A randomised controlled trial of psychological debriefing for victims of road traffic accidents

BMJ 1996; 313 doi: (Published 07 December 1996) Cite this as: BMJ 1996;313:1438
  1. Michael Hobbs, consultant psychotherapista,
  2. Richard Mayou, reader in psychiatrya,
  3. Beverly Harrison, research assistanta,
  4. Peter Worlock, consultant trauma and orthopaedic surgeonb
  1. a Warneford Hospital, Oxford OX3 7JX
  2. b John Radcliffe Hospital, Oxford OX3 9DU
  1. Correspondence to: Dr Hobbs.
  • Accepted 28 October 1996

Psychiatric problems are common after major or minor road accidents. 1 2 3 4 Mayou et al found that one fifth of victims developed an acute stress reaction and one quarter displayed psychiatric problems within the first year.1 Long term psychiatric problems were mainly mood disorder (10%), phobic travel anxiety (20%), and post-traumatic stress disorder (11%). Interest has been stimulated in preventive interventions, especially routine psychological “debriefing.” Although widely used after trauma, no randomised controlled trials of debriefing have been reported, and debriefing may sometimes increase distress.5 This randomised controlled study aimed to test whether a single debriefing could reduce post-traumatic psychopathology in road accident victims.

Subjects, methods, and results

The subjects were victims of road accidents admitted consecutively to the John Radcliffe Hospital. We excluded those who could not remember the accident and those with no psychological symptoms, who are at low risk of later problems. Others were excluded because they had been discharged or were not available when the researcher came to visit them. Eight people refused to participate. Those who agreed were allocated randomly to intervention or control group using a random number table. All subjects were screened initially using a semistructured interview based on previous research and two standard self report questionnaires, the impact of events scale for specific post-traumatic symptoms and the brief symptom inventory, which generates a global emotional distress score.

Interventions were undertaken within 24–48 hours of the accident in most cases; they comprised an hour's debriefing combining a review of the traumatic experience, encouragement of emotional expression, and promotion of cognitive processing of the experience. Advice was provided about common emotional reactions, the value of talking about the experience, and early graded return to normal road travel. All experimental subjects received an information leaflet consolidating this advice and encouraging the support of family and friends. Both groups were reassessed by interview and the self report questionnaires at four months.

The intervention group (n = 54) had a higher mean injury severity score and longer hospital stay than the controls (n = 52), but there was no significant difference in baseline post-traumatic or other psychiatric symptoms. It proved difficult to interview patients during brief admissions to busy wards. Despite intensive efforts, follow up data were not obtained from 22% of the intervention group and 6% of controls. Non-respondents did not differ on baseline characteristics.

Neither group showed a significant reduction in specific post-traumatic symptoms, mood disorder, anxiety, the self reported measures, interview ratings of intrusive thoughts or travel anxiety, or clinical diagnosis of posttraumatic stress disorder or phobic anxiety (table 1). The intervention group had a worse outcome (P <0.05) on two subscales of the brief symptom inventory, however, and a (non-significantly) poorer outcome in terms of emotional distress.

Table I

Baseline data and assessment of road accident victims before and after psychological intervention

View this table:


The study showed the practical difficulties of offering an early psychological intervention to road accident victims. Despite the small numbers and lack of complete follow up data, the findings are clinically convincing. Psychiatric morbidity was substantial four months after injury, with no evidence that debriefing had helped—and, indeed, indications that it might have been disadvantageous.

Several explanations are possible: rapid discharge from hospital necessitated very early intervention and some patients were still too numbed or distressed to be receptive; the intervention did not seem relevant to subjects expecting an uneventful recovery; a single intervention was inadequate for major emotional problems; and early interventions may disturb natural psychological “defences” against fear and distress.

We believe that later psychological intervention might better reduce psychiatric morbidity and that selective interventions designed to address specific emerging problems—for example, phobic avoidance of road travel—would be more effective than generalised psychological debriefing. Identifying patients at risk after discharge would, however, require effective collaboration between hospitals and general practitioners.


  • Funding Oxford regional research fund.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract