Education and debate

Randomised controlled trials in psychiatry: important but poorly accepted

Gavin Andrews, professor. 

School of Psychiatry, University of New South Wales at St Vincent's Hospital, Sydney, Australia

Correspondence to: G Andrews, Anxiety Disorders Unit, Darlinghurst, New South Wales, 2010 Australia gavina{at}crufad.unsw.edu.au

Identifying the efficacy of treatments from the data from randomised controlled trials is the first step in improving the efficiency of medical care. The implementation of clinical practice guidelines on the basis of this evidence, and prioritising of practice on the basis of cost effectiveness studies, are the next steps. Psychiatry seems nervous about trusting itself to this process. It is 50 years since the first randomised controlled trial in medicine, and doctors in other specialties seem sufficiently comfortable with the method to form research practice consortia and to conduct mega-trials ^{1 2} that are comparatively immune to the methodological and political hazards that can beset small randomised controlled trials and meta-analyses. ^{3 4} Why, then, have mega-trials not begun in psychiatry? There are two important reasons: the types of treatment used, and the nature of the endpoints. Psychological treatments can be effective and, in certain disorders, more effective than drugs.⁵ It is easy to be certain about the quality of drugs, but psychological treatments, unless manualised for both physician and patient, depend on the fidelity with which practitioners carry out the treatment. Next, the 22% of the global burden of disease attributed to mental disorders is made up of 21% from morbidity and only 1% from mortality.⁶ Measuring morbidity or disability is much more complex than measuring mortality, the usual endpoint in mega-trials of physical disorders.

	Influence of randomised controlled trials

Given such problems, has the data from randomised controlled trials changed mental health practice? Randomised controlled trials have been essential in the introduction of new proprietary remedies in psychiatry, largely owing to the activities of the regulatory authorities in each country. Even in non-proprietary remedies they have been critical. Excellent trials were critical in the phasing out of insulin coma therapy for schizophrenia⁷ and in the retention of electroconvulsive therapy as a treatment for severe depression,⁸ but the evidence has not always been so persuasive. Nathan and Gorman⁹ found evidence for efficacy in type 1 randomised controlled trials in 32 drug and 13 psychological treatments for specific mental disorders. Data from randomised controlled trials on efficacy clearly exist. Nathan and Gorman's project, originally sponsored by the American Psychological Association, was disowned before publication, which may be evidence that psychiatry is nervous about identifying with the results of science. And yet clinical care will be harmed if it does not. More than evidence about efficacy is required to introduce a treatment; certain barriers must be overcome, for example, professional competency, organisational structure and social attitudes that affect the preparedness of the profession to carry out, and the public to accept, good new treatments.

Considerable evidence shows that family interventions double the efficacy of drug therapy in schizophrenia,¹⁰ but the interventions have been largely ignored in practice. The investigators trained community mental health service staff to conduct family interventions in patients with schizophrenia. When the investigators reviewed progress,¹¹ no staff were using the intervention. Two reasons were postulated. Firstly, the administrators' refusal to allow time in lieu for staff to work at night and weekends when families are together, and secondly, that staff were pleased to go back to the therapies that they were comfortable with, as though their comfort was more important than the patients' well being. In this example, the adoption of scientific evidence was inhibited by organisational and professional barriers. Long term dynamic psychotherapy derived from the ideas of Freud has not, in the 50 years since the first randomised controlled trial, been shown to be superior to talking to a mature and kindly adviser.¹² There are sufficient negative randomised controlled trials to conclude that dynamic psychotherapy is not efficacious. Yet many patients expect to benefit from such therapy, and many psychiatrists only offer psychotherapy, despite evidence from randomised controlled trials that drugs or cognitive behaviour therapy would be more beneficial. Patients have a right to effective treatments, and physicians who choose to offer unproved treatments, no matter how accepted such treatments are, do so at their own risk.¹³ In this example, evidence of ineffectiveness has not been sufficient to limit the use of a treatment. What people think about a treatment seems to be as important as the evidence from the randomised controlled trial in determining what is actually done, at least in a specialty in which some of the treatments are difficult to standardise, endpoints are difficult to measure, the size of the trials is small, and the profession is reluctant to trust the results.

If what people think about a treatment is as important as the evidence of efficacy then the introduction of evidence based medicine to psychiatry, although essential, will not be straightforward. From 1981 to 1991 a project under the aegis of the Royal Australian and New Zealand College of Psychiatrists published 10 evidence based clinical practice guidelines for the common mental disorders.¹⁴ By the time the guidelines were ready, the priorities of the government had shifted, and the guidelines were never implemented by the college. In subsequent years evidence based treatment protocols in mental health have been prepared in New Zealand, the United States, and the United Kingdom but, as in Australia, changes in the medical agenda have meant that none became the vade mecum.

