Return to 2001 Abstacts List

JF McCarthy, SMITREC; M Valenstein, HSR&D COE Ann Arbor/SMITREC; FC Blow, SMITREC; J Zeber, SMITREC; L Gillon, SMITREC; R Bingham, SMITREC

Objectives: Health care delivery systems can provide effective treatment only if they are accessible to consumers. Patients with psychoses may be particularly sensitive to accessibility limitations. In prior studies, schizophrenic patients report on average that 3.6 miles is a reasonable travel distance, whereas bipolar patients report 8.5 miles as reasonable.

The impact of accessibility barriers is expected to vary depending on type of psychosis and other patient characteristics. We present a national VA study of the influence of accessibility, measured in miles to the nearest VA service provider, on outpatient care utilization among patients with schizophrenia, bipolar disorder, and other psychoses. We also examine the interrelationships between distance from service, psychosis type, and utilization.

Methods: Using inpatient and outpatient diagnosis data from the Austin Automation Center, we identified all VA patients with psychosis diagnoses in FY99 (N=191,606). We categorized each patient as having primarily schizophrenia, bipolar disorder, or other psychoses. For each patient we calculated the total outpatient visit days and, using data from the VA Planning Systems Support Group, we assessed the straight-line distance to the nearest VA service provider. We created three age groupings: under 45, 45 to 65, and greater than 65. We excluded 11,949 patients (6%) who had some missing data. We thus analyzed data for 179,657 patients (98,161 having schizophrenia, 56,003 bipolar disorders, and 25,493 other psychoses). Controlling for age, minority status, and marital status, we use least squares regression to assess the impact of distance on outpatient utilization. Type of psychosis and interactions with distance were also included in the model.

Results: Median distance to nearest VA provider was 5.9 miles among patients with schizophrenia, 6.9 miles among patients with bipolar disorders, and 7.3 miles among patient with other psychoses. Distance to nearest service provider was associated with significant decreases in annual outpatient utilization (p<0.0001). Controlling for the other predictors in our model, patients had 3.5 fewer visit days per year for every ten miles separating them from care. Patients with schizophrenia had more visit days per year (p<0.0001) than patients with bipolar disorders. However, interactions between distance and diagnosis indicated that patients with schizophrenia were more likely to decrease utilization with increased distance than were patients with bipolar disorder.

Conclusions: Distance to service provider presents real obstacles to care among the special patient population of veterans with psychoses. The impact is greatest among patients with schizophrenia.

Impact: This work presents national data on the impact of distance on outpatient utilization for a special vulnerable population, veterans with psychoses. Despite the rapid expansion in access points through expansion of CBOCs, distance to care for these patients remains a potential barrier to use. This study demonstrates the critical importance of assuring accessibility for psychotic patients, and particularly those with schizophrenia.