Publication Type: Abstract
Abstract: Rational:
Dual use of the Veterans Affairs health care system and non-VA care may affect depression treatment outcomes by fragmenting care across providers and reducing coordination and continuity of care.

Objective:
We examined non-VA utilization for physical and emotional health problems and factors associated with non-VA use among depressed primary care patients.

Methods:
WAVES is a group randomized trial of a depression quality improvement collaborative care intervention implemented in 7 VA outpatient primary care clinics across 3 VISNs during 2002-2003. Patients screened positive for depression using PHQ9 were given a baseline Patient Assessment Questionnaire via telephone addressing non-VA health care in the prior 6 months. We examined non-VA use by type of service and by visit purpose.

Results:
47% of the 768 depressed patients reported non-VA use in the previous six months, with 31% of their outpatient visits being non-VA. Non-VA users had more outpatient visits (6.4 vs 3.8, p< 0.01), ER visits (1.2 vs 0.1, p<0.01), and inpatient admissions (0.9 vs 0.3, p<0.01) than VA-only users. Among non-VA users, 36% of outpatient visits for physical problems were non-VA, compared to 18% of visits for emotional problems. 36% of all depressed patients visited mental health specialists in VA, while only 5% had such visits outside VA. Factors significantly associated with higher odds of any non-VA use were: other health insurance (OR=5.28), dissatisfaction with physical health services in general (OR=2.39), and comorbid conditions of pneumonia (OR=0.73), stroke (OR=1.86), or congestive heart failure (OR=1.89). Factors significantly associated with odds of any non-VA outpatient use for emotional problems were: some college education (OR=10.94-16.80), other insurance (OR=3.09), and comorbid conditions of heart attack (OR=2.24) or PTSD (OR=4.25).

Conclusions:
Almost half of depressed primary care patients used non-VA care. Non-VA users were high health care utilizers compared to VA-only users, although most of their non-VA outpatient use was for physical rather than emotional problems. Non-VA use may complicate chronic disease care for veteran dual users more than mental health care, where other public and private sector coverage is likely to be more limited. Implications for primary care–based interventions, such as depression collaborative care, need to be further evaluated.