Education and debate

All you need is cognitive behaviour therapy?

Commentary: Benevolent scepticism is just what the doctor ordered

Commentary: Yes, cognitive behaviour therapy may well be all you need

Commentary: Symptoms or relationships

Commentary: The "evidence" is weaker than claimed

All you need is cognitive behaviour therapy?

Jeremy Holmes, consultant psychiatrist in psychotherapy. 

Department of Mental Health, University of Exeter, North Devon District Hospital, Barnstaple EX31 4JB

j.a.holmes{at}btinternet.com

Psychotherapy, traditionally psychiatry's Cinderella treatment, has finally reached the consciousness of mental health policy makers. The trend started with the 1996 NHS Strategic Review, Psychotherapy Services in England.¹ This set out a programme for coordinated, evidence based, comprehensive, safe, and equitable provision of psychotherapyand pointed to the gap between these ideals and current reality. A sister publication, What Works for Whom? summarised the evidence for psychotherapy "best buys" in all the major psychiatric diagnoses.² Then came the Mental Health National Service Framework and the National Plan, which emphasised psychological therapies as equal players alongside physical and social measures in the management and prevention of mental illness. ^{3 4} Most recently, the Department of Health's Treatment Choice in Psychological Therapies and Counselling provides an evidence based guideline to help family doctors and psychotherapists allocate common mental disorders to appropriate psychological therapies.⁵

	Cognitive behaviour therapy as treatment of choice?

In each of these publications due homage is paid to psychotherapy as a multifaceted, pluralistic enterprise in which a range of therapies is required to meet patients' various needs. Yet, when detailed recommendations are examined there is no doubt that cognitive behaviour therapy is promoted as the therapy of choice. Thus the national service framework cites cognitive behaviour therapy as the first line treatment for depression, eating disorders, panic disorder, obsessive-compulsive disorder, and deliberate self harm. This follows from the framework's practice of classifying quality of evidence. For most diagnoses, cognitive behaviour therapy tends to get the accolade of "level 1" evidenceat least one randomised controlled trial and one good systematic review. Other therapies achieve honourable mentions, but usually as also rans.

A similar theme emerges in the Department of Health's guidelines: cognitive behaviour therapy comes first for depressive disorders, panic disorder, agoraphobia, generalised anxiety disorder, post-traumatic stress disorder, bulimia, and chronic fatigue.⁵ It seems that the traditional "Dodo bird verdict" for psychotherapy research"everyone has won, and all must have prizes"⁶has finally been superseded.

What are analytic, systemic, eclectic, or pluralistically minded therapists to make of this? Should they abandon hope, and immediately devote their continuing professional development time to retrain in cognitive behaviour therapy? Or, like the late Douglas Adams' Arthur Dent when faced with the imminent destruction of his planet,⁷ is it still appropriate to say "Don't panic"?

Cognitive behaviour therapy undoubtedly has much in its favour. It is an attractive, efficient therapy that is relatively easy to learn and deliver and produces good results in many instances. In addition, cognitive behaviour therapy researchers have set standards in detailed descriptions of their methods ("manualisation"), monitoring of adherence, and tailoring treatments to specific disorders that have had a major impact on psychotherapy practice and research generally. Psychoanalytic resistance to quantitative investigation, and consequent marginalisation in an increasingly evidence based world, has been successfully challenged. Psychoanalytic and systemic therapists now recognise the importance of high quality research, including randomised controlled trials, and many investigations are under way or near completion⁸ that would have been unthinkable a decade ago. Cognitive behaviour therapy is the therapy to beat, and this has sharpened the minds of psychotherapy researchers worldwide.

	Limitations of cognitive behaviour therapy

When it comes to making mental health policy, however, several aspects of cognitive behaviour therapy are open to question. Firstly, the foundations on which it rests are not as secure as some of its proponents would have us believe. The National Institute of Mental Health study of depression, the largest of its kind in the world, is now 20 years old, although its findings are still being digested.⁹ In this study, cognitive behaviour therapy fared less well than the two other main treatment arms, interpersonal therapy and clinical management plus antidepressants.

Secondly, there is still much to learn about the impact of different psychotherapies, including cognitive behaviour therapy, on the long term course of psychiatric illnesses. Thus, depression is increasingly seen as a relapsing chronic illness, and without long term comparative follow up studies it is surely premature to champion any one therapy.

(Credit: ULRIKE PREUSS) Cognitive behavioural therapy is the therapy to beat for depression, panic disorder, agoraphobia . . .

