Despite carrying the vast majority of the global mental disorder burden, 75% of adults with mental disorders in Low and Middle Income Countries have no access to services. This study will test strategies for integrating first and second line evidence-based depression and trauma-related disorder treatments with primary care services at a large public sector hospital and conduct robust cost and cost-benefit analyses of each treatment to produce a "menu" of cost-benefit options for personalized, integrated mental health care with corresponding effectiveness and implementation values.

Mental disorders are a leading cause of global disability, driven by depression and anxiety. Most of the disease burden is in Low and Middle Income Countries (LMICs), where 75% of adults with mental disorders have no service access. Despite nearly 15 years of efficacy research showing that local non-specialists can provide evidence-based care for depression and anxiety in LMICs, few studies have advanced to the critical next step: identifying strategies for sustainable "real world" non-specialist treatment including integration with existing healthcare platforms and response to common clinical dilemmas, such as what treatment to start with and how to modify it.

Given the need to personalize treatment to achieve remission (absence of disease) and the scarcity of mental health specialists in LMICs, successful reduction of population-level disability caused by depression and anxiety requires (1) evidence-based strategies for first-line and second-line (non-remitter) treatment delivered by non-specialists, with (2) confirmation of presumed mechanism of action and (3) patient-level moderators of treatment outcome to inform personalized, non-specialist treatment algorithms.

The research team has worked in western Kenya for 6 years with a UCSF-Kenya collaboration that supports integrated HIV services at over 70 primary healthcare facilities in Kisumu County (Family AIDS Care and Education Services [FACES]). Primary care populations in Kenya have high prevalence of Major Depressive Disorder (MDD) (26%) and Posttraumatic Stress Disorder (PTSD) (35%). Kenyan leaders lack an evidence base for two essential treatments - psychotherapy and second generation antidepressants- without which scale-up will fall short of its potential. We conducted a randomized, controlled trial in Kisumu County of Interpersonal Psychotherapy (IPT) delivered by non-specialists for HIV-positive patients with MDD and PTSD. In our study, IPT achieved full remission of MDD and PTSD in the majority of participants.

Given the high prevalence of MDD-PTSD co-morbidity, we will collaborate with the FACES team providing services to Kisumu County Hospital (KCH) primary care outpatient clinic (~10,000 patients/month) to conduct a randomized trial of IPT versus fluoxetine for MDD and/or PTSD. Local non-specialists will be trained in mental health care for the SMART and hired through the Kenyan Ministry of Health to work at KCH. SMART participants will be randomized to: (1) first line treatment with IPT or fluoxetine; (2) second line treatment for non-remitters- treatment "switch" (e.g., IPT to fluoxetine) or treatment "combination" (e.g., addition of IPT to fluoxetine). Research with mental health specialists in high income countries suggests that antidepressants and psychotherapy have equivalent short-term efficacy and that psychotherapy yields superior long-term relapse prevention. We will test the role of previously identified mechanisms in mediating remission and key moderators of treatment effect. Results of moderator and Q learning analyses will produce first and second-line non-specialist treatment algorithms.

Inclusion Criteria:

1. Kisumu County Hospital (KCH) adult primary care outpatient clinic attendees who screen positive for depression and/or PTSD
2. Ability to attend weekly IPT sessions/fluoxetine monitoring; (3) 18 years or older

Exclusion Criteria:

1. Cognitive dysfunction compromising ability to participate in IPT or accurately take fluoxetine (lack of orientation to person, place, time and situation)
2. acute suicidality requiring higher level of care
3. drug/alcohol use disorders requiring substance use treatment (AUDIT score of 8 or higher, DAST score of 3 or higher)
4. history of mania or requiring treatment for hypomania
5. Outside mental health treatment during the study treatment phases (any mental health treatment is allowed during follow-up phases and is recorded by study team).

NIMH Data Archive The National Institutes of Health (NIH) and NIMH have developed a federation of data repositories called the NIMH Data Archive (NDA) to store the collection of data from participants in research studies related to mental health, regardless of the source of funding. The extensive information collected by these studies, and subsequently stored in the National Database for Autism Research (NDAR), the NIH Pediatric MRI Repository (PedsMRI), the National Database for Clinical Trials Related to Mental Illness (NDCT), and the Research Domain Criteria Database (RDoCdb) provides a rare and valuable scientific resource. The NIH and the NIMH seek to encourage the use of these resources to achieve rapid scientific progress. In order to take full advantage of such resources and maximize their research value, it is important that data be made available, on appropriate terms and conditions, to the largest possible number of qualified investigators in a timely manner.

Data will be made available to the NIMH data archive as soon as it is verified as clean and complete, with at least annual updates throughout the duration of the study. Data will be be available indefinitely or for as long as the NIMH repositories supports the electronic data storage.

The investigative team and our NIMH and GACD partners will review requests for data. Evaluation will include scientific merit of the proposal, overlap versus building upon the study goals.

participants will be evaluated by an outcomes assessor that is not aware of which treatment the participant is receiving. This will be achieved by keeping the randomization key locked and accessible only to the study coordinator and investigators. Participants who are scheduled for assessments will be reminded not to spontaneously disclose their treatment modality to the outcomes assessor.