Inferred Attractiveness Gravity-Based Models for Estimating Realized Access at Rural Hospitals.

Area of Science:

• Health Services Research
• Rural Health
• Healthcare Equity

Background:

• Operating rural obstetric units is financially challenging, often linked to low birth volumes.
• Birther bypass of local hospitals for farther facilities exacerbates financial strain and impacts healthcare equity.
• Understanding subgroup utilization is crucial for equitable resource allocation in obstetric care.

Purpose of the Study:

• To estimate realized access to obstetric care in rural Montana using an inferred attractiveness gravity-based model (GBM).
• To assess how hospital attractiveness varies by obstetric care level and birther subgroups.
• To evaluate the performance of different GBM variations, particularly for low-volume hospitals.

Main Methods:

• Employed a nonlinear programming optimization model, specifically an inferred attractiveness gravity-based model (GBM).
• Compared three variations of the GBM against a regression-based conditional logit model.
• Utilized bootstrapping and subgroup analyses to assess model accuracy and population heterogeneity.

Main Results:

• Hospital attractiveness for obstetric care differs based on the level of care and the specific birther subgroup considered.
• All GBM variations showed improved accuracy for higher birth volume hospitals, but a novel GBM variant excelled for low-volume facilities.
• Subgroup analyses revealed significant variations in hospital attractiveness, highlighting the importance of population heterogeneity.

Conclusions:

• Findings enhance understanding of equity in rural obstetric care access.
• Emphasizes the necessity of incorporating population heterogeneity into GBMs for accurate healthcare access modeling.
• Demonstrates the value of hospital demand-based thresholds within GBMs for rural healthcare planning.