2005 HSR&D National Meeting Abstract
1062 — Access, Utilization, and Provider Selection Patterns of US Veterans
Cowper DC (VA HSR&D (RORC))
Use a national survey of the US veteran population (1) to describe veterans who were non-users, VA-only users, Non-VA only users, and Multiple System users; and (2) to model veteran selection of medical care provider.
This study employs the Andersen Behavioral Model to investigate veterans’ use of any health care during a one year period and to explore whether provider selection can be predicted based on veterans’ predisposing, enabling, and need characteristics. The data source was the National Survey of Veterans 2001. The random digit dialing sample totaled 12,905 individuals.
The dependent variable is user type. Independent variables include: age, race, marital status, education, and military experience (predisposing); income, presence of health insurance, type of health insurance, presence of Service-Connected disability, and meets VA “means test” threshold (enabling); and perceived health status, number of health problems, and ADLs (need). Descriptive statistics on all variables were obtained. Sequential conditional logistic regression models were used to distinguish between user types.
Results show 90.5% of veterans surveyed reported receiving some type of medical care in the previous year. Most veterans use non-VA providers only (76.4%); 21.5% used at least some VA use, only 6% were exclusive VA users. Comparing veterans who used some VA to non-VA only users, veterans with lower self-reported health status (OR=1.418), higher numbers of chronic health conditions (OR=1.154), and Priority 1-6 eligibility (OR=2.833) have higher odds of VA use. Coverage by Medicare (OR=.894), Medicaid (OR=.530) or private insurance (.310) reduce the odds of VA use.
The percentage of veterans using at least some VA has increased since the last national survey in 1992. There appears to be considerable dual utilization between the VA and non-VA health care sectors, especially for veterans with Medicare coverage, indicating that these programs should begin to coordinate their benefit packages.
Studies that more precisely delineate the individuals who will use VA services, the reasons for their using (or not using) VA care, and the scope and intensity of the services they will use can help to refine current practice in the estimation of VA health care utilization.