Erosion in the Safety Net 1990-2003: Findings for Different Population Groups

Significance: Closure of safety net hospitals (SNHs) impedes access to care for vulnerable populations.

Objective(s): Examine the impact of contraction in SNHs 1990-2000 on access to care by vulnerable populations, focusing on distance to events and remaining facilities and disparities across race or ethnicity and insurance type.

Methods: We employ logistic regression to examine rates of ambulatory-care sensitive (preventable) admissions relative to marker (emergency) admissions as indicators of impeded access, using HCUP discharge data from two states (CA, FL) in two cross sections (1990 - 2003). We include individual characteristics, hospital characteristics, and area characteristics and control for distance to a SNH contraction event and the closest remaining SNH. We calculate the conditional marginal probabilities of an ACSC versus marker admission by race or ethnicity, insurance type, and distance to a contraction event. Higher probabilities are interpreted as more impeded access to primary care services.

Discussion: We find that distance to a safety net hospital contraction event, race or ethnicity, and insurance type/status are all inter-related in impacting the probabilities of ACSC versus marker condition admissions. For uninsured people of all races and ethnicities who lived closer to safety net contraction events, the increasingly impeded access over time is greater than for those more distant from these events. Insured people were not impacted much by the SNH contraction. These effects were found for uninsured persons of all races and ethnicities. However, effects were larger for uninsured African Americans than for whites and Hispanics, showing a persistence of disparities in access.