Evaluation of the effectiveness of suicide prevention programs: a methodological perspective

Many studies that attempt to evaluate the effectiveness of treatment programs for suicidal patients are methodologically deficient in one or more areas. This paper outlines six criteria that should be met in designing such investigations: Patients should be randomly assigned to groups; at least 80% of the subjects who were initially enrolled should be followed up; the results should be both statistically significant and clinically important; all clinically relevant outcomes should be reported; the patients must be adequately described; and the new intervention should be feasible in other settings. It is shown that these criteria are both necessary and achievable.