Although the use of acute inpatient psychiatric facilities for the treatment of adolescent emotional disorders has increased during the past decade, no research to date has assessed the outcome of this treatment modality. The current study proposes to: (1) assess the post-discharge course, including diagnostic stability over time, of adolescents treated in an acute-care psychiatric inpatient setting via a prospective, longitudinal design; (2) discern the differential impacts of demographic, diagnostic, and psychosocial factors which have post-discharge prognostic value in predicting outcome; (3) compare the clinical course of adolescents discharged prematurely from acute-care hospitalization with those who complete a recommended period of hospital treatment, permitting some assessment of the utility of a completed hospitalization course; and (4) test the hypothesized model that risk factors of parent symptomatology, severity of patient psychopathology and negative life stress, in combination with resistance factors of completed hospital course, adherence to follow-up treatment, IQ, family supports and social skills, will significantly predict outcome symptoms at six months and one year following hospital discharge. Subjects will be drawn from all adolescent admissions to an acute care inpatient unit over a one-year period, with total sample size, following attrition, expected to be 100. Outcome data will be collected at discharge, and at follow-up periods of six months and one year. Diagnostic specificity will be established by structured clinical interview (DISC), which is expected to impact on outcome. The primary dependent variables of the study will be standardized behavioral checklists of symptoms (i.e. Youth Self-Report and CBCL), as reported by patient and parents. DISCs at one year follow-up will establish the continuing presence or absence of psychiatric diagnosis. Outcomes of patients who terminate treatment prematurely via 72-hour letter or AWOL will be compared to those of patients who complete inpatient treatment in accordance with clinical recommendations. Repeated-measures ANOVA will be used to assess differences in symptom severity over time, and multiple regression analysis will be used to assess the effects of the eight primary risk and resistance factors on outcome. A path model summarizing the interrelationships of the predictor variables is presented for heuristic purposes, with the estimated path coefficients to yield some measure of the utility of the model's promise in explaining variance in outcome.
