Mayou et al (Reference Mayou, Ehlers and Hobbs2000) conclude in their 3-year follow-up study of road traffic accident victims that psychological debriefing is ineffective and has, in fact, adverse long-term effects. The intervention group reported significantly worse outcome at 3 years in terms of more severe psychiatric symptoms, impact of event symptoms, anxiety, depression, obsessive-compulsive problems and hostility, pain, major chronic health problems and financial problems. The findings support the suggestion that routine use of psychological debriefing among trauma victims should be discontinued (Reference Bisson, Jenkins and AlexanderBisson et al, 1997).

However, this conclusion is premature. A most serious problem in previous research is that the term psychological debriefing has been used for different types of interventions, for example, in terms of number of sessions and individual or group debriefing. Mayou et al offered individual one-session intervention, without any followup. This kind of intervention is contrary to most clinical thinking: first, assess the trauma; second, offer treatment accordingly. Nobody would recommend that all victims of traffic accidents should be given a standard surgical procedure of 15 minutes in the operating room. For patients with major traumas, the results may be worse than having no operation. The conclusion based on such an approach might easily be that surgery after traffic accidents should not be performed.

A flexible and individual approach is a much more reasonable and appropriate strategy (Reference Rose, Brewin and AndrewsRose et al, 1999). Future studies of psychological debriefing should use an individualised design including screening of psychopathology before intervention, if any, is offered. To assess the effect of one session of debriefing, only subjects who are likely to benefit from such a limited intervention should be included (i.e. those who are at greatest risk for post-traumatic stress disorder should be excluded).

The Impact of Event Scale scores for patients with high initial scores was 25.9 v. 11.8 in the control group (Reference Mayou, Ehlers and HobbsMayou et al, 2000). This may indicate that the trauma of meeting a debriefer for 1 hour was comparable to the trauma of the traffic accident itself.

If the findings of Mayou et al are valid, it shows a tremendous potential for psychological intervention. However, we are still not convinced that a 1-hour psychological intervention unintentionally can do so much harm.