Which therapies produce the best outcomes? Are directive practices more effective than non-directive ones? What is the relative contribution of the therapist and the client to the outcomes of therapy? There are increasing demands for counsellors and psychotherapists to answer questions such as these and to base their practice on a comprehensive body of research knowledge. For instance, in its Standards of Education and Training, the UK’s Health Professions Council, which looks set to regulate counselling and psychotherapy, stipulates that courses will need to assist students in the development of evidencebased practice¹. And the recently launched Improving Access to Psychological Therapies programme, responsible for the allocation of £173 million to the development of mental health services in the UK, will focus exclusively on those therapies for which there is clear evidence of effectiveness through randomised clinical trials².

For counsellors and psychotherapists, then, it is becoming less and less sufficient to justify practice on the grounds that, ‘I know that what I do works.’ And, to some extent, why should it be? Snake oil salespeople and advocates of now-abandoned treatments, such as insulin coma therapy for schizophrenia, would claim much the same thing.

The reality is, therapists do get it wrong. For instance, ninety per cent of therapists put themselves in the top 25 per cent in terms of service delivery³. So therapists’ perceptions, experiences or beliefs that their therapies are effective do not necessarily make them so. Even direct positive feedback from clients has its limitations: research into the phenomenon of client deference shows that clients will often withhold more critical or negative judgements from their therapists, such that clients who have not found their therapy helpful may still tell their therapists how useful it was⁴.

The good news for counsellors and psychotherapists is that a vast body of empirical evidence does exist to support their practice – and not just CBT, but a wide array of psychological practices. The bad news is that very few therapists are actually aware of this evidence, or draw on it to develop their work.

One survey of American psychotherapists found that only four per cent ranked research literature as the most useful source of information on how to practise; with 48 per cent giving top ranking to ‘ongoing experiences with clients’; 10 per cent ranking theoretical literature as the most useful source; and eight per cent ranking their own experiences as clients most highly⁵.

One reason why counsellors and psychotherapists seem not to draw on the research evidence is because it is seldom communicated in a ‘clear and relevant fashion⁵’. Hence there is a need for texts that can communicate research findings in an accessible and jargon-free way. In 2005 BACP funded me to write such a book – Essential Research Findings in Counselling and Psychotherapy: The Facts are Friendly. This article summarises some of its key findings.

Before beginning, however, it is worth saying a few words about my own background. Clearly, it is not possible to present an entirely objective summary of the data, and knowing where an author ‘comes from’ can help to identify potential biases and omissions in any account.

In recent years I have come to see that the touchstone for my therapeutic work is a progressive political outlook⁶, and I am particularly drawn towards those therapies that advocate a relatively egalitarian client-therapist relationship, such as personcentred and existential therapies. At the same time, coming from a position that wants to emphasise the uniqueness of each individual client, I believe passionately that there is no one ‘best’ therapy for everyone⁷.

To some extent, this pluralistic bias also comes from my own experiences as a client, where I have experienced, and found helpful, a wide range of different therapies, including behavioural, person-centred and psychodynamic. Finally, with respect to research, I see it as having a valuable role to play in informing, challenging and stimulating therapists, because it can give voice to clients’ experiences, preferences and needs. That is not to say that engaging with the research findings is always comfortable or reassuring. But, as Carl Rogers wrote over 50 years ago, ‘the facts are always friendly,’ for ‘Every bit of evidence one can acquire, in any area, leads one that much closer to what is true8.’

Overall effectiveness
Does therapy work? Fortunately there is a simple answer to this question: yes. Studies which look at clients’ behaviours, feelings or psychological functioning before and after therapy nearly always find that, on average, they are better off by the end of it. For instance, one study found that prior to a course of family therapy, clients with anorexia nervosa weighed, on average, 40kg. After 12 months of therapy they weighed, on average, 48.2kg⁹.

