The current application has been submitted as a Small Grant to pilot new methodology by a new investigator. The proposed project has two principal aims: (1) To develop an efficacious self-report screener for identifying delinquent adolescents who have engaged in recent suicidal behavior. (2) To take the first step in testing a model of the psychosocial markers associated with suicidal behavior in delinquent adolescents. The 24 months of the proposed study are divided into three phrases. The first four months will be spent on instrument development, piloting all research procedures, and creating a distribution of scores on the initial screener questionnaire. During months 5-20, the questionnaire incorporating items which will become the screener will be administered at two time points (one week apart) to over 500 adolescents (ages 13-19) in a county juvenile detention center. The screener questionnaire will be administered in a group format to all participants during the school class period. To evaluate components of the screener questionnaire and to gather more detailed information, 50 high scoring (i.e., above the 66th percentile on the screener) and 50 low scoring adolescents will be randomly selected to participate in an extensive diagnostic interview. The diagnostic interview will consist of a semi-structured interview and an additional questionnaire. During the last four months of the study, data entry and analysis will be completed and manuscripts for publication will be prepared. Data will be examined relative to a conception of suicidal behavior that postulates the presence of distal factors (e.g., comorbid psychiatric disorders, prior suicide attempt) that contribute to the development of a high-risk youth. More proximal factors exist, including potentiating factors (primarily major life events) and protective factors (i.e., perceived social support, effective coping style). It is hypothesized that the absence of protective factors will exacerbate the impact of major life events and lead to suicidal ideation. The path from suicidal ideation to attempts is hypothesized to be mediated by two factors: (a) impulsivity (a result of personality style and/or substance abuse), and (b) dangerous setting (due to access to lethal means, lack of parental monitoring, and suicidal behavior by others). Prior to our primary analyses, we will evaluate the representatives of our sample. Analyses pertaining to the psychometric properties of the screener questionnaire will focus on internal consistency, test-retest reliability, sensitivity, specificity and positive predictive value. Initial models of suicidal ideation and suicide attempts will be tested using logistic regression approaches.
