The OptimizeD study aims to improve outcomes in depression in primary care in India. This study will randomize 1500 patients with moderate to severe depression to either psychotherapy based on behavioral activation called the Healthy Activity Program (HAP) or antidepressant medication (fluoxetine).

The study has two primary objectives:

1. Use patient characteristics to generate a precision treatment rule based on baseline information for predicting in advance what works best for whom (and which patients are unlikely to respond to either treatment and should be referred to specialist care).
2. Conduct a cost-effectiveness analysis by comparing relative costs and effectiveness between those who were randomly allocated to their optimal treatment with those who were randomly allocated to a non-optimal treatment based on the precision treatment rule.

Depression is the leading mental health contributor to the Global Burden of Disease. The World Health Organization's mhGAP initiative advocates the use of brief psychological therapies such as behavioral activation or antidepressant medications as first-line options for the treatment of moderate to severe depression in primary care settings, but not all patients will fully remit on either treatment. It is likely that different patients will respond to different treatments, but the optimal treatment for each individual remains unknown (and which patients are unlikely to respond to either treatment and should be referred to specialist care). Enhancing our ability to determine the optimal intervention for a particular patient has the potential to enhance the overall effectiveness of mental health care delivery in a more cost-efficient manner. This is a critical gap in knowledge in the treatment of depression across clinical settings globally.

The main objective of the OptimizeD study is to determine whether different patients respond differentially to brief psychological treatment or a widely used generic SSRI and, if so, whether one can optimize outcomes in a cost-effective fashion for primary care patients with moderate to severe depression.

The study has two specific aims and two exploratory aims:

• Specific Aim 1 (Clinical and Functional Outcomes): To evaluate the effectiveness of optimization via generating a precision treatment rule (PTR) on patients with moderate to severe depression randomized to either psychotherapy based on behavioral activation called the Healthy Activity Program (HAP) or antidepressant medication (fluoxetine). The study will use machine learning to develop the PTR, using a wide range of clinical, socio-economic, and neuro-cognitive characteristics measured at baseline as predictors. The investigators hypothesize that patients randomized by chance to their optimal intervention will be more likely to remit and recover than patients who are not.
• Specific Aim 2 (Cost-effectiveness Outcomes): To assess the costs of optimal vs. non-optimal treatments and to conduct a cost-effectiveness analysis by comparing relative costs and effectiveness between those who were randomly allocated to their optimal treatment with those who were randomly allocated to a non-optimal treatment, based on the PTR developed in Aim 1. The investigators hypothesize that optimizing will be more cost-effective than not.
• Exploratory Aim 1 (Mediators): To explore whether one can use the PTR to make our tests of mediation more precise. Patients who respond differentially to different treatments adhere to different causal mechanisms, and inclusion of the PTR in interaction terms with the purported mediators should facilitate the detection of moderated mediation among patients who show specificity of response. The investigators will also consider whether treatment-related factors (e.g., adherence, quality) act as mediators of the effects of each treatment on remission and recovery. This exploratory aim will offer insights into mechanisms of action for each treatment.
• Exploratory Aim 2 (Genetic Predictors): To explore whether polygenic risk scores and other biomarkers can enhance the prediction of both general and differential response to either treatment.

As a secondary objective, the study will evaluate the effectiveness of optimization in the long-term. The investigators hypothesize that patients allocated to their optimized treatment will be more likely to recover than patients who are allocated to their non-optimal treatment.

All trial procedures used in the main trial were first evaluated in a pilot study (NCT06153004), which employed the same design and methods to assess feasibility across recruitment, randomization, retention, treatment adherence and fidelity, and data quality assurance.

Inclusion Criteria:

• Participants will be adults aged 18 or over of any gender attending one of the selected Primary Health Care Centers with a "diagnosis" of moderate to severe depression based on scores of 10 or above on the Patient Health Questionnaire-9 (PHQ-9).

Exclusion Criteria:

1. Women who are pregnant or are breastfeeding or lactating
2. Patients with a history of psychosis, including schizophrenia spectrum disorders or bipolar disorder.
3. Participants planning to permanently move out of the study area during the follow-up period.
4. Patients with evidence of cognitive impairment.
5. Patients who do not speak either English or Hindi.
6. Patients who are undergoing treatment for depression at the time of recruitment or who completed treatment within one month prior to recruitment
7. Patients at imminent risk for suicide
8. Patients from households in which another member has been recruited into the study

Individual participant data will be shared through the NIMH Data Archive. To ensure participant privacy and data security, all shared data will be de-identified according to established protocols and guidelines. The data sharing plan will be periodically reviewed and updated, as necessary, to align with evolving best practices, policies, and guidelines for data sharing.

Data will be available indefinitely or for as long as the NIMH repositories support it.

Researchers and interested parties who wish to access the data can submit a formal data access request to the NIMH Data Archive. Researchers who gain access to the data will be required to comply with the data usage policies and guidelines set forth by the NIMH Data Archive.

Masked (blinded) field-based assessors will conduct all outcome assessments