Of course, the problem with such evidence is that we do not know whether the clients would have got better anyway without therapy, and so a more rigorous test comes from conducting controlled trials in which changes over a course of therapy are compared with changes over a similar period of time for people who do not have any therapy. What does the research show here? Pretty much the same thing. For instance, King and colleagues found that depressed clients who received non-directive therapy dropped 13.9 points on a measure of depression, while individuals who received treatment as usual from their GPs dropped by only 9.3 points¹⁰.

Combining findings from a wide range of controlled trials, meta-analytic studies have shown that, on average, counselling and psychotherapy has a large positive effect¹¹ – greater, indeed, than the average surgical or medical procedure¹². Put more precisely, research shows that approximately 80 per cent of people will do better after therapy than the average person who has not had therapy.

To illustrate, imagine ‘Frank’ going to his GP with depression and being encouraged to wait and see how things improve. Now imagine Frank two months down the line, possibly feeling a little better, but still relatively depressed. And now imagine another 10 people going to their GP, but this time being referred to therapy. So the research shows that in two months time approximately eight of these people will be feeling better than Frank, while two of them will be feeling worse.

In terms of general effectiveness, what we also know from the research evidence is that:
■ improvements in mental health tend to be maintained one or two years after therapy has ended, though the longer-term impact of psychological interventions is less clear¹³;
■ talking therapies are generally as effective as pharmacological treatments for psychological distress, and seem to have lower relapse and drop-out rates¹⁴;
■ counselling and psychotherapy are relatively cost-effective forms of mental health treatment – particularly for more psychologically distressed individuals – with an economic advantage above and beyond their contribution to psychological health and wellbeing¹⁵;
■ approximately five to 10 per cent of clients deteriorate as a result of therapy¹⁶.

Orientation and technique factors
In general, then, therapy is effective. But is this true for all therapies, or are some therapies more effective than others? This is probably the most controversial question in the counselling and psychotherapy research field and, to a great extent, can be answered in very different ways depending on how you read the evidence.

If, on the one hand, you look at the particular therapies that have been shown to be effective for particular psychological problems – as advocates of empirically supported treatments have done – there is no question that the evidence base is strongest for CBT. While, for instance, there are scores of highquality controlled trials demonstrating the effectiveness of CBT for depression¹⁷, there are just a handful of studies demonstrating the same thing for person-centred therapy. And while CBT has been shown to be effective for numerous psychological difficulties – such as phobias, panic, PTSD, bulimia, sexual problems and deliberate selfharm – there is little equivalent evidence for the vast array of non-CBT practices¹⁸.

At the same time, it is essential to note that ‘greater evidence of effectiveness’ is not the same as ‘evidence of greater effectiveness’ and, in the vast majority of instances, the reason why non-CBT therapies are not considered effective for particular psychological problems is simply because no one has studied them yet.

As Westen and colleagues write: ‘Perhaps the best predictors of whether a treatment finds its way to the empirically supported list are whether anyone has been motivated (and funded) to test it and whether it is readily testable in a brief manner¹⁹.’

Indeed, when studies are carried out to compare the effectiveness of different psychological therapies, the almost unanimous finding is that they are of about equivalent – or only marginally different – effectiveness¹¹ (this is particularly the case when the studies are carried out by independent bodies, and when bona fide therapies are compared²⁰). Within the counselling and psychotherapy research literature, this finding has come to be known as the dodo bird verdict – after the dodo bird in Alice in Wonderland who, having judged a race around a lake, declares that ‘Everyone has won and so all must have prizes.’

The dodo bird verdict is a remarkably ubiquitous finding in the counselling and psychotherapy research field.
Not only, for instance, do different therapies seem to be of about equivalent effectiveness, but relatively similar outcomes tend to be found when comparing group therapies against individual therapies²¹; ‘complete’ therapies against specific components of those therapies (for instance, a full CBT package against just its behavioural element²²); professionally-delivered therapies against paraprofessionally-delivered therapies²³; and self-help practices against interpersonal therapies²⁴.

