hospital admissions for COVID-19 and complicated COVID-19 would also suggest a greater potential benefit for pregnant people of advanced age or with comorbidities receiving booster doses.26

However, majority of evidence suggesting the possible beneficial effect of vaccination comes from descriptive observational studies. A large population-based descriptive study included more than 180 000 pregnant people in Scotland. The results suggested a possible benefit of vaccination. Thus, 77.4% of all COVID-19 infections, 90.9% of hospital admissions, 98% of ICU hospitalisations and 100% of neonatal deaths occurred in pregnant people diagnosed with COVID who had not been vaccinated.27 In addition, descriptive data from six European countries from the International Network of Obstetric Survey Systems have also reported that the highest proportion of ICU admissions of pregnant people corresponded to those who had not been vaccinated (80%–100% of the total).28 On the other hand, in a study in the USA from electronic medical record data in women who had miscarriages, no differences were found in the probability of having been vaccinated (mRNA vaccines) 28 days before the occurrence of the miscarriage.29

At present, the benefits of maternal immunisation are considered to outweigh the risks and for that reason vaccination is recommended at any time during pregnancy.1 30 31 The recommendation is supported, on the one hand, by the robust evidence indicating a higher morbidity and mortality associated with COVID-19 among pregnant people compared with non-pregnant people; and on the other hand, by the protective effect of maternal immunisation observed against complications of COVID-19. Moreover, no pharmacovigilance signals associated with maternal immunisation have been detected13 14 32 33 Finally, some studies suggest a decreased risk of occurrence for certain adverse fetal and neonatal outcomes associated with maternal immunisation, although these are considerations based mostly on observational evidence, and in many cases with the caveat that further studies are needed. However, the great difference in the amount of evidence available according to the type of vaccine is striking. Most studies in pregnant people have been carried out with vaccines developed from the newest mRNA platform (Pfizer-BioNTech, Moderna), with very little evidence available for the other vaccines. Perhaps for this reason, some countries prioritise the use of mRNA vaccines in their recommendations, leaving the administration of other (non-mRNA) vaccines with a weaker recommendation based on clinical criteria, weighing the possible risk of infection of the mother considering certain circumstances (eg, health workers, comorbidities) with the theoretical risks that the uncertainties of immunisation may entail.34 35 In spite of the available evidence, most countries that have reported vaccination coverage in pregnant people had low vaccinations rates.36 For example, in Scotland, only 32.3% of pregnant people had received the full scheme (vs 77.4% of non-pregnant people of childbearing age).27 Other countries have also reported low coverage rates among pregnant people such as Italy (20%) or England (22%) or 30%–50% in the Netherlands.28

The Latin American Center of Perinatology (CLAP) coordinates a NETWORK of sentinel centres in Latin American and Caribbean countries for the surveillance of maternal health related issues (see table 1). All centres use the Perinatal Information System (SIP, for its acronym in Spanish) on its plus version (SIP-PLUS) as a common data collection system. SIP-PLUS database contains the perinatal electronic health record data (online supplemental material 1). The perinatal medical record prospectively and systematically records individual data on the clinical follow-up of pregnant people seen from the date of the first antenatal visit to the date of discharge following delivery. It is a healthcare database of perinatal electronic medical records that centralises the demographic, diagnostic, obstetric and neonatal clinical data of mothers and their offspring. In addition, it incorporates the systematic collection of information on the use of vaccines (tetanus/diphtheria, Tdap, influenza, rubella, hepatitis B, hepatitis A), as well as information on dosage, date of vaccination and time of vaccination (before or during pregnancy, or after delivery).37 In response to the outbreak of the COVID-19 pandemic, CLAP developed a specific SIP form to monitor COVID vaccination of pregnant people, along with other related variables, such as COVID-19 history, which have become part of the routine follow-up of pregnant people (online supplemental material 2).

Therefore, this project aims to retrospectively describe the use of COVID-19 vaccines among pregnant people and explore the risk associated with vaccination and selected adverse events, adapting the SIP-PLUS database for surveillance of the use and safety of COVID-19 vaccines from data routinely collected during obstetric and neonatal monitoring of pregnant people in a network of Latin American centres.

Methods and analysis

Population, outcome variables, exposure and others

Retrospective description of vaccination and event frequency in the pregnant people cohort

Inclusion criteria

All pregnant people admitted for delivery (singleton) or miscarriage from October 2022 to July 2023.

Exclusion criteria

Twin pregnancies will be excluded.

Adverse events of interest

1. Premature births (preterm delivery)

   Born alive at less than 37 weeks gestational age as registered in the SIP form clinical record.

2. Small for gestational age. Live birth classified as small by country standards.

3. Low birth weight. Born alive weighing less than 2500 g.

4. Stillbirth (at any time).

5. Non-extended neonatal death (max 48 hours post partum).

6. Congenital malformations (major and minor).

7. Maternal death (at any time).

8. Maternal near miss (MNM). Pregnant people with an MNM condition according to WHO MNM approach.38

9. Abortion (spontaneous abortion and therapeutic abortion).

Covariates

Sociodemographic variables:

• Mother’s age (years).

• Ethnicity (white, indigenous, mixed, other).

• Marital status (married, common-law marriage, single, other).

• Educational level (none, elementary, high school, college, university).

• Country

• Hospital.

Obstetric variables and risk factors (definitions)

• Date of last menstrual period (LMP).

• Gestational age (weeks)

• Smoking (yes/no).