	Implementing clinical practice guidelines

Given that whether a treatment is used or not depends on efficacy as determined from randomised controlled trials and acceptability to the clinicians and their patients, how would one go about ensuring that a new set of evidence based medicine guidelines were implemented? Science, in the form of a body of randomised controlled trials, is necessary, but not sufficient, to cause clinicians to alter the way they practise. Firstly, guidelines need to be based on immaculate science; then clinicians need to be informed, exhorted and rewarded; then the procedures should fit within the work environment; and patients should be told what to expect so they can monitor the treatment they are being offered. Finally, whether the adoption of the guidelines has been effective needs to be evaluated. And the cost? Perhaps a sixth of the budget would be used for the preparation of guidelines, and the remainder of the budget would be used for promotion and evaluation.

(Credit: LIANE PAYNE)

Australia has two new evidence based guideline initiatives in mental health that originated from the complaints of dissatisfied groups of funders or consumers. So far there is money for guideline preparation, but not for implementation. Perhaps guideline preparation at times fulfils the important political purpose of satisfying the complainants that something is being done. Other guidelines serve the political needs of a professional group to show their unique professional role¹⁵ and, among other things, what they think they should be funded to do. Some guidelines have been funded by payers to define the professional activities they are prepared to fund,¹⁶ and other guidelines have been linked to the promotion of a proprietary product. Yet the purpose of clinical practice guidelines is to improve the management of disease, whether treated by the various specialist groups, by primary care physicians, or by the patients themselves.

Two years ago a Practice Guideline Coalition of 39 organisations flagged their interest in preparing guidelines.¹⁷ The first guidelines, for panic disorder, are in draft form and favour no profession or treatment modality, but rather are directly related to treatment of the disorder. Perhaps we should follow this model and prepare unbiased guidelines for the treatment of a disease, not for the behaviour of the professionals or funders involved.

No group seems to want to do the evaluation. Yet it is unwise to prepare guidelines and not evaluate them to see if they improve effectiveness. Making certain that a new system of treatment is better is exactly what randomised controlled trials were designed to do, and exactly what Cochrane said should be done,¹⁸ even though effectiveness research is more complex than efficacy research.¹⁹ Medicine is moving in this direction. Psychiatry, which was the first of the specialties to develop clinical practice guidelines, should not let the intrinsic difficulties deter continued commitment to implementation and evaluation of effectiveness.

	Footnotes

Competing interests: None declared.

	References

1.	Collins R, Peto R, Gray R, Parish S. Large scale randomised evidence: trials and overviews. In: Maynard A, Chalmers I, eds. Non-random reflections on health services research. London: BMJ Publishing Group, 1997:197-230.
2.	Early Breast Cancer Trialists' Collaborative Group. Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomised controlled trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Lancet 1992; 339: 1-15[Medline], 71-85.
3.	Wells J. Mammography and the politics of randomised controlled trials. BMJ 1998; 317: 1224-1230[Free Full Text].
4.	Deyo RA, Psaty BM, Simon G, Wagner EH, Omenn GS. The messenger under attack: intimidation of researchers by special-interest groups. N Engl J Med 1997; 336: 1176-1180[Free Full Text].
5.	Andrews G. Talking therapies that work: the rise of cognitive behaviour therapy. BMJ 1996; 313: 1501-1502[Free Full Text].
6.	Murray CIL, Lopez A. Mortality by cause for eight regions of the world: global burden of disease study. Lancet 1997; 349: 1269-1276[Medline].
7.	Ackner B, Harris A, Oldham AL. Insulin treatment of schizophrenia. Lancet 1957; i: 607-611.
8.	Gregory S, Shawcross CR, Gill D. The Nottingham ECT study. Br J Psychiatry 1985; 146: 520-524[Abstract/Free Full Text].
9.	Nathan PE, Gorman JM, eds. A guide to treatments that work. New York: Oxford University Press, 1998.
10.	Mari JDJ, Streiner DL. An overview of family interventions and relapse on schizophrenia. Psychol Med 1994; 24: 565-578[Medline].
11.	Kavanagh DJ, Piatkowska O, Clark D, O'Halloran P. Application of cognitive-behavioural family intervention for schizophrenia in multidisciplinary teams. Aust Psychol 1994; 28: 181-188.
12.	Andrews O. The essential psychotherapies. Br J Psychiatry 1993; 162: 447-451[Free Full Text].
13.	Klerman GL. The psychiatric patient's right to effective treatment. Am J Psychiatry 1990; 147: 409-418[Abstract/Free Full Text].
14.	Andrews G. Treatment outlines for Australian psychiatry. Epidemiol Psichiatria Sociale 1993; 2: 161-164.
15.	American Psychiatric Association Work Group on Panic Disorder. Practice guidelines for the treatment of patients with panic disorder. Am J Psychiatry 1998; 155(suppl): 1-34[Free Full Text].
16.	National Health Committee (New Zealand). Guidelines for assessing and treating anxiety disorders. In: Wellington, NZ: NHC , 1998.
17.	Practice Guideline Coalition. http://www.unr.edu/psych/pgc/shorthis.html 1999. (Accessed 30 January 1999.)
18.	Cochrane AL. Effectiveness and efficiency: random reflections on health services. London: BMJ and the Nuffield Provincial Hospitals Trust, 1989.
19.	Wells KB. Treatment research at the crossroads: the scientific interface of clinical trials and effectiveness research. Am J Psychiatry 1999; 156: 5-10[Abstract/Free Full Text].

(Accepted 30 April 1999)