COVID-19 infection history as a risk factor for early pregnancy loss: results from the electronic health record-based Southeast Texas COVID and Pregnancy Cohort Study

Abstract

Background: The effects of SARS-CoV-2 infection before or during pregnancy on pregnancy outcomes are still largely unknown. We hypothesized that COVID-19 in early pregnancy is a risk factor for adverse pregnancy outcomes, particularly miscarriage.

Methods: We examined the relationship between COVID-19 and adverse pregnancy outcomes, including spontaneous abortion, ectopic pregnancy, and preterm delivery in a large, retrospective, electronic health record (EHR)-based cohort, from 2019 to 2023. Generalized estimating equation modeling was performed to identify risk factors for adverse pregnancy outcomes. Study exposures included COVID-19 before pregnancy, COVID-19 during pregnancy, age, race/ethnicity, comorbidity burden, and neighborhood-level social vulnerability.

Results: In the Southeast Texas Pregnancy and COVID Cohort (26,783 pregnancy episodes), the risk of miscarriage among pregnancy episodes with a miscarriage, livebirth, or delivery outcome was 6.3% (1514/ 24,119). In multivariable modeling, history of both mild and moderate to severe COVID-19 before pregnancy were associated with miscarriage (adjusted odds ratio (aOR) 2.48, confidence interval (CI) 2.21-2.78 and aOR 2.81, CI 1.8-4.38, respectively). Additionally, in the same model, both mild and moderate to severe COVID-19 in the first trimester were associated with miscarriage (aOR 2.31, CI 1.96-2.72 and aOR 2.45, CI 1.12-5.35, respectively).

Conclusions: COVID-19 both prior to and during pregnancy was identified as a risk factor for spontaneous abortion in this study sample. These findings highlight the importance of COVID-19 vaccination and post-COVID management for pregnant people and those planning a pregnancy.

Keywords: COVID-19; Electronic health records; Epidemiology; Infectious disease; Miscarriage; Pregnancy.

Fig. 1

Study flow

Fig. 2

Study population with inclusion criteria

Fig. 3

COVID-19 and pregnancy episode temporal categorization

Fig. 4

COVID-19 infections relative to pregnancy. COVID infection episodes are defined as 30 days from the date of the index COVID-19 report (captured from diagnoses/public health registry report/lab tests). Any COVID report dated within this 30-day period was discarded, and any COVID report dated after this 30-day period became the index date of a new COVID episode. In this figure, the start date of each COVID episode was compared to the start dates of each pregnancy episode

Fig. 5

Pregnancy episodes identified within the Southeast Texas COVID Cohort by date with adverse pregnancy outcomes. CI: Confidence interval. Miscarriage risk calculated as the proportion of miscarriage among Miscarriage models include spontaneous abortions among pregnancy episodes with known outcomes; ectopic pregnancies, abortions of unclear etiology (induced vs spontaneous), and induced abortions are excluded from this analysis (1514 events/ 24,119 episodes). Ectopic risk calculated as the proportion of ectopic pregnancy among all pregnancy episodes (371 events/25,058 episodes). Preterm birth risk calculated as the proportion of deliveries occurring between 20 and 36 weeks gestation among singleton pregnancy episodes of >20 weeks gestation with gestational age at delivery available (2896/17,331)

Fig. 6

Forest plot: Univariable and multivariable generalized estimating equation models for miscarriage. Results of univariable models reported as crude odds ratios [95% confidence intervals] and multivariable models reported as adjusted odds ratios [95% confidence intervals]. Miscarriage models examine spontaneous abortions among pregnancy episodes with known outcomes. Ectopic pregnancies, abortions of unclear etiology (induced vs spontaneous), and induced abortions are excluded from this analysis