External Determinants of Veterans' VA Healthcare Utilization
In 2014 Congress created the VA Choice program, through which the US Department of Veterans Affairs (VA) pays for care provided to veterans by non-VA providers. To manage this program, the VA needs to anticipate and budget for changes in the composition of the use of healthcare services from VA enrollees. In developing projections, we propose that models of VA healthcare utilization should include factors external to the VA because local economic conditions and non-VA insurance coverage may influence VA use, particularly among non-elderly veterans. Using a national cohort, we measured sensitivity of VA healthcare utilization to these external factors.
We identified veterans enrolled for VA healthcare and examined annual VA healthcare utilization during 2008-2013. Our primary outcome measure was overall utilization ($), combining inpatient care, outpatient care and medication use. We mapped individual-level annual utilization data to area-level annual data on external determinants based on veterans’ residence location. We studied the effects of six potential external determinants: unemployment ratio, poverty rate, % Medicaid eligible, employer-sponsored private insurance offer, non-VA provider availability, and housing price index. Based on longitudinal linear models, we estimated sensitivity of VA healthcare utilization separately for those under and over age 65, adjusting for patient demographics and patient-level fixed effects.
We examined healthcare utilization of 10.2 million veterans, of whom 60% were under 65 years old. During the study period, annual per capita healthcare utilization increased 13.2% from $3,798 (2008) to $4,298 (2013), unemployment ratio increased 15.2%, poverty increased 20%, % Medicaid eligible increased 13.3%, housing price index increased 1.3%, non-VA provider availability increased 2.8%, and private insurance offer did not change. All external factors were associated with VA utilization, with the exception of housing price index among the under 65 and private insurance offer among the over 65. Between 2008 and 2013, changes in external factors were associated with an increase in VA spending of $244 million, accounting for 4.8% of the overall increase in VA spending for the cohort. Unemployment ratio, poverty rate and % Medicaid eligible had the largest effects on spending for both age groups.
VA healthcare utilization is affected by policy and economic factors beyond the control of VA policymakers and managers. Failure to account for these factors in past budget projections may have contributed to imbalances in supply and demand. Future VA budget models should account for the effects of these external determinants.