Motor vehicle accidents (MVAs) represent one of the most common causes of posttraumatic stress disorder (PTSD) worldwide. Predicting those MVA survivors who are likely to experience PTSD in the longer-term has attracted substantial research attention, but most investigators have concluded that early traumatic stress symptoms have only moderate predictive power. The present study focuses on a decision-tree approach that might be useful to clinicians attempting to identify subgroups of MVA survivors with graduated degrees of risk. Eighty-three consecutive MVA admissions (response rate, 65%) were assessed by structured clinical interview within 2 weeks of the accident and at 18 months follow-up. Meeting full criteria for PTSD (other than the time criterion) at baseline achieved a positive predictive power of 0.92 in identifying those who had PTSD over the following 18 months. For the remainder, the PTSD "arousal" domain achieved a positive predictive power of 0.81 for predicting those with either subthreshold PTSD or full PTSD over the 18-month follow-up period. The implications for further research into a stepped approach to intervention and monitoring are discussed. (Author/publisher)
Abstract