Thirdly, there is continuing uncertainty about the effectiveness of different psychotherapies (that is, their clinical relevance) as opposed to their efficacy (ability to produce change under "laboratory" conditions). Cognitive behaviour therapy works well in university based clinical trials with subjects recruited from advertisements, but the evidence about how effective it can be in the real world of clinical practice is less secure. In the London depression trial, for example, couple therapy performed better than antidepressants for treating severe depression in patients living with partners, but cognitive behaviour therapy came nowhere, having been discontinued early in the trial because of poor compliance from a particularly problematic (but clinically typical) group of patients.¹⁰

Fourthly, as the Department of Health's guidelines suggest,⁵ absence of evidence is not the same as evidence of absence. Most studies show absolute rather than relative efficacythat is, cognitive behaviour therapy is usually compared with waiting list controls, no therapy at all, or some sort of bland pseudotherapy rather than with another form of psychotherapy. As in drug trials, comparing good treatments with those that may be better is a much greater research challenge that demonstrating that a treatment is better than nothing.

Finally, and perhaps most important, there are signs that leading cognitive behaviour therapists themselves are starting to question aspects of their discipline and recognise some of its limitations. Linehan argues that standard cognitive behaviour therapy for patients with conditions as complex as borderline personality disorder is unlikely to be effective.¹¹ Her integrative therapy, dialectical behaviour therapy, combines acceptance and acknowledgement of defences (a psychoanalytic idea laced with Zen Buddhism) with cognitive and behavioural techniques for change. Similarly, Teasdale questions the "zap the negative cognitions" approach in major depressive disorder, believing that "mindfulness techniques" such as meditation are also needed to help patients divorce themselves from their emotional pain.¹² In the treatment of personality disorder Young argues for a "schema-based" approach,¹³ taking account of transference, which looks increasingly psychoanalytic in flavour.¹⁴ We are entering a "post-cognitive behaviour therapy" world, which goes beyond brand name therapies to considering the active ingredients of therapy, specific competencies and techniques, and the similarities and differences between different approaches at both theoretical and practical levels.¹⁵

In sum, it is hard to escape the suspicion that cognitive behaviour therapy seems so far ahead of the field in part because of its research and marketing strategy rather than because it is intrinsically superior to other therapies.

	Cognitive behaviour therapy in primary care

If this is so, it puts further pressure on psychoanalytic therapy, counselling, and systemic therapies to prove their worth. Recent studies comparing cognitive behaviour therapy with counselling in primary care showed no significant differences in outcomes.¹⁶ Such results show the dangers of tying government policies too closely to specific research findings. An earlier review had suggested that counselling was ineffective.¹⁷ That conclusion is now clearly open to question, but health authorities and commissioners or health maintenance organisations react slowly and are unlikely to follow the latest psychotherapy research literature, despite the updating mechanisms built into the national service frameworks. The situation is analogous to that with monetarism as part of previous governments' economic policies. Both cognitive behaviour therapy and monetarism could be seen as the medicine needed to sweep aside complacency and old fashioned practice, especially where resources are limited. But, just as Western countries have moved on from monetarism to a more mixed economy while the World Bank continues to insist on outdated monetarist practice in the developing world,¹⁸ so there is a danger of imposing cognitive behaviour therapy in general practice just as its therapists in secondary care are moving beyond it.

	Beyond psychotherapy brand names

The drug treatment paradigm has enormous power in medicine. Research in psychological therapies, especially cognitive behaviour therapy, has been shaped by the "drug metaphor."¹⁹ This implies specific treatments for specific conditionsand that anything that lies outside the paradigm lies beyond the bounds of science. This approach has been beneficial in that it has forced psychotherapies to submit themselves to randomised controlled trials if they are to claim scientific credibility, and means that consumers and practitioners can be secure in the knowledge that their therapies are of proved worth.

However, the drug metaphor has also had a distorting and potentially damaging effect in psychological medicine. Psychotherapy is essentially concerned with people, not conditions or disorders, and its methods arise out of an intimate relationship between two people that cannot easily be reduced to a set of prescribed techniques. One of the most robust findings in psychotherapy research is that a good therapeutic alliance is the best predictor of outcome in psychotherapy,²⁰ which suggests that specificity needs to be sought at a much deeper level than therapy brand names. Indeed, it may well be that cognitive behaviour therapywith its clear structure, optimistic outlook, and active involvement of the patientis successful precisely because of its power to create a good alliance.

