Development and validation of a 20-item screening scale to detect major depressive disorder among adolescents with HIV in rural Uganda: A mixed-methods study

Abstract

Background: Depression is a major cause of disability among children and adolescents and is associated with elevated risks for substance abuse, HIV transmission risk behavior, and suicide. Among adolescents living with HIV (ALWH), depression undermines adherence to antiretroviral treatment, leading to poorer health outcomes. However, there are few instruments available for depression screening among ALWH in sub-Saharan Africa.

Methods: Using mixed methods we developed and validated a 20-item depression screening scale to be used among ALWH in rural Uganda. First, we conducted focus group discussions and in-depth interviews with adolescents and adult caregivers (n = 80) to elicit participant perspectives about mental health challenges facing HIV-affected children and adolescents. We generated an initial pool of 40 items, pilot tested it with ALWH and adolescents of unknown serostatus (n = 40), and then administered the items to a validation sample of ALWH (n = 224). Exploratory factor analysis was used to examine the factor structure of the scale. We evaluated the scale for its reliability, and validity.

Results: The mean age of the participants in the validation sample was 14.9 years (standard deviation [SD] 1.4), 131 (58%) were girls and 48 (21%) were orphans. Exploratory factor analysis revealed two factors related to affective and cognitive symptoms of depression. The 20-item depression scale was internally consistent (Cronbach's alpha = 0.91) with moderate test-retest and inter-rater reliability. Construct validity was excellent, as demonstrated through correlation with related constructs like stigma (P< 0.001) and bullying (P< 0.001). At the optimized cutoff score, 64 (29%) participants screened positive for probable depression. Using the Mini-International Neuropsychiatric Interview for Children and Adolescents, we found that 37 participants (17%) were diagnosed with major depressive disorder. In reference to the criterion standard, the depression scale showed excellent discrimination (c-statistic = 0.84).

Conclusion: This new 20-item depression scale was reliable and valid for detecting major depressive disorder among ALWH in rural Uganda.

Keywords: Adolescent; Depression; HIV; Screening; Uganda.

Fig. 1

Scree. The graph shows the eigenvalues of the identified factors, in declining order of magnitude. As is shown in the figure, the eigenvalue was largest for factor 1, followed by a steep decline for factor 2.

Fig. 2

Receiver operating characteristic curve. The graph shows the diagnostic ability of the depression scale as the discrimination threshold is varied. The point on the graph that maximizes the sum of sensitivity and specificity (Youden's Index) corresponds to a total depression score threshold of ≥ 10.