Therapist factors
So if therapeutic orientation is not the principal determinant of whether or not therapies are effective, what is? One possibility, as argued by Scott Miller and colleagues in the April 2008 edition of therapy today²⁵ is that it is something to do with the therapists themselves: their personalities, openness to feedback, or personal and professional experiences. In support of this hypothesis, research shows substantial differences in outcomes across therapists. One study, for instance, found that the clients of the most effective therapist in a university counselling centre showed a rate of improvement 10 times that of the average, while the clients of the least effective therapist showed an average worsening of problems²⁶.

Across a range of studies research suggests that around five to 10 per cent of the variance in outcomes is related to differences across therapists²⁷, and this compares to just one per cent or so attributable to the therapists’ particular orientations11. This means that the differences in effectiveness from one CBT practitioner to another, or from one psychodynamic therapist to another, is considerably greater than the differences in effectiveness between all CBT practitioners and all psychodynamic therapists, or all therapists of any other orientation.

Less clear, however, are the particular therapist characteristics that relate to outcomes. Research suggests, for instance, that the following therapist characteristics are only moderately related to therapists’ effectiveness: personality traits; levels of psychological wellbeing; demographic characteristics (gender, ethnicity, age, sexual orientation); amount of professional training; experience as a therapist; life-experience²⁸.

With respect to therapists’ characteristics, however, one somewhat stronger finding to emerge from the research is that clients from marginalised social groups (such as lesbian clients) and clients with strong values do seem to do better with, stay in therapy for longer with, and express a preference for therapists with matching characteristics. One study found that African American clients averaged 17 sessions with white therapists, compared with 25 sessions with African American therapists²⁹. What the research also indicates, however, is that the key issue here may be a perception that such therapists will be, or are, more accepting, rather than the particular characteristics, per se.

One study³⁰ found that 45 per cent of Orthodox Jewish respondents would prefer to see an Orthodox Jewish therapist, and much of this was to do with fears that non- Orthodox therapists would react negatively to them. One respondent said, ‘Someone not frum [Orthodox] would try to channel me in nonfrum directions.’

Interestingly, however, those respondents who expressed a preference not to see an Orthodox Jewish therapist (20 per cent of respondents) did so for similar reasons – they feared that an Orthodox Jewish therapist would judge or criticise them. For instance: ‘I really chose someone… who was really “off-the-wall” in Judaism, because I kind of felt I could say some of the things I wanted to without any fear of… “How can you even think of such things?!”’

Relational factors
Such evidence suggests that the key contribution that therapists make to the outcomes of therapy may be less to do with who they are and more to do with how they relate to their clients. Consistent with this, Michael Lambert, one of the world’s leading psychotherapy researchers, estimates that the therapeutic relationship accounts for as much as 30 per cent of the variance in outcomes³¹. And, while other reviewers have given more modest figures, such as seven to 17 per cent32, there are few authorities in the field who would question the importance of the relationship altogether.

It is also important to note that relational qualities appear to be as important in non-relationally-oriented therapies (e.g. CBT) as they are in relationally-oriented ones. In fact, in a review of five retrospective studies in which clients were asked what had been the most helpful aspect of their CBT, it was consistently found that clients rated their relationship with their therapist as more helpful than the cognitive-behavioural techniques employed³³.

In terms of which aspects of the therapeutic relationship are most closely linked to outcomes, an extensive review of the research by a Task Force of the American Psychological Association’s Division of Psychotherapy³⁴ identified four ‘demonstrably effective’ elements and seven ‘promising and probably effective elements’. These are presented below in roughly descending order of how strong the relationship with outcomes was rated as being (i.e. strongest predictors first).

Demonstrably effective:
■ goal consensus and collaboration;
■ cohesion in group therapy;
■ therapeutic alliance;
■ empathy.

