This study compares two models for implementing antidepressant treatment in 10 HIV clinics in Uganda. Using a cluster randomization, 5 clinics implement a task-shifting, protocolized model, and 5 others rely on clinical acumen. The protocolized model includes (1) routine depression screening at each clinic visit for all adult patients by trained expert patients at triage, (2) training nurses to diagnose depression and prescribe and monitor antidepressant treatment using an algorithm-based protocol, and (3) monthly supervision and monitoring by hired study psychiatrists. The clinical acumen model also includes routine depression screening and ongoing supervision, but it relies on the clinical acumen of trained primary care providers to further evaluate and treat patients who show signs of depression at screening, as opposed to a structured protocol. The primary aim is to test the hypothesis that the nurse-driven protocolized model will result in greater uptake of antidepressant treatment and better quality of depression care outcomes. The study will also test the hypotheses that treatment of depression results in improved HIV treatment adherence, work functioning and consistent condom use.

The study evaluates a task-shifting approach to depression treatment that uses an algorithm-based, nurse-driven model for managing antidepressant treatment in 10 Ugandan HIV clinics. The model will include (1) case identification facilitated by routine depression screening at each clinic visit for all patients, (2) training nurses to assist primary care providers in implementing antidepressant treatment by performing the initial evaluation, monitoring symptoms and side effects, and making algorithm-based dose recommendations, and (3) layers of supervision and monitoring by psychiatric specialists to ensure safety and quality of care. This model will be implemented at 5 randomly selected clinics (protocolized arm), while 5 other clinics will be selected to also use routine depression screening but will rely on primary care providers to decide whether to further evaluate and treat depression (clinical acumen arm). At each site, random samples of 150 patients (total n=1500) who have screened positive for possible depression will be followed for 12 months. We will compare the two arms on depression evaluation, uptake of antidepressant treatment, and change in depression (treatment response). We will also examine the relationship between change in depression and key economic and public health outcomes (e.g. work status, condom use, HIV treatment adherence).

Inclusion Criteria:

• age 18 years or above
• screens positive for depression (scores > 2 on PHQ-2)
• medically stable
• client at the HIV clinic (and hence, HIV positive)

Exclusion Criteria:

• active acute opportunistic infection
• about to start HIV antiretroviral therapy or started antiretroviral therapy within past 4 weeks