Another negative aspect of the drug metaphor is the neglect of a developmental perspective in psychiatry. Psychiatry is not just about patients' disorders, but people who are on a developmental trajectory. Past experience shapes present reactions in ways that cannot be captured simply by the Diagnostic and Statistical Manual notion of comorbidity between "Axis 1 disorders" (illnesses like depression) and "Axis 11 disorders" (personality).²¹ For example, there is good evidence that insecure attachment patterns in childhood act as vulnerability factors for adult psychiatric illness and, in particular, that patients with borderline personality disorder are likely to have had disorganised attachment patterns.²² We need a psychotherapy that is sophisticated enough to take the development of the mind into account and to capture the often unexpected "emergent meanings" that arise in therapythe antithesis of the predetermined narrowly technical approach with which cognitive behaviour therapy is sometimes identified.

Even if the drug analogy is to some extent inescapable, there are still grounds for questioning the current undue emphasis on cognitive behaviour therapy. Just as no one antipsychotic or antidepressant drug can cure all ills, so a wide range of psychological therapies are needed if we are to meet the variety of psychiatric illnesses and human developmental experience. The current apparent triumph of cognitive behaviour therapy harks back to the ideological divide between behaviourism and psychoanalysis in the 1920s. Patients in the 21st century deserve therapies that transcend old rivalries and concentrate on effectiveness, common factors, the search for active ingredients that go beyond brand names, and development of the skills needed to deliver them.

	Acknowledgments

I thank Anthony Bateman, Else Guthrie, Iona Heath, Peter Hobson, and Paul Salkovskis for their help with the preparation of this paper.

	Footnotes

Competing interests: None declared.

	References

1.	NHS Executive. Psychotherapy services in England. London: HMSO, 1996.
2.	Roth A, Fonagy P. What works for whom? New York: Guilford, 1996.
3.	Department of Health. National service framework for mental health. London: HMSO, 2000.
4.	Department of Health. The national plan. London: HMSO, 2000.
5.	Department of Health. Treatment choice in psychological therapies and counselling. London: HMSO, 2001.
6.	Luborsky L, Singer B, Luborsky B. Comparative studies of psychotherapies: is it true that `everyone has won and all must have prizes'? Arch Gen Psychiatry 1975; 32: 995-1008[Abstract].
7.	Adams D. Hitchhiker's guide to the galaxy. London: Methuen, 1978.
8.	Fonagy P, ed. An open door review of outcome studies in psychoanalysis. London: International Psychoanalytic Association, 1999.
9.	Elkin I. The NUMH treatment of depression collaborative research programme; where we began and where are we? In: Bergin A, Garfield S, eds. Handbook of psychotherapy and behaviour change. Chichester: Wiley, 1994:114-142.
10.	Leff J, Vearnals S, Wolff G. The London depression intervention trial: randomised controlled trial of antidepressants v. couple therapy in the treatment and maintenance of people with depression living with a partner. Br J Psychiatry 2000; 177: 95-100[Abstract/Free Full Text].
11.	Linehan M. Cognitive-behavioural treatment of borderline personality disorder. New York: Guilford, 1993.
12.	Teasdale J. Prevention of relapse/recurrence in manic depressive psychosis by mindfulness-based cognitive therapy. J Counsel Clin Psychol 2000; 68: 615-621.
13.	Young J. Cognitive-behaviour therapy for personality disorders: a schema-focussed approach. Sarasota, FL: Professional Resource Exchange, 1990.
14.	Bateman A. Integration in psychotherapy: an evolving reality in personality disorder. Br J Psychotherapy 2000; 17: 147-156.
15.	Holmes J, Bateman A. Psychotherapy integration: models and methods. Oxford: Oxford University Press (in press).
16.	Hemmings A. Counselling in primary care: a review of the practice evidence. Br J Guidance Counsel 2000; 28: 233-252[CrossRef].
17.	Freidli K, King M. Counselling in general practicea review. Primary Care Psychiatry 1996; 2: 205-216.
18.	Hutton W. The state we're in. London: Verso, 1996.
19.	Shapiro D. Finding out about how psychotherapies help people change. Psychother Res 1996; 5: 1-21[ISI].
20.	Horvath A, Symonds D. Relationship between working alliance and outcome in psychotherapy: a meta-analysis. J Counselling Psychology 1991; 38: 139-149.
21.	American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington DC: APA, 1994.
22.	Holmes J. The search for the secure base: attachment theory and psychotherapy. London: Routledge, 2001.

(Accepted 28 August 2001)

Commentary: Benevolent scepticism is just what the doctor ordered

Roger Neighbour, general practitioner. 

