Feasibility Versus Vulnerability in the Geographic Distribution of Obstetric Services

Policy Points Service gaps align more closely with feasibility-related inputs (birth volume and workforce capacity) than with vulnerability alone. As a result, vulnerability-targeted programs may have a limited impact on service distribution unless observed operational conditions also change (payment/readiness support, staffing depth, liability/fixed costs). Equity strategies differ: modify operational conditions where local provision is desired or formalize regional access (transport/referral networks and non-hospital supports) where local provision is not observed.

Geographic gaps in hospital-based obstetric services are often framed as "maternity care deserts," emphasizing community vulnerability and motivating targeted investments. This study was conducted to test whether service geography is more closely aligned with an operational feasibility regime: the birth volume and workforce capacity associated with the observed presence of obstetric services.

A national county-year panel (2015-2023) was constructed and used to compare vulnerability-oriented versus feasibility-oriented explanations for the presence of obstetric services. Feasibility was measured using prior-year birth volume and physician workforce supply, as well as their interaction (ln[births] × workforce). Logistic regression models were estimated with county-clustered standard errors, and non-nested specifications were compared using the Akaike information criterion (AIC) and the Bayesian information criterion. Analyses were also stratified by metropolitan status.

Feasibility-based specifications were found to substantially outperform vulnerability-only models (ΔAIC > 10,000). A hybrid specification including feasibility inputs and maternal vulnerability achieved the strongest overall fit, indicating that vulnerability contributes explanatory information conditional on observed feasibility conditions. In nonmetropolitan counties, an increase in the birth volume from the 25th to the 75th percentile was associated with a 41 percentage point increase in the predicted presence of services, compared with 20 percentage points for a corresponding change in the workforce.

Obstetric service distribution was found to be most consistent with a hybrid structure in which feasibility-related inputs aligned more strongly with the presence of services and vulnerability contributed additional explanatory information conditional on those inputs. These findings suggest that vulnerability-targeted approaches may imply different policy approaches than interventions targeting observed operational conditions. Within settings consistent with observed feasibility conditions, or where operational arrangements are modified, vulnerability gradients can guide where stabilization investments are most needed.