Neonatal and maternal outcomes following SARS-CoV-2 infection and COVID-19 vaccination: a population-based matched cohort study

Understanding the impact of SARS-CoV-2 infection and COVID-19 vaccination in pregnancy on neonatal and maternal outcomes informs clinical decision-making. Here we report a national, population-based, matched cohort study to investigate associations between SARS-CoV-2 infection and, separately, COVID-19 vaccination just before or during pregnancy and the risk of adverse neonatal and maternal outcomes among women in Scotland with a singleton pregnancy ending at ≥20 weeks gestation. Neonatal outcomes are stillbirth, neonatal death, extended perinatal mortality, preterm birth (overall, spontaneous, and provider-initiated), small-for-gestational age, and low Apgar score. Maternal outcomes are admission to critical care or death, venous thromboembolism, hypertensive disorders of pregnancy, and pregnancy-related bleeding. We use conditional logistic regression to derive odds ratios adjusted for socio-demographic and clinical characteristics (aORs). We find that infection is associated with an increased risk of preterm (aOR=1.36, 95% Confidence Interval [CI] = 1.16–1.59) and very preterm birth (aOR = 1.90, 95% CI 1.20–3.02), maternal admission to critical care or death (aOR=1.72, 95% CI = 1.39–2.12), and venous thromboembolism (aOR = 2.53, 95% CI = 1.47–4.35). We find no evidence of increased risk for any of our outcomes following vaccination. These data suggest SARS-CoV-2 infection during pregnancy is associated with adverse neonatal and maternal outcomes, and COVID-19 vaccination remains a safe way for pregnant women to protect themselves and their babies against infection.

**Adjustment for deprivation only for analyses with the following outcomes: spontaneous very preterm birth and provider-initiated very preterm birth.Adjustment for parity and deprivation only for analyses with the following outcomes: stillbirth, extended perinatal death, very low Apgar score and venous thromboembolism.Adjustment for all covariates apart from parity for the following outcomes: preterm birth, spontaneous preterm birth, provider-initiated preterm birth and very preterm birth.Adjustment for all covariates for analyses for the following outcomes: small for gestational age, very small for gestational age, low Apgar score, maternal critical care or death, hypertensive disorders of pregnancy and pregnancy-related bleeding.***For these outcomes there is not an exact match of three uninfected controls to each neonate exposed to infection as neonates with missing outcome data have been removed from analysis; see Supplementary Table 11 for further details on levels of missing data.
Supplementary Table 5: Association between exposure to SARS-CoV-2 during pregnancy at 6 weeks preconception to ≤19+6 weeks gestation and neonatal and maternal outcomes (reinfections spanning both gestational periods excluded from subgroup analysis), calculated using conditional logistic regression.**Adjustment for deprivation only for analyses with the following outcomes: very preterm birth, spontaneous very preterm birth and providerinitiated very preterm birth.Adjustment for parity and deprivation only for analyses with the following outcomes: extended perinatal death, very small for gestational age and low Apgar score.Adjustment for all covariates for analyses apart from parity for preterm birth, spontaneous preterm birth and provider-initiated preterm birth.Adjustment for all covariates for analyses for the following outcomes: small for gestational age, very small for gestational age, low Apgar score, maternal critical care and/or death, hypertensive disorders of pregnancy and pregnancyrelated bleeding.***For these outcomes there is not an exact match of three uninfected controls to each neonate exposed to infection as babies with missing outcome data have been removed from analysis; see Supplementary Table 11 for further details on levels of missing data.

Cohort
Supplementary Table 6: Association between exposure to SARS-CoV-2 during pregnancy at ≥20 weeks gestation and neonatal and maternal outcomes (reinfections spanning both gestational periods excluded from subgroup analysis), calculated using conditional logistic regression.Lanarkshire versus all other health boards) for hypertensive disorders of pregnancy and pregnancy-related bleeding.**Adjustment for deprivation only for analyses with the following outcomes: very preterm birth, spontaneous very preterm birth and providerinitiated very preterm birth.Adjustment for parity and deprivation only for analyses with the following outcomes: stillbirth, extended perinatal death and venous thromboembolism.Adjustment for all covariates for analyses apart from parity for preterm birth, spontaneous preterm birth and provider-initiated preterm birth.Adjustment for all covariates for analyses for the following outcomes: small for gestational age, very small for gestational age, low Apgar score maternal critical care and/or death, hypertensive disorders of pregnancy and pregnancy-related bleeding.***For these outcomes there is not an exact match of three uninfected controls to each neonate exposed to infection as neonates with missing outcome data have been removed from analysis; see Supplementary Table 11 for further details on levels of missing data.7: Association between SARS-CoV-2 infection and hypertension/bleeding outcomes, stratified by health board of residence, calculated using conditional logistic regression.*Multiple exclusion criteria may have applied for a given pregnancy, but each was only counted in the first exclusion criteria in the order of appearance in this box.**Each cohort was drawn independently, so the controls will vary between the different cohorts.***After cohort matching, we removed any of the infected group with missing data for a specific outcome and their controls, as well as any additional controls with missing data.****Data only missing for delivery onset, so affecting only analyses looking at spontaneous or provider-initiated (very) preterm birth.No missing data for any (very) preterm birth.*****Exposed pregnancies could only be matched to uninfected control pregnancies that had not had the outcome prior to the gestational week at matching.*Multiple exclusion criteria may have applied for a given pregnancy, but each was only counted in the first exclusion criteria in the order of appearance in this box.**Each cohort was drawn independently, so the controls vary between the different cohorts.***After cohort matching, we removed any of the vaccinated group with missing data for a specific outcome and their controls, as well as any additional controls with missing data.****Data only missing for delivery onset, so affecting only analyses looking at spontaneous or provider-initiated (very) preterm birth.No missing data for any (very) preterm birth.*****Exposed pregnancies could only be matched to unvaccinated control pregnancies that had not had the outcome prior to the gestational week at matching.

Table 2 :
Lag in days between infection (or matching) and end of pregnancy for pregnancies ending in preterm (<37+0 gestation) and very preterm births (<32+0 gestation).

Table 3 :
Sub-types of provider-initiated preterm births for Cohort 5.
*Matched for maternal age, gestation at infection/matching and seasons of conception and additionally for health board of residence (Greater Glasgow & Clyde or Lanarkshire versus all other health boards) for hypertensive disorders of pregnancy and pregnancy-related bleeding.
Matched for maternal age, gestation at infection/matching and seasons of conception and additionally for health board (Glasgow or Lanarkshire versus all other health boards) for hypertensive disorders of pregnancy and pregnancy-related bleeding.
Matched for maternal age, gestation at infection/matching and seasons of conception and additionally for health board (Glasgow or

Table 8 :
Matched for maternal age, gestation at infection/matching and seasons of conception.**Adjustment for all covariates for both maternal outcomes.Association between vaccination and hypertension/bleeding outcomes stratified by health board, calculated using conditional logistic regression.Matched for maternal age, gestation at infection/matching and seasons of conception.**Adjustment for all covariates for both maternal outcomes.Supplementary Table10: ICD-10 codes and critical care admission codes (drawn from the Scottish Intensive Care Society Audit Group [SICSAG] dataset) used to identify hypertensive disorders of pregnancy, pregnancy-related bleeding, and venous thromboembolism.

Table 11 :
Missing outcome data in matched cohorts.