Vine House Health Centre, Abbots Langley WD5 0AJ

roger.neighbour{at}dial.pipex.com

To open the BMJ is, for a general practitioner, often to find yourself the target of a batch of rather bossy papers competing for the right to tell you what to do. It makes me feel like some Arthurian princess watching, with dogged interest and an ill concealed yawn, a succession of knights strut their stuff in the joust. With cries of "Evidence rules OK," the Lancelots and Galahads of the research world tilt at each other with their controlled trials until a winner emerges to claim the lady's hand. "All this shouting and rushing around is very flattering," the princess might murmur, "but is it any basis for a lasting marriage?"

It is generally wise to keep such unworthy thoughts private. In public, we general practitioners are expected to fall like terriers on every latest pronouncement from our betters and implement it gratefullya case of the double blind leading the blind. A favourite question in the MRCGP examination is to present candidates with a recent research paper claiming some new treatment is a marginal advance on the old and to ask them, "What would be the implications for your practice of adopting these conclusions?" The platitudes flow: team meetings, educational programmes, audit cycles. The right answer, of course, is, "Pandemonium."

I confess my hackles are quick to rise when I see anything flaunting the slogan, "As endorsed by evidence based medicine." It's not that I'm a Luddite. Far from it; of course our practice should reflect and incorporate the best available scientific knowledge. But, where knowledge is concerned, especially in the slippery arena of psychotherapy, newest isn't always truest. Evidence that looks compelling from one viewpoint is often less impressive from another. Too many researchers, in their rush to publish, haven't stopped to ask themselves what good and useful evidence would actually look like. I'm pleased to see how well Jeremy Holmes understands this.

Let us allow that, on a scale of effectiveness from "rubbish" to "pure gold," cognitive behaviour therapy is well towards the latter. It has a rational basis in applied learning theory and is not that difficult to carry out. With its relatively short time scale and the systematic structure of its interventions, it lends itself to, and does relatively well in, controlled trials against its more protracted or idiosyncratic rival therapies. It is cheap and often cheerful, in the sense of not insisting on a cathartic show of misery as part of the therapeutic ritual. Patients often like it. And you don't have to be a doctor to do it. One can readily see why it would suit our health politicians for cognitive behaviour therapy to become the default psychotherapy for whatever unhappiness can't be coped with by a patient centred general practitioner and a course of Prozac.

Holmes' evaluation is balanced and sensible. But his piece has reminded me of all the other instances, less obvious and therefore more worrying, where poor research is used to twist the arm of general practice into naively accepting politically or financially convenient oversimplifications.

We need to remember that research is no more than the amalgamation of numerous anecdotesindividual human stories with the detail deleted. But for general practitioners, the individual differences are where the interest and the value lie. Every treatment has its "successes" (which live on in the memory and publications of therapists) and its "failures" (who come back to consult us). The difference in outcome between the best and the worst therapists within a discipline is often greater than the mean difference between disciplines.

What general practitioners should take from Holmes' article is his attitude of benevolent scepticism. We should apply it to all the quick fix, one-size-fits-all proposals handed down in the name of evidence based medicine. As someone said, "To a man with a hammer, an awful lot of things look like nails."

Commentary: Yes, cognitive behaviour therapy may well be all you need

Nicholas Tarrier, professor of clinical psychology. 

Academic Division of Clinical Psychology, School of Psychiatry and Behavioural Science, University of Manchester, Wythenshawe Hospital, Manchester M23 9LT

ntarrier{at}man.ac.uk

Holmes bemoans the fact that the evidence for the efficacy of cognitive behaviour therapy is considerable whereas there is little evidence for the more traditional psychotherapies. Because of the evidence base for it, cognitive behaviour therapy has advanced as the treatment of choice, leaving traditional psychotherapies as "also rans." Not surprisingly, as a traditional psychotherapist, he is unhappy with this situation. Well, he is unhappy that those who make policy in the NHS have become wise to this and may well act on it. Do psychotherapists abandon hope or panic? Neither, they fight back. Having sounded his wake up call, Holmes leads the charge. He fires his criticisms from the hip with merry abandon: cognitive behaviour therapy has its limitations, the evidence base is less secure than supposed and its value in the real world doubtful, leading proponents of cognitive behaviour therapy are unsettled, cognitive behaviour therapy is simplistic, and so on.

