2019 HSR&D/QUERI National Conference

1071 — The Impact of the Affordable Care Act on Veterans' Use of VA Outpatient Care

Lead/Presenter: Elizabeth Tarlov,  COIN - Hines
All Authors: Tarlov E (Hines), Gonzalez B (Hines), Smith B (Hines) Silva A (Hines) Huo Z (Hines) Stroupe K (Hines)

In January 2014, the Affordable Care Act (ACA) increased health insurance options for many non-elderly Veterans. The objective of this study was to ascertain changes in use of VA outpatient care associated with ACA implementation. We evaluated changes in utilization in a high-need group, individuals with diabetes and at least one comorbid chronic condition, as well as in the general population of VA patients under age 65.

In a retrospective cohort study of a 10% random sample of male and 100% sample of female Veterans using VA healthcare, we assembled information on VA outpatient visits, state of residence, VA enrollment priority group (PG), and demographic and health covariates in each calendar year quarter, 2012 through 2015. We used mixed effects logit regression models to estimate relationships between utilizing VA services ( > = 1 versus 0 visits) and ACA implementation while accounting for secular trends in utilization and covariates. Analyses were stratified based on gender and PG.

For men, we found no immediate or longer-term changes associated with the ACA among those in the diabetes-with-comorbidity group (N = 27,159) while the full sample (N = 233,255) experienced immediate and longer-term declines in utilization. For example, in PG 5, the odds of having an outpatient visit declined by 15% (OR 0.85, CI95:0.83-0.88) in 2014 Quarter 1 compared to the previous 8 quarters. Subsequently, a pre-ACA upward trend in utilization was reversed (OR 0.95, CI95:0.95-0.96 per quarter) throughout 2014-2015. Results for other PGs were similar. For women in the diabetes-with-comorbidity group (N = 22,092), we found an upward utilization trend (OR 1.03, CI95:1.01-1.04) post-ACA in PG 5 only, reversing a pre-ACA downward trend. Results for all women (N = 352,996) were very similar to those for men.

For Veterans with diabetes and comorbidities, we found little evidence that the ACA impacted use of VA healthcare among pre-ACA VA users. One exception may be low-income women without a service-connected disability. For other VA users, ACA implementation was associated with declines in outpatient care, reversing secular upward trends.

Increasing Veterans' access to non-VA care may induce little change in reliance on VA care among high-need patients. Impacts on resource use and per-capita costs deserve study.