Promising and probably effective:
■ management of countertransference;
■ feedback;
■ positive regard;
■ congruence;
■ self-disclosure;
■ relational interpretations;
■ repair of alliance ruptures.

So does this mean that Carl Rogers got it all basically right back in 1957³⁵, when he hypothesised that a set of therapist-provided relational conditions (empathy, acceptance and congruence), alongside psychological contact, client vulnerability and the communication of these conditions to the client, are necessary and sufficient for therapeutic personality development to occur? On the basis of today’s evidence, the answer to this question is almost certainly no, and this is for a number of reasons. First, there is clear evidence that many noninterpersonal therapies, such as web-based therapeutic programs and self-help manuals, can be highly efficacious^{24, 36}. Second, the fact that relational factors (or, indeed, any other factors) are related to outcomes does not prove that the former caused the latter. It may be that clients who feel they are doing well in therapy then start to feel more positive about their therapists. Third, while a good therapeutic relationship is predictive of good therapeutic outcomes, it is essential to remember that this is not something that therapists ‘provide’ for clients, but something that emerges in the clienttherapist interaction. Indeed, there is actually more evidence that clients’ contributions to the therapeutic relationship predict outcomes than there is for therapists’ contributions³⁷. For instance, while 56 per cent of studies show a positive relationship between therapists’ affirmation of their clients and outcomes, 69 per cent show a positive relationship between clients’ affirmation of their therapists and outcomes³⁷.

Client factors
Paradoxically, then, clientcentred therapists such as Rogers may have actually underestimated the contribution that clients make to the outcomes of therapy; for what the research suggests is that client factors are probably the most important determinants of therapeutic outcomes, accounting for 70 per cent or more of the overall effectiveness of counselling and psychotherapy²⁸.

Here, a particularly strong predictor of effectiveness seems to be the extent to which clients actively participate in therapy. Reviewing the evidence across a vast range of client, therapist and relational variables, Orlinsky and colleagues suggest that client participation is possibly ‘the most important determinant’ of outcomes³⁸, accounting for 20 per cent or more of improvement alone³⁹. A good example of this comes from an early study by Heine and Trosman⁴⁰, who found that 67 per cent of clients who saw themselves as having an active part to play in the therapeutic process continued with psychotherapy beyond six weeks, compared with just 28 per cent of clients who placed responsibility completely in the hands of their therapists. As with relational factors, this association would seem to be as strong in the more technique-orientated therapies as it is in the less directive ones⁴¹.

Closely related to client participation, more positive outcomes have also been associated with clients who are more proactive in choosing to enter therapy; are more willing to adopt the client role; have higher (but not unrealistically high) expectations of therapeutic outcomes; believe that psychological treatments will be of help to them; and have realistic expectations about what will happen in therapy²⁸.

What the research also indicates is that clients who have higher levels of psychosocial functioning tend to get the most out of therapy. More specifically, better therapeutic outcomes have been associated with clients who have more secure styles of attachment; are more able to think about themselves in psychological terms (psychological mindedness); are not diagnosed with personality disorders; are less perfectionist; and have greater social support²⁸.

Why this is the case is not clear, but it may be that clients who come into therapy with good relational and psychological abilities are more able to form strong therapeutic alliances and engage effectively in the therapeutic process, while clients who are less able in these areas may struggle to make use of therapy.

In this respect, what the research seems to indicate is that therapy tends to help clients capitalise on their strengths, as opposed to compensating for their deficiencies. And, indeed, this does not seem to just take place at the level of overall effectiveness, but also with respect to specific domains of functioning. For instance, there is evidence that clients who have higher levels of cognitive functioning tend to do better in cognitive therapies, while those with higher levels of social functioning tend to do better in more interpersonal therapies⁴².

