eMJA: Should we debrief and counsel people who have had psychological shock?

EBM: Trials on trial

Should we debrief and counsel people who have had psychological shock?

Christopher B Del Mar

MJA 2002 177 (5): 258-259

Abstract

—

Introduction

—

Question

—

Commentary

—

Rationale for the trial

—

Trial methods

—

New information

—

Implications for clinical practice

—

References

—

Author details

Trial: Hobbs M, Mayou R, Harrison B, Worlock P. A randomised controlled trial of psychological debriefing for victims of road traffic accidents. BMJ 1996; 313: 1438–1439.
Mayou R, Ehlers A, Hobbs M. Psychological debriefing for victims of road traffic: three year follow-up of a randomised controlled trial. Br J Psychiatry 2000; 176: 589–593.

QuestionHow effective is psychological debriefing for people subject to unpleasant psychological shock?

Trial details

Design: Randomised controlled trial of psychological debriefing.

Setting: A British teaching hospital (the Radcliffe Hospital, Oxford).

Patients: 66 men and 40 women, aged 17–69 years, admitted to hospital after a motor vehicle accident. Most had been the driver of a car. Median admission duration was four days for the 52 control patients and eight days for the 54 who underwent the intervention.

Interventions: A debriefing of about one hour on Day 2 of admission, encouraging patients to describe the accident and express their emotions, followed by a cognitive appraisal which included describing common reactions to traumatic experiences and suggesting a range of people who might be able to assist in the future, including the patient's general practitioner. 91 patients were assessed at four months and 61 were assessed at three years. Control patients had no debriefing or counselling.

Main outcome measures: Impact of Event Scale (IES, which focuses on intrusive thoughts and avoidance of similar situations to the event); Brief Symptom Inventory (BSI, a measure of 53 symptoms); and other questions related to physical pain and functional activities.

Main results: At four months there was still considerable psychological morbidity among the patients who were followed up. There was a significant difference (P < 0.05) in changes of IES between the 42 who received the intervention, in whom it increased from 15 (standard deviation [SD], 15) to 16 (SD, 15), and the 49 controls, in whom it fell from 15 (SD, 12) to 13 (SD, 14). Similarly, two subscales of the BSI score changed significantly between the intervention group, among whom it deteriorated from 0.5 (SD, 0.5) to 0.6 (SD, 0.8), and the control s, in whom it hardly changed from 0.4 (SD, 0.3) to 0.4 (SD, 0.4). Among the 61 patients followed for three years, the 30 randomised to receive the intervention were significantly worse, by self-report, both psychologically and physically. Their mean IES score deteriorated from a baseline of 15 (SD, 14) to 16 (SD, 18). In comparison, scores for the 31 control patients improved from 16 (SD, 12) to 13 (SD, 17). The difference in change was significant (P < 0.05). Among all patients with high initial scores, these decreased among the controls but not among those receiving the intervention.

Conclusion: Psychological counselling should only be used in the context of trials rather than routine care.

Commentary

Rationale for the trial

Victims of psychological shock are vulnerable for many years after the incident to a cluster of symptoms (including intrusive thoughts, avoidance symptoms, functional problems extending to social problems, and pain and other physical symptoms), commonly called "post-traumatic stress" disorder. The response of society has been to try to deflect this.1 For example, after the Port Arthur tragedy, teams of counsellors were flown to the scene and made available for those who wanted to talk through the horrors of the incident. The trial under consideration was a clever way of testing such debriefing for people subject to a sudden, unexpected event in which some were physically hurt.

Trial methods

The methods appear to have been well constructed. However, despite random allocation by random number tables, the patients in the intervention group were worse affected (nearly double the hospital stay and higher IES scores). Presumably, this happened by chance, but it might explain some of the result. Also, the rate of follow-up at four months was higher among control patients (49 out of 52) than intervention patients (42 of 54), although both follow-up rates were equally poor at three years. It is reassuring that the different measures of severity were the same for those followed up and those not.

New information

The results are compelling. There was a very strong suggestion that the sort of counselling we expect should be effective was actually harmful. It might be that we have some sort of mechanism ("denial", or a method of dismissing disturbing memories from our minds) that protects us from further psychological effects that interfere with normal living.

Moreover, this finding is supported by evidence from another trial in which women who had had operative childbirth were debriefed to defer postnatal depression. This intervention also failed to be effective, with a trend in the direction of causing more harm than good.2

Implications for clinical practice

This trial's findings challenge an assumed and conventional approach to protecting people against post-traumatic stress disorder. It may be another case of the empirical informing the psychopathological process.

We should reconsider the impulse to assist people who have undergone an unpleasant psychological event by counselling them and getting them to talk about it and relive the event. It may be harmful. Could it be that the unfashionable British "stiff upper lip" is the better approach after all?

We should provide such psychological counselling only in the context of trials rather than routine care.

The stroke of genius here was that the authors thought to challenge an intervention whose benefits seemed self-evident. From this we learn that we must continue to do so with many other interventions whose benefits we take for granted.

References

Ornstein R. Approach to the patient following a traumatic event. In: Stern TA, Herman JB, Slavin PL, editors. The MGH guide to psychiatry in primary care. New York: McGraw-Hill; 1998: 327–334.
Small R, Lumley J, Donohue L, et al. Randomised controlled trial of midwife led debriefing to reduce maternal depression after operative childbirth. BMJ 2000; 321: 1043–1047. <PubMed>

(Received 9 Jan 2002, accepted 1 Mar 2002)