Reducing the risk of suicide attempt and suicide is a public health priority. There has been no substantial decrease in suicide for the past two decades. Prevention efforts to date have focused on primary (e.g. public service announcements) and tertiary (e.g. interventions following suicide attempt) methods. Secondary or indicated prevention has been relatively unexplored. Secondary prevention requires both accurate screening methods and effective interventions. These essential elements are now available. Our recent research demonstrates that responses to the suicidal ideation item of the PHQ depression scale are a powerful predictor of subsequent suicide attempt and death. Availability of the PHQ in electronic medical records creates an opportunity for accurate population-level screening. Dialectical behavior therapy (DBT) has strong evidence of tertiary efficacy for preventing suicide attempts in clinical populations. Brief outreach tertiary interventions such as caring messages have some evidence for preventing suicide attempts. These clinical or tertiary prevention interventions provide the best available evidence for use in building secondary prevention programs. Resulting programs could then be linked to population-level screening data in electronic medical records. The first goal of the research plan is to adapt DBT and caring messages to fit delivery models suitable for the large numbers of at-risk patients identifiable in healthcare settings. The second goal is to evaluate the feasibility, acceptability, and safety of these alternative online suicide and self-injury secondary prevention programs. Online delivery models are suitable because they are scalable and can be provided securely, cheaply, and utilizing existing systems in healthcare. Three brief online interventions will be evaluated: caring email (CE); CE + DBT online program; and CE + DBT online program + coach. Each intervention will supplement usual care among high- risk patients identified via PHQ depression scales collected at outpatient primary care and mental health visits. Intervention content will be drawn from the principles of DBT and caring messages. The study design will be additive in order to examine alternative intervention models that vary widely in resources required for large- scale delivery. However, these interventions will require vastly fewer resources than in-person or telephone interventions. The project will involve intervention feasibility (pretesting, N=60) and acceptability and safety testing (pilot, N=400). Acceptability wll be assessed by patient intervention engagement levels (requiring the large pilot sample) and qualitative/formative intervention feedback from patients. Safety will be assessed via rates of psychiatric hospitalizations and self-injury diagnoses in the medical record in the intervention conditions compared to those receiving just continued usual care. Results of the pilot study will inform the design of a full-scale effectiveness trial examining the impact of one or more of these interventions on risk of suicide attempt and/or suicide death.