Discussion
What the research seems to indicate is that at the heart of most successful therapeutic journeys is a client who is willing and able to become involved in making changes to her life.  If that client then encounters a therapist who she trusts, likes and feels able to collaborate with, she can make use of a wide range of techniques and practices to move closer towards her goals. For different clients, different kinds of therapist input may be more or less helpful and there may be certain kinds of input that are particularly helpful for clients with specific psychological difficulties.

But the evidence suggests that the key predictor of outcomes remains the extent to which the client is willing and able to make use of whatever the therapist provides. The old joke, then, would seem to have got it right: how many therapists does it take to change a light bulb? One, but the light bulb has really got to want to change.

1. Health Professions Council. Standards of education and training. London: Health Professions Council; 2005.
2. National Institute for Mental Health in England. Choices in mental health. Leeds: National Institute for Mental Health in England; 2008.
3. Dew, Raimer (2003) cited in Worthen VE, Lambert MJ. Outcome oriented supervision: advantages of adding systematic client tracking to supportive consultations. Counselling and Psychotherapy Research. 2007; 7(1):48-53.
4. Rennie DL. Person-centred counselling: an experiential approach. London: Sage; 1998.
5. Morrow-Bradley C, Elliott R. Utilization of psychotherapy research by practicing psychotherapists. American Psychologist. 1986; 41(2):188-197. 6. Cooper M. Socialist humanism: a progressive politics for the twenty-first century, in Proctor G et al (eds). Politicising the person-centred approach: an agenda for social change. Ross-on- Wye: PCCS Book. 2006: 80-94.
7. Cooper M, McLeod J. A pluralistic framework for counselling and psychotherapy: implications for research. Counselling and Psychotherapy Research. 2007; 7(3):135-143.
8. Rogers CR. On becoming a person: a therapist's view of therapy. London: Constable and Co; 1961.
9. Eisler I et al. Family therapy for adolescent anorexia nervosa: the results of a controlled comparison of two family interventions. Journal of Child Psychology and Psychiatry. 2000; 41(6):727-736.
10. King M et al. Randomised controlled trial of non-directive counselling, cognitive-behaviour therapy and usual general practitioner care in the management of depression as well as mixed anxiety and depression in primary care. Health Technology Assessment. 2000: 4(19).
11. Wampold BE. The great psychotherapy debate: models, methods and findings. Mahwah, NJ: Erlbaum; 2001.
12. Caspi (2004) cited in Carr A. The effectiveness of psychotherapy: a review of research. The Irish Council for Psychotherapy: Dublin; 2007.
13. Nicholson RA, Berman JS. Is follow-up necessary in evaluating psychotherapy? Psychological Bulletin. 1983; 93(2):261-278.
14. Chilvers C et al. Antidepressant drugs and generic counselling for treatment of major depression in primary care: randomised trial with patient preference arms. British Medical Journal. 2001; 322(7289):772-775.
15. Gabbard GO et al. The economic impact of psychotherapy: a review. American Journal of Psychiatry. 1997; 154(2):147-155.
16. Lambert MJ, Ogles BM. The efficacy and effectiveness of psychotherapy, in Bergin AE, Garfield SL, Lambert MJ (ed). Handbook of psychotherapy and behavior change. Chicago: John Wiley. 2004: 139-193.
17. Gloaguen V et al. A meta-analysis of the effects of cognitive therapy in depressed patients. Journal of Affective Disorders. 1998; 49(1):59-72.
18. Roth A, Fonagy P. What works for whom? A critical review of psychotherapy research. New York: Guilford; 2005.
19. Westen D, Novotny CA, Thompson-Brenner H. The empirical status of empirically supported psychotherapies: assumptions, findings, and reporting in controlled clinical trials. Psychological Bulletin. 2004; 130(4):631-663.
