Utilization of medical services by Veterans Health Study (VHS) respondents

The first objective of this study was to profile Veterans Health Study (VHS) respondents' use of medical services-the types of services used, use of a regular source of care, and the propensity to use services for selected symptoms. We focused on differential use of VA and non-VA services and highlighted differences in use by age group. The second objective was to use multivariate analysis to identify factors associated with respondents' use of any medical services and with VA services specifically. We incorporated 2 self-reported variables not used in previous studies of VA utilization-health status and disease burden. Patients receiving ambulatory care services in 4 VA ambulatory outpatient clinics in the greater Boston area were eligible for inclusion in the VHS. A sample of 2425 community-dwelling male veterans was randomly selected from among veterans receiving ambulatory services at Boston-area VA facilities. This analysis focuses on 1909 respondents for whom we had complete data. Interviews and questionnaires were used to collect cross-sectional, observational data on sociodemographic, economic, and clinical characteristics; health status; disease burden; and service-connected disability (SCD) rating. To measure health status, we used 2 summary measures, the Physical Component Summary (PCS) and the Mental Component Summary (MCS), derived from the 8 scales of the Medical Outcomes Study Short Form 36-item Health Survey (MOS SF-36). To measure disease burden, we used the Physical Comorbidity Index (PHYCI) and Mental Comorbidity Index (MENCI), composed of 30 physical and 6 mental health conditions and symptoms, respectively. Information on the availability of non-VA insurance was obtained from administrative VA files. Information on utilization prior to the interview was self-reported. Recall periods of 3 and 12 months were used for ambulatory and inpatient services, respectively. We used descriptive statistics to profile respondents and their utilization patterns. We used multivariate probit models to identify respondent characteristics associated with use of any medical services, medical visits, mental health visits, and hospital stays. Independent variables used in the models were socioeconomic and demographic characteristics, and measures of disease burden, health status, and VA eligibility. The respondents relied heavily on the VA for medical care: 74% of the respondents said the VA was their regular source of care; 72% of all the respondents and 87% of those who had used any medical service in the recall period had used a VA service; 68% of those who were hospitalized used a VA hospital; and 76% of the medical care the respondents received and 60% of their hospital stays were in VA facilities. Younger veterans (aged 22-44) used substantially more mental health services than older respondents, but they were less likely than older veterans to have seen a doctor recently for most of the medical symptoms studied. PHYCI and PCS were significantly related to use of any medical services and to use of inpatient services; MENCI and MCS were significantly related to use of mental health services (P<.05 for each, respectively). Lower income and lack of alternatives to VA care were directly related to use of any VA services and VA inpatient services. Information on the reasons for differential use of VA and non-VA services can be useful to the VA as it serves an aging veteran population, seeks to provide comprehensive care to a wider spectrum of veterans, and moves into a more competitive healthcare marketplace.