Population-level changes in perinatal death for pregnancies prior to and during the COVID-19 pandemic: A pregnancy cohort analysis.

Abstract

Results of population-level studies examining the effect of the COVID-19 pandemic on the risks of perinatal death have varied considerably. To explore trends in the risk of perinatal death among pregnancies beginning prior to and during the pandemic using a pregnancy cohort approach. This secondary analysis included data from singleton pregnancies ≥20 weeks' gestation in Alberta, Canada, beginning between 5 March 2017 and 4 March 2021. Perinatal death (i.e. stillbirth or neonatal death) was the primary outcome considered. The risk of this outcome was calculated for pregnancies with varying gestational overlap with the pandemic (i.e. none, 0-20 weeks, entire pregnancy). Interrupted time series analysis was used to further determine temporal trends in the outcome by time period of interest. There were 190,853 pregnancies during the analysis period. Overall, the risk of perinatal death decreased with increasing levels of pandemic exposure; this outcome was experienced in 1.0% (95% confidence interval [CI] 0.9, 1.0), 0.9% (95% CI 0.8, 1.1) and 0.8% (95% CI 0.7, 0.9) of pregnancies with no overlap, partial overlap and complete pandemic overlap respectively. Pregnancies beginning during the pandemic that had high antepartum risk scores less frequently led to perinatal death compared to those beginning prior; 3.3% (95% CI 2.7, 3.9) versus 5.7% (95% CI 5.0, 6.5) respectively. Interrupted time-series analysis revealed a decreasing temporal trend in perinatal death for pregnancies beginning ≤40 weeks prior to the start of the COVID-19 pandemic (i.e. with pandemic exposure), with no trend for pregnancies beginning >40 weeks pre-pandemic (i.e. no pandemic exposure). We observed a decrease in perinatal death for pregnancies overlapping with the COVID-19 pandemic in Alberta, particularly among those at high risk of these outcomes. Specific pandemic control measures and government response programmes in our setting may have contributed to this finding.