Risk Factors for Loss of Hospital-Based Obstetric Care in Rural and Urban Hospitals.

IMPORTANCE

Obstetric care access is declining in the US, with rural and urban hospitals closing obstetric units or closing entirely.

OBJECTIVE

To assess hospital-level factors associated with obstetric services loss by unit or hospital closure and between rural and urban hospitals.

DESIGN AND

SETTING

Retrospective cohort study of all US rural and urban hospitals that provided obstetric services in 2010 as identified using 2010 to 2023 data from American Hospital Association Annual Surveys, the Centers for Medicare & Medicaid Services' Provider of Services Files, and primary data from hospital websites and news articles. Data were analyzed November 2025 to April 2026.

EXPOSURES: Distance to nearest obstetric hospital, birth volume, average daily census, nurse staffing, ownership, Critical Access Hospital status, unprofitability, county unemployment in 2010.

MAIN OUTCOMES AND MEASURES: Outcomes using adjusted multinomial logistic regression included whether each hospital maintained obstetric services from 2010 to 2023, remained open but closed the obstetric unit by 2023, or closed entirely by 2023; and hospital characteristics in 2023. Outcomes are reported as estimated risk-adjusted predicted probabilities and marginal effects as percentage-point (pp) differences.

RESULTS

Among 1126 rural obstetric hospitals open in 2010, 239 (21.2%) closed their unit and 42 (3.7%) closed entirely by 2023; among urban hospitals (n = 1982), 250 (12.6%) units and 92 (4.6%) hospitals closed. Rural and urban hospitals with lower birth volumes had from 8-pp (95% CI, 5.4-10.6 pp) higher predicted probabilities of unit closure for the second highest birth volume quartile to 31-pp (95% CI, 23.2-39.1 pp) higher predicted probabilities for the lowest birth volume quartile and 4-pp (95% CI, 0.7-7.1 pp) higher predicted probabilities of hospital closure for the second lowest birth volume quartile to 8 pp (95% CI, 3.3-12.4 pp) for the lowest birth volume quartile; birth volume had a larger marginal effect in unit than hospital closure. Unprofitable rural hospitals had a 12-pp (95% CI, 6.8-17.5 pp) higher predicted probability of unit closure than profitable ones; for urban hospitals, predicted probabilities were 4 pp (95% CI, 0.1-7.1 pp) higher for unit and 7 pp (95% CI, 3.9-9.7 pp) higher for hospital closure. Rural and urban hospitals near another obstetric hospital had greater than 7-pp (95% CI, 1.6-12.7 pp) higher predicted probabilities of unit closure than those farther away. In rural areas, for-profit hospitals had 13-pp (95% CI, 2.8-22.2 pp) higher predicted probabilities of unit closure and 6-pp (95% CI, 1.6-10.3 pp) higher predicted probabilities for hospital closure vs government-owned hospitals. In urban areas, for-profit status was only associated with hospital closure (4 pp; 95% CI, 1.2-7.6 pp). Among hospitals that maintained obstetric services through 2023, 41% had high-risk factors in 2023 for future obstetric loss.

CONCLUSIONS AND RELEVANCE: In this study, lower birth volume, unprofitability, proximity to another obstetric hospital, and for-profit status were factors associated with obstetric loss. The importance of specific risk factors varied by type of loss (unit or hospital closure) and between rural and urban hospitals. Many hospitals with obstetric services in 2023 may be at high risk for future closure.