This study investigates the relationship between major affective disorder and substance abuse, the presence of intoxicated states at time of death, and health care utilization patterns in a sample of 108 consecutive completed adolescent suicides occurring in from contiguous Illinois counties. We will study 2 comparison groups using the same instruments and procedures: (a) matched adolescents living in the same neighborhood as the suicide victim, and (b) adolescents who die as passengers auto accidents. The Neighborhood Sample will allow us to estimate base rates of psychopathology among adolescents in the same communities as the suicide victim so that we can evaluate the plausibility of our method and raters with reference to published epidemiological studies, and so we can assess the suicide victims against the background of community-based peers. The Car Accident Sample wig allow us to apply the same method (multiple informants, retrospective recall) to family informants who are acutely bereaved after the sudden and violent loss of a child. We have chosen these comparison groups to learn the degree to which methodological strategies (e.g., method of combining informant reports, impact of violent death & grief on the report of informants, lapse of time since death) confound a psychological autopsy of adolescent suicide. The use of structured clinical interviews with DSM-M-R criteria & experienced clinical interviewers will enable us to estimate sturdy prevalence rates of major affective disorders, substance abuse/dependence, and other major psychiatric disorders, so we will be able to compare our findings with those from other clinical studies of adolescent populations. Standardized toxicology assays performed in one centralized laboratory will establish the presence, types, and levels of alcohol and psychoactive drugs in the blood of all suicide victims at death. We will estimate the prevalence rates for major affective disorder and substance abuse/dependence in a sample of consecutive adolescent suicide victims, adolescent car accident victims, and normal adolescents matched to the suicide victims. We hypothesize that: (a) the one-year prevalence rates for concomitant affective disorder and substance abuse/dependence are higher in the Suicide Sample than in either of the other two, and that a family history of concomitant affective disorder and substance abuse/dependence is more common in the Suicide Sample than in either of the other two; (b) the prevalence of alcohol and psychoactive drug use during the hours preceding death is not higher in the Suicide Sample than in the Car Accident Sample; and (c) health care utilization rates during the year prior to death are higher in the Suicide Sample than in either of the other two. The linking of psychiatric factors implicated in adolescent suicide with information about health care utilization may offer new insights of immediate practical value to primary
