Validity of the Beck Depression Inventory among women with HIV engaged in care: Implications for transdiagnostic psychiatry in HIV treatment
MetadataShow full item record
Major depressive disorder is highly prevalent among women with HIV and is associated with decreased antiretroviral therapy adherence and increased risk of mortality. As major depressive disorder often goes undiagnosed among women with HIV, one proposed solution to this problem has been to establish routine depression screening and integrated psychiatric care as part of HIV treatment. However, the high prevalence of other psychiatric and substance use disorders among women with HIV complicates the accurate diagnosis of major depressive disorder, as depressive symptoms are common across these distinct disorders with varying treatment recommendations. While previous studies have validated screening tools, such as the Beck Depression Inventory, for major depressive disorder among women with HIV, none have addressed the extent to which the presence of these other conditions results in psychiatric misdiagnosis. This thesis addresses this gap by examining the impact of psychiatric disorders and substance use behaviors on the validity of the Beck Depression Inventory in screening for major depressive disorder among women with HIV. Women in our sample were engaged in care at the Johns Hopkins HIV Clinic and were recruited between April 2006 and July 2010 to join a trial of a brief intervention for hazardous alcohol use. Using baseline trial data, we compare women’s Beck Depression Inventory scores to their major depressive disorder diagnosis obtained from a structured clinical interview and calculate sensitivity and specificity in the full study population and stratified by post-traumatic stress disorder diagnosis, bipolar affective disorder diagnosis, and alcohol use risk classification. Beck Depression Inventory sensitivity and specificity were 65% and 75% overall. Sensitivity was higher among women with current (79%) or lifetime (74%) PTSD compared to those without a current (55%) or lifetime (53%) diagnosis. Sensitivity did not vary by alcohol use risk classification. Specificity was higher among women without current (78%) or lifetime (81%) PTSD, without BPAD (78%), or with low-risk alcohol use (80%), and lower among women with these diagnoses (59%, 66%, 62%) or with hazardous alcohol use (68%). We discuss the implications of these results for transdiagnostic psychiatry in HIV care and highlight the need for evaluating major depressive disorder in the context of other psychiatric and behavioral factors prevalent among women with HIV.