Some of these criticisms are reasonable, some are not. A clinical trial is considered efficacious when tested under ideal conditions and effective when beneficial under routine circumstances.¹ Holmes concedes that the demonstrated efficacy of cognitive behaviour therapy across a range of disorders is substantial. Not to do so would be difficult.² Perusal through psychology journals such as Behavior Therapy, Behaviour Research and Therapy, and Journal of Consulting and Clinical Psychology, as well as psychiatric and medical journals, will confirm this. There are also good examples of effectiveness in the literature.^3-6 Yes, there is much to do in terms of understanding effectiveness rather than just efficacy, but cognitive behaviour therapy practitioners and researchers are addressing these issues and have the scientific background to do so.

In recent years the area in which cognitive behaviour therapy has made a real impact, and in which Britain has a world lead, is in the treatment of psychoses, ^{7 8} but this is quite specialist work. As an example of the broader applicability of cognitive behaviour therapy, when quite simple behavioural skills were taught to care staff in residential and nursing homes for elderly people there were beneficial effects for the residents compared with the control situation.⁹ The strength of cognitive behaviour therapy is its broad application in many settings, not just psychiatric, and with many different client groups and at different levels of expertise.

Holmes relies on the specious old adage that absence of evidence is not evidence of absence. Absence of evidence is absence of evidence. It may be theoretically accurate that many things could be efficacious, including psychotherapy, but I would have more enthusiasm for this argument if traditional psychotherapy were new. It has been around for 100 years or so. The argument, therefore, becomes a little less compelling when psychotherapy's late arrival at the table of science has been triggered by a threat to pull the plug on public funding because of the absence of evidence.

Holmes tells us that relationships are important. Of course they are. Humans have evolved as social animals, and they exist in complex social systems in which social relationships and ties are paramount and intimately linked to mental health.¹⁰ We don't need 100 years of psychotherapy to tell us that. Many of those who suffer mental distress and disorder have disrupted social lives and restricted social networks¹¹ and would benefit from having stable relationships.¹² It is improbable that traditional psychotherapyexpensive, time consuming, and unproved as it isis either the only or best way to achieve this.

Cognitive behaviour therapy is not just about psychiatry nor is it just a set of atheoretical techniques.¹³ It has a strong theoretical base in the discipline of psychology, which informs treatment development through a scientific understanding of both normal and abnormal mental functioning. Most of its proponents are not medical. Of the 4156 members of the British Association for Behavioural and Cognitive Psychotherapy, the relevant professional organisation, 41% are psychologists and 28% are nurses while only 6% are psychiatrists (personal communication, H Lomas, membership secretary). Thus, most cognitive behaviour therapy theorists and practitioners come from outside the profession of psychiatry or medicine.

Furthermore, cognitive behaviour therapy is collaborativethe patient is an equal, and information is shared. There is a telling quote from the BMJ 's Editor's choice last September: "Doctors can't just declare things to be true, they must produce evidence and they must share their knowledge with their patients."¹⁴ Little more needs to be said.

Much of mental distress no doubt has its roots in, or is at least exacerbated by, social deprivation and inequality and their psychological consequences.¹⁵ A good dose of social justice and redistribution of wealth would do the world's health a lot of good. In the meantime, any psychological treatment can only be a sticking plaster over the wound of such inequality, but, as far as evidence goes, cognitive behaviour therapy is the best plaster available.

