Date of Award


Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Amy Justice


Background: Despite a higher prevalence of depression among HIV-infected veterans, previous research has shown that infectious disease (ID) providers report substantially less comfort with depression treatment than do general medicine (GM) providers. We examined whether HIV-infected veterans who are treated by ID providers are less likely to have their depressive symptoms treated compared to uninfected controls managed by GM providers. Methods: We used survey, service utilization, and pharmacy data on veterans from the Veterans Aging Cohort Study (VACS), a prospective cohort study of HIV-infected and age-, race- and site-matched uninfected subjects at 8 Veterans Affairs Healthcare Centers. We used the Patient Health Questionnaire (PHQ-9) to identify veterans with depressive symptoms. Each of nine survey items was rated by the veteran as being present "0" (not at all) to "3" (nearly every day). Veterans were considered to have active depressive symptoms if they had a PHQ-9 score of 10 or greater, which constituted a positive screen for major depressive disorder. Of the 5998 VACS patients, 19.7% of uninfected and 21.3% of HIV-infected veterans had PHQ-9 scores of 10 or greater. Of these veterans with active depressive symptoms, those receiving mono-amine oxidase inhibitors (MAOIs) (n=3), female veterans, and men with diagnoses of schizophrenia (n=511) or PTSD (n=689), were excluded. A small number of patients receiving tricyclic antidepressants (TCAs) were excluded for criteria other than TCA use. Depression treatment was defined as receipt of a selective serotonin reuptake inhibitor (SSRI) or any VA mental health utilization in the 6 months prior to or after survey. Bivariate comparisons by clinic type were assessed using chi-square and t-tests. Logistic regression was used to determine whether clinic type was associated with receipt of SSRI, adjusting for potential confounding variables such as demographics and clinical factors. Results: Of the 5998 veterans in VACS, 732 met our criteria with PHQ-9 scores greater than 10, male gender, without schizophrenia, PTSD or MAOI use. Of the 732 eligible veterans, 59% were HIV-infected and 41% were uninfected. The sample was predominantly African-American (58%) and had a median age of 48 years. There was no significant difference in the proportion of veterans with depressive symptoms who were treated by HIV status (38% of HIV-infected veterans vs. 34% of uninfected veterans, p=0.4). This remained true even when mental health service utilization was included (48% vs. 49%, p=0.8). Caucasian veterans were significantly more likely than African-Americans to have received SSRI (48% vs. 30%, p<0.01). After controlling for veteran age, race, and comorbid conditions, HIV-infected veterans did not differ significantly in receipt of SSRI (OR=1.16, 95% CI=0.84, 1.58). However, there were significant differences in treatment rates by site and by individual clinic. Conclusions: Despite previous analysis demonstrating substantial differences in provider comfort with depression treatment, both HIV-infected and uninfected veterans were equally unlikely to be treated for depressive symptoms. While treatment rates did not vary by HIV status, they varied significantly by geographic site and individual clinic, suggesting that provider practices have considerable influence over receipt of treatment.

Open Access

This Article is Open Access