Contributors to disparities in postpartum readmission rates between safety‐net and non‐safety‐net hospitals: A decomposition analysis

Background Safety‐net hospitals (SNHs) treat more maternal patients with risk factors for postpartum readmission. Objective To assess how patient, hospital, and community characteristics explain the SNH/non‐SNH disparity in postpartum readmission rates. Design A linear probability model assessed covariates associated with postpartum readmissions. Oaxaca–Blinder decomposition estimates quantified the contribution of covariates to the SNH/non‐SNH disparity in postpartum readmission rates. Setting Healthcare Cost and Utilization Project 2016–2018 State Inpatient Databases from 25 states. Participants 3.5 million maternal delivery stays. Measurements The outcome was inpatient readmission within 42 days of delivery. SNHs had a share of Medicaid/uninsured stays in the top quartile. A range of patient, hospital, and community characteristics was considered as covariates. Results The unadjusted postpartum readmission rate was 4.2 per 1000 index deliveries higher at SNHs than at non‐SNHs (19.1 vs. 14.9, p