Are preventable hospitalizations sensitive to changes in access to primary care? The case of the Oregon Health Plan

Objective: To examine whether preventable hospitalization (PH) rates are sensitive to the impact of policies aimed at improving access, such as the Oregon Health Plan (OHP), which expanded Medicaid coverage to all adults with incomes under the federal poverty level. Study Design: We conducted a retrospective, time series analysis of PH rates in Oregon from 1990 to 2000. We calculated hospitalization rates for ambulatory-care sensitive conditions for the Medicaid + uninsured population and compared average annual rates from 1990 to 1993 (pre-OHP implementation) to those from 1995 to 2000 (post-OHP implementation). We compared changes in PH rates over time in the Medicaid + uninsured group to changes in the nonMedicaid insured population. We standardized rates by age and sex and used logistic regression models to compare rates. Results: Contrary to our hypothesis, annual PH rates in the Medicaid + uninsured population increased after the eligibility expansion, from an average of 46.1 to 54.9 per 10,000 persons. This rise was significant compared with the non-Medicaid insured population, who experienced a slight decline in annual PH rates, from 26.9 to 26.1 per 10,000 (P < 0.001, after adjusting for age, sex, and rates of unpreventable hospitalizations). The increase in overall PH rates for the Medicaid + uninsured population can be explained by an increase in PH rates for the newly insured group. Conclusions: Our results suggest that PH rates may vary not only with access to primary care (inversely) but also with access to hospital care (directly). The use of PH rates as a marker of health care access should take into account these dual influences. Limita tions in available data may also contribute to perceived variation in PH rates unrelated to health care access.