Esketamine Versus Crisis Response Planning Versus Enhanced Treatment as Usual for Suicide Prevention: A Pragmatic Randomized Trial in a Brazilian Municipality (SAVE)

The SAVE study will test whether a single esketamine infusion or a single Crisis Response Planning session, each added to enhanced treatment as usual, reduces future suicide-related events compared with enhanced treatment as usual alone. The study takes place in the public emergency care network of Indaiatuba, São Paulo, Brazil, and includes adolescents and adults aged 14 years or older who have attempted suicide within the previous 30 days or currently have severe suicidal thoughts with intent to act. The main questions are whether either intervention reduces suicide-related events over 12 months and whether the interventions improve suicidal thoughts, depression, anxiety, sleep, well-being, hopelessness, and quality of life. Researchers will also evaluate acceptability, feasibility, use of health services, and costs. Anticipated total enrollment is 478 participants: 10 participants in a separate pilot cohort and 468 in the main cohort used to test the study hypotheses. The main cohort will be assigned by chance, in equal numbers of 156 participants, to one of three groups: 1. Esketamine plus enhanced treatment as usual: one intravenous infusion of esketamine at 0.375 mg/kg over 40 minutes, with medical supervision and monitoring of heart rhythm, blood pressure, and oxygen levels. Participants remain under clinical observation, with discharge after 24 hours if clinically stable. 2. Crisis Response Planning plus enhanced treatment as usual: one 20-to-45-minute session with a trained clinician to develop a personalized plan covering warning signs, coping strategies, reasons for living, support contacts, and emergency resources. Participants receive a printed plan and a digital copy. 3. Enhanced treatment as usual alone: early outpatient psychiatric consultation, arranged to take place within seven days of randomization, plus lethal means safety counseling to reduce access to potentially lethal means. Both components are offered to participants in all three groups, alongside routine emergency care. The study records whether each component was delivered; booking a consultation alone does not count as receiving it. Participants will complete assessments at enrollment, 24 hours, seven days, and weeks 2, 4, 8, 16, 24, 32, 40, and 52. Blood samples will be collected at enrollment for exploratory analyses of biological factors that may be associated with treatment response. Safety monitoring and contact to identify new events will continue throughout follow-up. Participants will also use a smartphone application to answer brief questions about their mood, thoughts, and experiences during three periods of 30 consecutive days, beginning at enrollment and at calendar months 4 and 8. There will be three prompts each day: two at fixed times, 09:00 and 21:00, and one at a randomly selected time between 10:00 and 20:00, using local time in Indaiatuba. This represents 90 assessment days and 270 scheduled prompts over the study. The study team will contact participants within 24 hours of a safety alert through a dedicated study mobile phone with WhatsApp. Participants will be instructed to seek emergency care immediately when needed rather than wait for a study response. The main outcome is the time to the first qualifying event: a suicide attempt, including an attempt stopped by the person or interrupted by someone else; a psychiatric admission to prevent suicide; death by suicide; or self-injury requiring emergency department care. An external adjudicator who does not know the assigned treatment will review suspected events and determine whether they meet the study definition. Outcome assessors will also be unaware of treatment allocation. Participants who experience a qualifying event may be offered rescue treatment combining esketamine and Crisis Response Planning, depending on clinical eligibility and safety. They will remain in follow-up and in analyses according to their original randomized group. The first qualifying event will still count in the main analysis. Pilot data will be described separately and will not be included in confirmatory efficacy analyses.