Use of the Win Ratio for Suicide Prevention Research

The Win Ratio (WR) is a novel approach to analyze composite outcomes that accounts for the relative priority of each outcome. In this study, we evaluated use of the WR for composite outcomes in suicide prevention studies to overcome challenges associated with low base rates of primary outcomes.

We illustrate the utility of the WR in suicide prevention studies through two re-analyses: 1) data from the Emergency Department Safety Assessment and Follow-up Evaluation (ED-SAFE) trial, and 2) Medicaid claims data comparing youths with concussions to those with orthopedic injuries. Statistical power of the WR is illustrated through simulations.

In the ED-SAFE trial, with ordered outcomes of suicide death, suicide attempt, interrupted attempt, aborted attempt, and preparatory acts, the active intervention had a WR of 1.26 (95% CI: 1.04-1.53, p = 0.018) compared to treatment as usual. For the Medicaid data, using a matched sample of 42,313 youths we found that youths with orthopedic injuries had a WR of 1.10, 95% CI (1.07, 1.12), p < 0.001 relative to concussed youths for the ordered outcomes of suicide; deliberate self-harm/suicide attempt; psychosis, externalizing, or internalizing disorders; migraines or sleep problems; and post-baseline concussions. Both results are consistent with the originally published findings.

The Win Ratio preserves the clinical hierarchy of events and offers a robust analytic approach to account for multiple outcomes. By addressing challenges related to low event rates, the WR can enhance study efficiency and interpretability, supporting its broader use in suicide prevention research.