HIV and depression in a primary care clinic in Johannesburg, South Africa Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 19, 2019
  • Cholera, Rushina
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • As rapidly expanding HIV care and treatment programs are implemented in sub-Saharan Africa, maintaining engagement in HIV care is proving to be a significant operational challenge. Depression is highly prevalent among HIV-infected people in sub-Saharan Africa and predicts a range of poor HIV-related clinical outcomes including faster disease progression and increased morbidity and mortality. Depression is strongly associated with non-adherence to anti-retroviral treatment (ART), and may also impact engagement in HIV care. Recognizing and treating depression among HIV-infected patients seeking care in primary health care settings, where most HIV testing and treatment takes place, could increase access to mental health services and may help to target patients at risk for negative outcomes. An observational study was conducted between September 2012 and April 2013 among 1683 randomly selected adult patients undergoing routine, opt-out HIV counseling and testing (HCT) at Witkoppen Health and Welfare Center (WHWC), a high-HIV burden primary care clinic in Johannesburg, South Africa. Patients were screened for depression immediately prior to HCT using the Patient Health Questionnaire-9 (PHQ-9), a 9-item brief screening tool administered by lay health workers. A subset of 400 patients was included in a blinded diagnostic validation study of the PHQ-9 (Aim 1). Sensitivity and specificity of the PHQ-9 were calculated with the Mini International Neuropsychiatric Interview (MINI) as the reference standard, and receiver operating characteristic (ROC) curve analyses were performed. Patients who tested positive for HIV were followed and linkage to care, defined as returning to WHWC within 3 months to obtain a CD4 count result, was assessed (Aim 2). Among patients who collected a CD4 count, ART initiation within 3 months was assessed. Multivariable Poisson regression with a robust variance estimator was used to assess the association between depression and linkage to care or ART initiation. Nearly all patients completed depression screening and 82% (n=1386) subsequently tested for HIV. Of the patients who tested for HIV and were of unknown HIV status prior to testing, 26% (n=340) were found to be HIV-infected. Nearly a quarter of all patients (22%) were depressed. Similar to other studies, depression was more common among patients who tested positive for HIV compared to those who tested negative for HIV (30.3% versus 19.7%, p<0.0001). In the validation sample included in Aim 1, the prevalence of depression was 11.8% and one-third of participants tested positive for HIV. Using the standard cut-off score of &#8805;10, the PHQ-9 demonstrated a sensitivity of 78.7% (95% CI: 64.3-89.3) and specificity of 83.4% (95% CI: 79.1-87.2) for identifying major depressive disorder. The area under the ROC curve was 0.88 (95% CI: 0.83-0.92). Test characteristics did not vary by HIV status or language and in sensitivity analyses, reference test bias associated with the MINI appeared unlikely. The PHQ-9 was easily implemented by lay health workers. The instrument performed reasonably well and may be a useful depression screening tool in high HIV-burden sub-Saharan African primary health care settings. Of the HIV-infected patients included in Aim 2, 30% were depressed. The proportion who linked to care was 80% among depressed patients and 73% among patients who were not depressed (Risk ratio: 1.08; 95% confidence interval: 0.96, 1.23). Of the participants who linked to care, 81% initiated ART within 3 months in both depressed and not depressed groups (Risk ratio: 0.99; 95% confidence interval: 0.86, 1.15). Conducting this study in a clinic-based population may have selected for patients who are high-utilizers of health care and unlikely to be at high risk for loss to HIV care. These unexpected results highlight the importance of population selection and the timing of HIV testing relative to depression screening when studying the complex relationship between depression and engagement in HIV care.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Pettifor, Audrey
  • Miller, William
  • Gaynes, Bradley
  • Pence, Brian
  • MacPhail, Catherine
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2014
Place of publication
  • Chapel Hill, NC
  • There are no restrictions to this item.

This work has no parents.