Predictors of response to treatment for suicidal behavior

Abstract

Military personnel who had exhibited serious suicidal behavior participated in a controlled study to investigate the effects of an intensive cognitive behavioral, problem-solving treatment intervention administered in a group format. This study examined five categories of variables which were considered to be potential indicators of treatment responsiveness. These five categories included demographic/psychosocial indices, suicidal symptoms, problem-solving components, personality features, and psychiatric diagnosis. Treatment outcome was determined by continued suicidal ideation as well as problem-solving appraisal. Treatment outcome was assessed over six different time intervals across a twelve-month period. Bivariate correlational analyses, joint contingency table probabilities, and regression analyses were the statistical procedures employed to-determine results. In general, problem-solving indices and psychiatric diagnosis were not predictive of treatment response. Unfavorable short-term treatment outcome was predicted by a higher number of previous suicide attempts, single marital status, negative life stress, hopelessness, and avoidant personality features. Long-term negative treatment response was predicted by higher levels of depression and avoidant personality features. These results demonstrated significant implications for treatment of suicidal individuals. The most critical finding from this study was the need to address the specific avoidant characteristics that suicidal individuals bring to the therapy situation. In addition, specific tactics to retain these subjects in therapy may also be necessary in order to provide adequate treatment.