	References

1.	Fletcher RH, Fletcher SW, Wagner EHW. Clinical epidemiology: the essentials. London: Williams and Wilkins, 1988.
2.	Clark DM, Fairburn CG. Science and practice of cognitive behaviour therapy. Oxford: Oxford University Press, 1997.
3.	Barrowclough C, Tarrier N, Sellwood W, Quinn J, Mainwaring J, Lewis S. A randomised controlled effectiveness trial of a needs based psychosocial intervention service for carers of schizophrenic patients. Br J Psychiatry 1999; 174: 506-511.
4.	Stuart GL, Treat TA, Wade WA. Effectiveness based treatment for panic disorder delivered in a service clinic setting. J Consult Clin Psychol 2000; 68: 371-377[CrossRef][ISI][Medline].
5.	Wade WA, Treat TT, Stuart GL. Transporting an empirically supported treatment for panic disorder to a service clinic setting: a benchmarking strategy. J Consult Clin Psychol 1998; 66: 231-239[CrossRef][ISI][Medline].
6.	Wiesrma D, Jenner JA, van de Willige G, Spakman M, Nienhuis FJ. Cognitive behaviour therapy with coping training for persistent auditory hallucinations in schizophrenia: a naturalistic follow-up study of the durability of effects. Acta Psychiatrica Scand 2001; 103: 393-399[CrossRef][ISI][Medline].
7.	Gould RA, Mueser KT, Bolton E, Mays V, Goff D. Cognitive therapy for psychosis in schizophrenia: an effect size analysis. Schizophrenia Res 2001; 48: 335-342[CrossRef][ISI][Medline].
8.	Rector NA, Beck AT. Cognitive behavioral therapy for schizophrenia: An empirical review. J Nervous Mental Dis 2001; 189: 278-287[CrossRef][ISI][Medline].
9.	Proctor R, Burns A, Stratton-Powell H, Tarrier N, Faragher B, Richardson G, et al. Behavioural management in nursing and residential homes: a randomised controlled trial. Lancet 1999; 354: 26-29[CrossRef][ISI][Medline].
10.	Sloman L, Gilbert P, eds. Subordination and defeat: an evolutionary approach to mood disorders and their therapy. New Jersey: Lawrence Erlbaum, 2000.
11.	Davidson L, Stayner D, Haglund KE. Phenomenological perspectives on the social functioning of people with schizophrenia. In: Mueser K, Tarrier N, eds. Handbook of social functioning in schizophrenia. Needham Heights, MA: Allyn and Baker, 1998.
12.	Tarrier N, Kinney C, McCarthy E, Humphreys L, Wittkowski A, Morris J. Two year follow-up of cognitive-behaviour therapy and supportive counselling in the treatment of persistent positive symptoms in chronic schizophrenia. J Consult Clin Psychol 2000; 68: 917-922[CrossRef][ISI][Medline].
13.	Tarrier N, Wells A, Haddock G, eds. Cognitive behaviour therapy for complex cases: an advanced guidebook for the practitioner. Chichester: Wiley, 1998.
14.	Smith R. Editor's choice: Doctors: the long march to accountability. BMJ 2000;321(9 Sep).
15.	Brown G, Harris T. The social origins of depression. London: Tavistock Publications, 1978.

Commentary: Symptoms or relationships

R D Hinshelwood, consultant psychotherapist. 

Cassel Hospital, Richmond TW10 7JF

bob{at}hinsh.freeserve.co.uk

Medical services assume that evidence for the profitable use of psychotherapy will flow from the standard drug trial model, the randomised controlled trial. Because of the emphasis on symptom change, certain psychotherapies, like cognitive behaviour therapy, "fit" that model, whereas others that are relationship based, like psychoanalytic psychotherapy, do not. The randomised controlled trial is almost completely helpless to assess relationship change.

For example, with severe personality disorder, a condition resistant to cognitive behaviour therapy, the core problem is that the person engages in abusive relationshipstypically, the relationship with care. Then cognitive behaviour therapy, which does not address relationship change, matches the randomised controlled trial paradigm, which cannot assess it. The focus on symptoms in cognitive behaviour therapy and in randomised controlled trials promotes a particular "fit," and Jeremy Holmes rightly argues that this gives cognitive behaviour therapy a head start over other psychotherapies in the race for government accolades. Reductionism to a single paradigm of "evidence" reduces the field to a single treatment not comparable with others. Other psychotherapies, like psychoanalytic psychotherapy, focus on changes in relationships and cannot be assessed in the same way. So, in psychotherapy research we compare two quite different treatment modalities. Holmes argues it is not enough to measure outcome of treatment against no treatment, yet this is often what happens when no real means of comparing treatment with treatment exists. There is no reason why psychoanalytic psychotherapy should not run in the race, but it does so with a handicap; the race is "fixed" for those therapies that focus on symptom change.

	Placebo or transference

When psychotherapy research addresses relationships it is in difficulty. In randomised controlled trials a patient's relationship with a doctor is viewed as suspectthe professional relationship is regarded simply as a placebo effect. Such trials were designed specifically to control out the effects of the relationship. However, the psychotherapies, adopting a relational approach to personal change, exploit precisely that placebo effect, as "treatment alliance" and "transference." Cognitive behaviour therapy is particularly dependent on compliance, and the method wins patients' cooperation. It is ineffective in non-compliant patients. Psychoanalytic psychotherapy, which is compliance-neutral therapy, specifically focuses on the destructive attitudes towards care and focuses treatment on the nature and extent of compliance itself (transference). Because the randomised controlled trial method was developed to eliminate the placebo effect, is it appropriate for assessing cognitive behaviour therapy, which must rely on a treatment alliance, or psychoanalytic psychotherapy, which investigates that alliance?

