Study design and population

In this retrospective cohort study, the records of all women who delivered between March 1, 2021, and July 4, 2021, in St George’s University Hospitals NHS Foundation Trust, London, United Kingdom, were screened for eligibility. The first of March was chosen as pregnant women with comorbidities were offered vaccination after this date. The inclusion criteria were pregnant women with known vaccination status and complete maternal and fetal outcome data. The exclusion criteria were women who were vaccinated entirely (ie, all doses) before pregnancy or after birth or women who had pregnancies complicated by fetal aneuploidy or genetic syndromes.

Data were obtained from electronic hospital records stored in the following systems: EuroKing E3 maternity information system, Chertsey, United Kingdom; ViewPoint Bildverarbeitung GmbH, Wessling, Germany; and Health Information Exchange—Coordinate My Care, GP Partner 2, One London Hub, One London SWL, MIG Partner, and TPP SystmOne—via Cerner, North Kansas City, Missouri. The quality of data in the later database was checked for accuracy of the COVID-19 vaccination record. For 1 week, women admitted to the postnatal ward were queried about their vaccination status, and their responses were cross-checked with the electronic records. The initial quality control showed no inconsistency between patient responses and electronic records.

Data collected included maternal age, parity, index of multiple deprivation (IMD) score, self-reported ethnicity, body mass index (BMI), alcohol and smoking habits, comorbidities, antenatal complications, and medications. The IMD combines multiple deprivation indices (income, employment, education, health, crime, barriers to housing and services, and living environment) into a single score and is widely used to assess deprivation in the United Kingdom. Other data variables included COVID-19 vaccination uptake, vaccination type, and gestational age (GA) at vaccination. Other data of interest included antenatal complications, including gestational diabetes mellitus, obstetrical cholestasis, and preeclampsia; venous thromboembolism (given the association with the Oxford-AstraZeneca vaccine) or myocarditis or pericarditis (given the association with the Pfizer-BioNTech vaccine); and antenatal medication, including medications for chronic prepregnancy disorders (such as hypothyroidism, epilepsy, diabetes mellitus, or depression) or pregnancy conditions (such as gestational hypertension or preeclampsia), but not nutritional supplements, multivitamins, iron replacement, antibiotics, antiemetics, analgesia or anti-D immunoglobulin.

Vaccine types included messenger RNA (mRNA) vaccines (Moderna, Pfizer-BioNTech) and viral vector vaccine (Oxford-AstraZeneca), which are approved for use in the United Kingdom. As for the general population, pregnant women were eligible for vaccination if they (1) were a health or social care worker, which increased their risk of SARS-Cov-2 infection, or (2) were at high risk of severe COVID-19 because of personal factors (eg, non-White ethnicity) or health conditions (eg, diabetes mellitus or gestational diabetes mellitus specifically). From April 16, 2021, vaccination was offered to those aged 45 years and above, with progressively younger groups offered vaccination from May 28, 2021. From April 16, 2021, the JCVI has advised that mRNA vaccines should be used in preference in the United Kingdom for pregnant women.

Study outcomes

The primary study outcome was COVID-19 vaccine uptake during pregnancy among women eligible for vaccination. Vaccination eligibility was assessed by comparing delivery date with vaccination eligibility date based on the mother’s age and priority category. In the United Kingdom (2021), women became eligible for vaccination based on clinical risk and maternal age, with vaccination offered to pregnant women at the same time as the rest of the population, as follows: >40 years (from April 30, 2021), >30 years (May 26, 2021), >25 years (June 8, 2021), and >18 years (June 18, 2021). ^, Women who delivered after the vaccination eligibility date for their age category were considered eligible for antenatal vaccination. Vaccination uptake rate was calculated as the number of women who received at least 1 dose of any COVID-19 vaccine during pregnancy in a certain group divided by all women eligible for vaccination in that group.

Secondary outcomes included perinatal outcomes to assess the safety of COVID-19 vaccination, which included stillbirth (fetal death at ≥24 weeks’ gestation), neonatal death, fetal abnormalities, preterm birth before 37 weeks’ gestation, GA at birth in weeks, intrapartum complications (pyrexia, suspected chorioamnionitis, placental abruption, and postpartum hemorrhage), birthweight z score, mode of birth (cesarean delivery, instrumental delivery, or unassisted vaginal delivery), maternal high-dependency unit or intensive care unit (ICU) admission (any indication), and neonatal ICU admission (any indication). Postpartum hemorrhage was defined as an estimated blood loss of ≥1 L regardless of the mode of birth.

GA was determined in the first trimester of pregnancy according to the crown-rump length of the fetus (singleton pregnancies) or the larger fetus (twin pregnancies) in cases of spontaneous conception and according to the embryonic age in pregnancies conceived via assisted reproductive technology (ART). After 14 weeks’ gestation, GA was determined using the head circumference of the fetus (singleton pregnancies) or the larger fetus (twin pregnancies) in cases of spontaneous conception and according to the embryonic age in pregnancies conceived via ART. Birthweight z scores were calculated using a reference standard published by Poon et al.

Statistical analysis

Continuous variables were represented as median and interquartile range (IQR) regardless of the distribution assumptions. Categorical variables were represented as numbers and percentages. The Shapiro-Wilk test and visual inspection of quantile-quantile plots were used for verifying the normality of continuous variables. Mann-Whitney U test, t test, chi-square test, or Fisher-Freeman-Halton test was used for group comparisons as appropriate.

Factors associated with antenatal vaccination were assessed among all women eligible for vaccination, by logistic regression, with factors significant in univariable analyses assessed in a multivariable model to calculate the adjusted odds ratios (ORs). Results of regression analyses were reported as mean difference (MD), OR, or hazard ratio (HR).

The effect of COVID-19 vaccination on perinatal outcomes was assessed among women who had antenatal vaccination, compared with those who did not have the vaccine. Propensity score matching was used to match cases and controls for factors identified from between-group comparisons as potentially differing ( P <.10). Groups were matched 1:3 to simulate the observed vaccination uptake. The success of propensity score matching was assessed with propensity score histograms. After matching, the effect of vaccination was estimated using generalized estimation equations using matched group identifiers as cluster labels. Effect magnitudes of vaccination on perinatal outcomes were reported as MD or OR with 95% confidence intervals (CIs). GA at delivery was assessed in a separate cohort matched for the expected date of delivery, and confounders were identified for other perinatal outcomes. Mixed-effects Cox regression was used to estimate the effect of vaccination on GA at birth <40 weeks’ gestation. In a sensitivity analysis for pregnancy outcomes, women with antenatal COVID-19 were excluded. All analyses were performed using R for statistical computing software (version 4.0.2; R Foundation for Statistical Computing, Vienna Austria).