• Drug use (yes/no).

• Alcoholism (yes/no).

• Induced labour (yes/no).

• Diabetes

  • Pre-existing (yes/no)

  • Gestational (yes/no).

• High blood pressure

  • Pre-existing (yes/no)

  • Gestational (yes/no).

• Eclampsia (yes/no).

• Pre-eclampsia (yes/no).

• Parity (nulliparous, multiparous).

• Intergestational period (months).

• COVID-19 infection (describe diagnosis, prior to or during pregnancy).

Administration of other recommended vaccines during pregnancy

Exposure of interest (immunisation with COVID-19 vaccines)

Type of vaccine, number of doses administered and date.

The vaccines of interest will be33

• Astra-Zeneca (AZD1222; ChAdOx1 nCoV-19; ChAdOx1-S; Vaxzevria, Covishield).

• Janssen (NJ-78436735; Ad26.COV2-S).

• Moderna (mRNA-1273; Elasomeran).

• Pfizer-BioNTech (BNT162b2; Tozinameran; Comirnaty).

• Sinopharm (Sinopharm/BIB).

• Sinovac (CoronaVac; VeroCell).

• Sputnik-Gamaleya (Gamaleya; Gam-COVIDVac; Adeno-based (rAd26-S+rAd5 s).

The rest of vaccines not considered in the list will be recorded in free text as ‘other vaccines’. When the administration of two different vaccines is included, the vaccination will be classified as ‘heterologous schedule’. On the other hand, it will be considered a complete schedule if the pregnant person received two doses (or one dose in case of Janssen’s vaccine) and a booster schedule if she has received three doses (or two doses in case of Janssen’s vaccine).

Analysis of the risk associated with vaccination for the occurrence of the neonatal events of interest through an exploration of case–controls nested in a cohort

A matched case–control approach nested in a cohort will be carried out for each of the events of interest.

Cohort definition

• Pregnant people (preterm to late term) and with mother’s age data.

• Live births with delivery date available based on the best available obstetric estimator.

• Perinatal medical record with ‘COVID-19 vaccine module’ implemented.

Case definition (events of interest)

1. Premature. Preterm delivery: live birth at less than 37 weeks gestational age according to the best available obstetric estimator.

2. Small for gestational age: live birth classified as small based on country standards.

3. Low birth weight: born alive weighing less than 2500 g.

Definition of controls

Two groups of three (uneventful) controls will be randomly selected for each case. The first group of controls will be randomly selected at the same delivery date of each case matched by delivery hospital and mother’s age (±3 years). The second group of controls will be births randomly selected from the estimated conception time of the cases based on the estimated delivery date (EDD). EDD will be estimated according to the best available estimator considering the following characteristics: (1) early ultrasound and (2) clinical (LMP). If EDD is not available, gestational age considered during newborn examination should be use. Second group of controls will be also matched by delivery hospital and mother’s age (±3 years).

Definition of exposure

Vaccines considered will be those mentioned in the Exposure of interest (immunisation with COVID-19 vaccines) section.

Exposure shall be classified according to the following exposure windows:

1. ‘Late immunisation’: pregnant people immunised on the index date or up to one trimester before the end of pregnancy (third trimester).

2. ‘Second trimester immunisation’: pregnant people immunised after completing the first trimester until the second trimester before the event.

3. ‘First trimester immunisation’: pregnant people immunised during the first trimester.

4. ‘Past immunisation’: immunised prior to pregnancy

5. ‘Immunisation at any time’: immunisation at any of the above times.

Moreover, gestational age (weeks) will be considered if available (online supplemental material 2)

Figure 1 summarises the overall conceptual framework and procedures of this study.

Figure 1

Overview of the study procedure for data collection and outcomes of interest.

Expected sample size and rationale for its determination

General considerations

In estimating the sample size, the frequencies of events observed in a previous analysis in the SIP-PLUS database were considered to estimate the number of potential events (online supplemental material 3). Given the lack of precise data on vaccination coverage in hospitals in the CLAP network, literature values from studies published in other countries (prevalences ranging from 20% to 50%) were used as a reference.27 28 The design of the contrast tests as bilateral (or two tailed) is because the hypothesis contemplates the scenario of a possible protection against adverse events associated with vaccines and also the opposite scenario, with an increased risk. The correlation coefficient (r) between (matched) cases and controls is not known and cannot be estimated from previous studies. Hence, a conservative value of ‘r=0.2’ has been assumed for the calculation of the sample size, following the recommendation of some authors for these cases.39 In this part, we will consider the analysis of events for which, a priori, there would be sufficient statistical power, anticipating that the measures of association (OR) may not be to large based on existing published studies.

Supplemental material

Main analysis of vaccination (any vaccine) with respect to ‘non-use’

For the matched case–control study (matched with three controls) with a power of 80%, alpha error=0.05 (bilateral), a case–control correlation coefficient of 0.2 and a vaccine exposure of 20% would require at least 1009 cases (3027 controls) to detect an increased probability of vaccine-associated event risk of 30% (OR=1.3) and at least 650 cases (1950 controls) to detect 30% protection (OR=0.7) (figures 2 and 3).

Figure 2

Power of the analysis as a function of the number of cases for different probabilities of exposure to vaccines from 20% to 50% for the detection of an OR=1.3 (increased risk).

Figure 3

Power as a function of the number of cases for different probabilities of exposure to vaccines from 20% to 50% for the detection of an OR=0.7 (protection).