BACKGROUND This study examined the ability of commonly used questionnaires and a structured clinical interview to predict PTSD in recent trauma survivors.

METHOD Horowitz's Impact of Event Scale (IES), Speilberger's State Anxiety (SANX) and the Peri Traumatic Dissociation Questionnaire (PDEQ) were administered one week post-trauma to 239 traumatised individuals recruited from a general hospital emergency room. The IES, the SANX, the civilian version of the Mississippi Scale for Combat Related PTSD (MISS), and the Clinician Administered PTSD Scale (CAPS) were administered one month and four months post-trauma. Receiver operator characteristic (ROC) analysis was used with these data.

RESULTS All questionnaires were better than chance at predicting PTSD. The so-called PTSD questionnaires (IES and MISS) were not better than the more general ones. No difference in predictive value was found when questionnaires were carried out one week or one month after a trauma. Recovery was better predicted than PTSD, and the CAPS was better than the questionnaires.

DISCUSSION The use of psychometrics and clinical interviews to predict PTSD should be guided by clinical relevance and by the availability of resources.