(Credit: WILL AND DENI MCINTYRE) A psychotherapeutic relationship between two people cannot easily be reduced to prescribed techniques

	Social construction

The relational context of psychotherapy has wider implications for assessment. Symptoms and their derivatives, syndromes, are in many cases socially constructed.¹ Socially expected categories of ill health, together with a negotiation between patient and doctor (or therapist), "construct" the diagnosis. However, in general medicine there is usually a failsafe mechanism in biological testing through objective physiological and anatomical findings. But no similar failsafe exists for problems of the mind. Objectifying psychiatric symptoms and diagnoses takes place inevitably in a framework made up of social attitudes held by doctors, patients, and society at large.² Psychiatric diagnoses are particularly prone to "fashion" in different historical periods, when society "constructs" certain symptoms and syndromes. For example, interest in, and diagnosis of, multiple personality disorder peaked in the 1890s and again in the 1990s. "Cures" are also socially constructed from the same ingredientssocial expectations and the care relationship. A cure based on such "subjective" ingredients is no less a cure, and objectifying such changes may be possible. But this challenges the view that objective phenomena exist in a way that transcends social and personal attitudesand it undermines those objective methods that ignore the relational framework of social attitudes.

	Assessing relational change

The real challenge in psychotherapy research is to develop the assessment of relational change. At present relational change is largely assessed intuitively by therapists. Recently, however, the adult attachment interview has been developed to objectify relational phenomena.³ This has the advantage of producing assessments of change within generalised categories, although that must be weighed against the loss of individuality. However, the technique is costly, both to train people and to carry out assessments. It is hardly for routine use but is an increasingly useful research tool. Other methods have failings too. The now defunct object relations test⁴ was problematic because it depended on intuitive assessment by the psychologist administering the test. The repertory grid ^{5 6} is potentially a sensitive method of objectifying a person's relationships and any change in them but is statistically complex. It respects the individuality of a person and his or her relationships, but such individuality makes it very difficult to compare the changes effected by different therapies, and by a therapy in different patients.

Currently, there is no method of assessing relationship change that compares with methods for assessing symptom change. Any that are developed will not have congruence with the traditional assumptions of medical research, embodied in the drug trial method. Some forms of relational therapies may gain a credence from a spurious conformity to the symptom change paradigm. Currently, the Henderson therapeutic community has received extensive government financing for expanding its work, and Grendon Prison is trying to "manualise" its work. Both organisations use reoffending rates to stand in for symptom change. Without a valid testing method that takes account of the psychosocial dimension of relationships and social attitudes, relational psychotherapies may in the long term be doomed to exclusion from public service funding, to the serious detriment of patient choice.

My comment on Holmes' critique of evidence has tried to place the debate in a context that is beyond the usual framework. The philosophical colouring may make pragmatists suspicious. However, practitioners addressing underlying factorswhether inappropriate cognitive ideation or the psychoanalytic unconsciousshould be prepared to accept a moment or two of reflection on factors that underlie our therapies and the comparisons between them.

	References

1.	Figlio K. How does illness mediate social relations? Workmen's compensation and medical-legal practices, 1890-1940. In: Treacher A, Wright P, eds. The problem of medical knowledge: examining the social construction of medicine. Edinburgh: Edinburgh University Press, 1982:174-224.
2.	Hinshelwood RD. The difficult patient: the role of `scientific' psychiatry in understanding patients with chronic schizophrenia or severe personality disorder. Br J Psychiatry 1999; 174: 187-190[Free Full Text].
3.	Main M. Discourse, prediction and recent studies in attachment: implications for psychoanalysis. J Am Psychoanal Assoc Suppl 1993; 41: 209-244.
4.	Laing RD, Philipson H, Lee AR. Interpersonal perception: a theory and a method of research. London: Tavistock, 1966.
5.	Kelly G. The psychology of personal constructs. New York: Norton, 1955.
6.	Bannister D. Inquiring man: the psychology of personal constructs. London: Penguin, 1980.

Commentary: The "evidence" is weaker than claimed

Nick Bolsover, consultant clinical psychologist in psychotherapy. 

Clinical Psychology Services, Hull and East Riding Community Health NHS Trust, Victoria House, Hull HU2 8TD

Holmes eloquently describes major issues in current psychotherapy research and practice. He is right to suspect cognitive behaviour therapy of appearing ahead of the field thanks to its research and marketing strategy rather than any intrinsic superiority, and he could have challenged the research evidence more forcefully.

