OBJECTIVES: This analysis utilized two administrative claims databases to examine reported associations between COVID-19 infection and adverse delivery outcomes.

METHODS: Women aged 15-50 with a delivery in 2020 were identified in the MarketScan Commercial or Multi-state Medicaid Databases. The index date was the last delivery claim in 2020 and women were continuously enrolled for the prior 280 days (pregnancy period). Diagnoses of COVID-19 and other conditions associated with negative delivery outcomes (diabetes, hypertension, and hemorrhage) were assessed over the pregnancy period among subsets of women with evidence of pre-term, early/threatened labor, or suspected maternal demise (defined as continuous eligibility for <8 or <43 days post-delivery).

RESULTS: Women with pre-term vs. full term birth – commercial 3.0% vs. 2.3% (p<0.0001) and Medicaid 2.8% vs. 2.4% (p=0.0017) – and early/threated labor vs. not – commercial 2.8% vs. 2.3% (p<0.0001) and Medicaid 2.7% vs. 2.3% (p=0.0007) – were significantly more likely to have a COVID-19 diagnosis. Women with pre-term birth and early/threatened labor also had significantly increased comorbidity burden during pregnancy. There were no differences in the rate of COVID-19 or comorbidity diagnoses between women with and without suspected maternal demise the commercial database. Conversely rates of COVID-19 infection were significantly higher among women with suspected maternal demise compared to those without in the Medicaid database at both the 8 day (9.0% of deaths had a COVID-19 diagnosis compared to 2.3% of live women (p<0.0001)) and 43 day cutoffs (4.6% and 1.8% respectively (p<0.0001)).

CONCLUSIONS: This claims-based analysis parallels prior reports indicating correlation between adverse delivery outcomes and COVID-19 infection. Findings for pre-term birth and early/threated labor subgroups are confounded by comorbidity burden. Conversely, comorbidity burden is not driving the correlation between COVID-19 infection and death in the Medicaid data; due to the seriousness of the outcome, additional research that includes mortality data is certainly warranted.