Clinical Decision-Making Following Disasters: Efficient Identification of PTSD Risk in Adolescents

Abstract

The present study aimed to utilize a Receiver Operating Characteristic (ROC) approach in order to improve clinical decision-making for adolescents at risk for the development of psychopathology in the aftermath of a natural disaster. Specifically we assessed theoretically-driven individual, interpersonal, and event-related vulnerability factors to determine which indices were most accurate in forecasting PTSD. Furthermore, we aimed to translate these etiological findings by identifying clinical cut-off recommendations for relevant vulnerability factors. Our study consisted of structured phone-based clinical interviews with 2000 adolescent-parent dyads living within a 5-mile radius of tornados that devastated Joplin, MO, and northern Alabama in Spring 2011. Demographics, tornado incident characteristics, prior trauma, mental health, and family support and conflict were assessed. A subset of youth completed two behavioral assessment tasks online to assess distress tolerance and risk-taking behavior. ROC analyses indicated four variables that significantly improved PTSD diagnostic efficiency: Lifetime depression (AUC = .90), trauma history (AUC = .76), social support (AUC = .70), and family conflict (AUC = .72). Youth were 2-3 times more likely to have PTSD if they had elevated scores on any of these variables. Of note, event-related characteristics (e.g., property damage) were not related to PTSD diagnostic status. The present study adds to the literature by making specific recommendations for empirically-based, efficient disaster-related PTSD assessment for adolescents following a natural disaster. Implications for practice and future trauma-related developmental psychopathology research are discussed.

Keywords: Adolescents; Evidence-based assessment; PTSD risk assessment; Stress disorders; Traumatic stress.

Figure 1

Population pyramids for predictor variable distributions for those with a diagnosis of PTSD (green bars, right of center axis) versus no diagnosis of PTSD (blue bars, left of center axis). Predictors represented in each cell are as follows: (A) Number of prior potentially traumatic events, PTEs. (B) Total scores on the Social Support for Adolescents questionnaire. (C) Lifetime total score on the CRAFFT substance abuse screener. (D) Total time to quit on the BIRD task. (E) Depression symptoms, count. (F) Disaster impact severity score. (G) Number of pumps on non-explosion trials of the BART task. (H) Total scores on the Conflict Beliefs Questionnaire. Figures generated in SPSS.

Figure 2

Nomograms for adolescents with high conflict and low social support