The hypothesis that people (and their thoughts, emotions, and behaviour) can be helped to change, in a meaningful sense, in a few sessions over a few months seems improbable, although it is not new. Freud regretted the widespread hope for brief treatments for mental disorders: "Doctors lend support to these fond hopes. Even the informed among them fail to estimate properly the severity of nervous disorders. A friend and colleague of mine . . . once wrote to me: `What we need is a short, convenient, out-patient treatment for obsessional neurosis.' I could not supply him with it and felt ashamed; so I tried to excuse myself with the remark that specialists in internal diseases, too, would probably be very glad of a treatment for tuberculosis or carcinoma which combined these advantages."¹

The evidence for effective brief therapies is not as strong as has been claimed, especially if you consider that many effects reported at the 5% level of significance will be false positives. For example, a report has claimed cognitive behaviour therapy to be effective in treating people with psychotic experiences.² Much of the cited evidence reported effects at the 5% significance level. The randomised controlled trial, which presented the more robust data, was not assessed "blind" and did not include a no treatment control group.

The evidence for the claimed pre-eminence of cognitive behaviour therapy in treating less severe mental health problems is similarly defective. One study gave from six to a maximum of 12 sessions of psychological therapy, either cognitive behaviour therapy or non-directive counselling, or usual general practitioner treatment.³ Both psychological interventions gave better results (on the Beck depression inventory but not on three other outcome measures) than usual general practitioner care at four months, but this advantage was not maintained at 12 months.

The difficulties that cognitive behaviour therapy encountered in another study⁴ (in which the cognitive behaviour therapy arm was abandoned after eight of 11 couples failed to attend therapy sessions) point to the complexities of conditions such as depression and the difficulties in providing effective therapy. The apparent brevity and simplicity of cognitive behaviour therapy seems not to acknowledge fully this complexity.

There are dangers in a therapy being presented in a more positive light than the data from clinical trials merit. For example, acronyms for clinical trials that suggest therapeutic benefits seem to have more influence than their actual results on prescribing habits.⁵

In addition to the lack of evidence for brief cognitive behaviour therapy having a clinically significant effect, the claim that such studies are randomised controlled trials does not stand up to scrutiny. They are not, and it is extremely difficult, perhaps impossible, to meet the criteria for randomised controlled trials in research into psychotherapy. Declaring that these flawed studies meet the "gold standard" of clinical research is a marketing rather than a research strategy. Taken with Holmes' points on the inappropriateness of the standard drug trial model, there is a case for redirecting the resources currently allocated to randomised controlled trials into devising more creative research designs in psychotherapy.⁶

Furthermore, there is robust evidence that longer therapies give better outcomes, regardless of treatment method,⁶ which has largely been ignored (most recently by Holmes). Emerging results support this view, showing that, for example, psychotherapy and psychoanalysis both have beneficial effects that increase with time in treatment.⁷ Such results seem to have had little impact on researchers and health providers in Britain, although in the United States substantial federal research funds have been allocated as a result.

Until the limitations of the evidence base for cognitive behaviour therapy are recognised, there is a risk that psychological treatments in the NHS will be guided by research that is not relevant to actual clinical practice and is less robust than claimed.

	References

1.	Freud S. On beginning the treatment. In: The standard edition of the complete psychological works of Sigmund Freud. vol 12. London: Hogarth Press, 1924:123-143.
2.	Kinderman P, Cooke A. Understanding mental illness: Recent advances in understanding mental illness and psychotic experiences. Leicester: British Psychological Society, 2000.
3.	Ward E, King M, Lloyd M, Bower P, Sibbald B, Farrelly S, et al. Randomised controlled trial of non-directive counselling, cognitive-behaviour therapy and usual general practitioner care for patients with depression. I: Clinical effectiveness. BMJ 2000; 321: 1383-1388[Abstract/Free Full Text].
4.	Leff J, Vearnals S, Brewin C, Wolff G, Alexander B, Asen E, et al. The London depression intervention trial. Br J Psychiatry 2001; 177: 95-100.
5.	Goodman B. Acronym acrimony. Scientific American 2001; 285(5): 16.
6.	Seligman M. The effectiveness of psychotherapy. The consumer reports study. Am Psychol 1995; 50: 965-974[CrossRef][Medline].
7.	Sandell R, Blomberg J, Lazar A, Carlsson J, Broberg J, Schubert J. Varieties of long-term outcome among patients in psychoanalysis and long-term psychotherapy: a review of findings in the Stockholm outcome of psychoanalysis and psychotherapy project (STOPPP). Int J Psychoanal 2000; 81: 921-942.