Study Population
We performed a cross-sectional analysis of all delivery-related hospital admissions using the Premier Healthcare Database (PHD) (Premier Applied Sciences, 2020). The PHD is a comprehensive, service-based, all-payor database that compiles discharge data from over 1419 non-profit, community-based, teaching hospital and hospital systems from geographically varied areas including both urban and rural areas. PHD represents approximately 20% of inpatient discharges in the United States. For this study, we included data from 912 hospitals with delivery hospitalizations during February 2020–August 2023. The methods build on previously published reports using the same data source (DeSisto et al., 2021; Ko et al., 2021; Simeone et al., 2022).
Diagnostic and procedure codes from the International Classification of Diseases, 10th revision (ICD-10) were used to identify obstetric delivery resulting in a singleton livebirth or stillbirth and diagnosis-related codes among women 12–55 years of age (Supplement, Appendix 1). COVID-19 during a delivery hospitalization was determined using ICD-10 codes U07.1 (COVID-19, virus identified) or B97.29 (other coronavirus as the cause of disease classified elsewhere). Adverse outcomes and pregnancy outcomes were identified using ICD-10 diagnostic and procedure codes (Supplement, Appendix 2 and 3). The following adverse outcomes were assessed: renal failure, cardiac event/outcome [including acute myocardial infarction, cardiomyopathy, heart failure, cardiac arrest, cardioversion, atrial fibrillation, ventricular tachycardia, ischemia, pulmonary edema], thromboembolic disease [including deep vein thrombosis and other thromboembolic disease], ARDS, shock, sepsis, ICU admission, mechanical ventilation, cesarean delivery, and in-hospital death. Any severe outcome included any of the above adverse outcomes. Mechanical ventilation was identified through a combination of ICD-10 procedure codes and hospital chargemaster records. Pregnancy outcomes assessed included preterm delivery and stillbirth. Data were restricted to females aged 12–55 years that resulted in a singleton live birth or stillbirth (Supplement, Appendix 1). Among individuals with more than one delivery hospitalization during the study period, one admission was selected at random to be included in the analysis.
Medical conditions were identified if an ICD-10 diagnosis or procedure code were present at the delivery hospitalization (Supplement, Appendix 4). Medical conditions included obesity, any diabetes (including type 1 and type 2, gestational diabetes, and unknown diabetes), pre-pregnancy diabetes, gestational diabetes, asthma, other chronic lung diseases (including chronic obstructive pulmonary disease, chronic respiratory failure, pulmonary fibrosis, cystic fibrosis, chronic bronchitis, obstructive sleep apnea, interstitial lung disease, and sarcoidosis), hypertensive disorders of pregnancy (including chronic hypertension, gestational hypertension, chronic hypertension with superimposed pre-eclampsia, pre-eclampsia, eclampsia, hemolysis, elevated liver enzymes, and low platelet count [HELLP] syndrome), chronic hypertension, and gestational hypertension.
Demographic variables included maternal age, race and ethnicity (Hispanic, non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, and non-Hispanic other), primary insurance payor (Medicaid, private insurance, other), hospital location/urbanicity (urban and rural based on U.S. Census where block groups have a population density of at least 1000 people per square mile and surrounding block groups a density of 500 people per sq/mi) (2020 Census Qualifying Urban Areas and Final Criteria Clarifications, 2022), and hospital region (based on U.S. census: Midwest, Northeast, South, West).* This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy; the activity was determined to meet the requirements of public health surveillance as defined in 45 CFR 46.102(l)(2).
* Hospital data were obtained from the PATDEMO table in PHD, which contains both patient demographic information and provider information.
Statistical analysis
Demographic, medical conditions, adverse outcome, and pregnancy outcome characteristics by COVID-19 variant period were summarized using frequencies. COVID-19 case counts were calculated using delivery hospital admission dates. Poisson regression models with robust standard errors were used to calculate the adjusted prevalence ratio (aPR) of adverse outcomes and pregnancy outcomes at the delivery hospitalization by COVID-19 diagnosis for each time period of variant predominance. For consistency with previous analyses using these data, regression models controlled for maternal age at delivery, obesity, diabetes, asthma, and chronic hypertension. Data were stratified by timing of variant predominance based on SARS-CoV-2 genomic surveillance (Lambrou, 2022). When the percent of infections by a specific variant constituted more than 50% of sequenced isolates, it was determined to be the predominant variant (Lambrou, 2022): pre-Delta period (before July 2021), Delta (July 2021–December 2021), and Omicron (January 2022–August 2023) periods. Variant periods were defined using whole months because of data aggregation in PHD. All analyses were performed using R Statistical Software (v4.1.2; R Core Team 2021).