20. Elkin I et al. National Institute of Mental Health Treatment of Depression Collaborative Research Program – general effectiveness of treatments. Archives of General Psychiatry. 1989; 46(11):971-982.
21. McRoberts C, Burlingame GM, Hoag MJ. Comparative efficacy of individual and group psychotherapy: a meta-analytic perspective. Group Dynamics: Theory Research and Practice. 1998; 2(2):101-117.
22. Jacobson NS et al. A component analysis of cognitivebehavioral treatment for depression. Journal of Consulting and Clinical Psychology. 1996; 64:295-304.
23. den Boer P et al. Paraprofessionals for anxiety and depressive disorders. Cochrane Database of Systematic Reviews. 2005; (2).
24. Barak A et al., A comprehensive review and a meta-analysis of the effectiveness of internet-based psychotherapeutic interventions. Journal of Technology in Human Services; in press.
25. Miller SD, Hubble M, Duncan B. Supershrinks. Therapy Today. 2008; 19(3):4-9.
26. Okiishi J et al. Waiting for supershrink: an empirical analysis of therapist effects. Clinical Psychology & Psychotherapy. 2003; 10(6):361-373.
27. Crits-Christoph P et al. Meta-analysis of therapist effects in psychotherapy outcome studies. Psychotherapy Research. 1991; 1(2):81-91. 28. Cooper M. Essential research findings in counselling and psychotherapy: the facts are friendly. London: Sage; 2008.
29. Rosenheck R, Fontana A, Cottrol C. Effect of clinician-veteran racial pairing in the treatment of posttraumatic stress disorder. American Journal of Psychiatry. 1995; 152(4):555-563.
30. Wikler M. The religion of the therapist: its meaning to Orthodox Jewish clients. Hillside Journal of Clinical Psychiatry. 1989; 11(2):131-146.
31. Asay TP, Lambert MJ. The empirical case for the common factors in therapy: quantitative findings, in Hubble M, Duncan BL, Miller SD (eds). The heart and soul of change: what works in therapy. Washington, DC: American Psychological Association. 1999: 33-55.
32. Beutler LE et al. Therapist variables, in Bergin AE, Garfield SL, Lambert MJ (ed). Handbook of psychotherapy and behavior change. Chicago: John Wiley. 2004: 227-306.
33. Keijsers GPJ, Schaap CPDR, Hoogduin CAL. The impact of interpersonal patient and therapist behaviour on outcome in cognitive-behavior therapy. Behaviour Modification. 2000; 24(2):264-297.
34. Norcross JC (ed). Psychotherapy relationships that work: therapist contributions and responsiveness to patients. New York: Oxford University Press; 2002.
35. Rogers CR. The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology. 1957; 21(2):95-103.
36. Gould WB. Viktor E. Frankl: life with meaning. Pacific Grove, CA: Brooks/Cole Publishing Company; 1993.
37. Orlinsky DE, Rønnestad MH, Willutzki U. Fifty years of psychotherapy process-outcome research: continuity and change, in Bergin AE, Garfield SL, Lambert MJ (ed). Handbook of psychotherapy and behavior change. Chicago: John Wiley. 2004:307-389.
38. Orlinsky DK, Grawe K, Parks B. Process and outcome in psychotherapy – noch einmal, in Bergin AE, Garfield SL (eds). Handbook of psychotherapy and Behaviour Change. New York: John Wiley. 1994: 270-376.
39. Bachelor A. Comparison and relationship to outcome of diverse dimensions of the helping alliance as seen by client and therapist. Psychotherapy: Theory, Research, Practice, Training. 1991; 28(4):534-549.
40. Heine RW, Trosman H. Initial expectations of the doctorpatient interaction as a factor in continuance in psychotherapy. Psychiatry. 1960; 23:275-8.
41. Nelson RA, Borkovec TD. Relationship of client participation to psychotherapy. Journal of Behavior Therapy and Experimental Psychiatry. 1989; 20(2):155-162.
42. Sotsky SM et al. Patient predictors of response to psychotherapy and pharmacotherapy – findings in the NIMH Treatment of Depression Collaborative Research-Program. American Journal of Psychiatry. 1991; 148(8):997-1008.