Our objective was to identify factors that predict occurrence and severity of post-traumatic stress disorder (PTSD) after a terrorism attack.

We evaluated 32 victims of a bomb attack in a Paris subway in December 1996 at 6 and 32 months.

Sociodemographic characteristics, clinical data and physical injuries were used to predict PTSD occurrence and severity in 32 victims. The Watson’s PTSD Inventory (PTSD-I) and the Impact of Event Scale (IES) by Horowitz were used to evaluate occurrence and severity of PTSD, respectively.

Thirty-nine percent of participants met PTSD criteria at 6 months, 25% still had PTSD at 32 months. Women had PTSD 32 months after the bomb attack more frequently than men. Employment predicted PTSD severity at 32 months. PTSD scores assessed by PTSD-I at 6 months were significantly and positively associated with IES scores at 32-month follow-up (r = 0.55, P = 0.004). Psychotropic drug use before the bomb attack significantly predicted PTSD occurrence and severity at 6 and 32 months. In a linear regression model, physical injuries, employment status and psychotropic drug use before the bomb attack were independent predictors of severity of PTSD at 32 months.

Bomb attack exposure resulted in persisting PTSD in a significant proportion of victims; the severity was predicted at 32 months by physical injuries and psychotropic drug use before the terrorism attack and by the PTSD score few months after the bomb attack.