Commentary on NICE Depression Draft Guideline Update

Entry for 12 April 2009:

The NICE (National Institute for Clinical Excellence) guidelines are a set of practice guidelines developed to guide the treatment of medical and psychological problems in the UK. Over the past few years, however, they have become increasingly controversial, and there has been quite a lot of fulmination and even some lawsuits over them. For example, recently by a group of people with ME (Myalgic Encephalomyelitis; the UK term for Chronic Fatigue Syndrome) sued NICE, claiming that the guidelines for ME restricted their treatment to CBT and graded exercise, which they said were ineffective for them (they lost...).

In February, NICE issued its long-awaited revised draft guideline on treatment of depression. Previously, “counselling” had been in the guidelines as a second line treatment if CBT or medication didn’t work. Now, however, this recommendation had been transformed into what amounts to a health warning (p. 197):

6.5.3.1 For people with persistent minor and mild to moderate depression who have declined a low intensity intervention or group CBT, counselling may be considered. However, practitioners should take care to explain the uncertainty about the effectiveness of counselling for people with
depression.
6.5.3.2 Counselling for people with persistent minor and mild to moderate
depression should be:
• based on a non-directive person-centred model
• typically in the range of 6 to 10 sessions over 8 to 12 weeks

In other words, if clients are pig-headed enough to turn down the recommended treatments, they should be permitted to receive counselling, but with a warning that nobody knows if it's effective for depression. The guideline goes on to say that the counselling should be “nondirective person-centred” and generally no more than 10 sessions, after which the person would once again to urged to get themselves to a more effective treatment, possibly with an “I told you so!”.

There are so many things wrong with this recommendation that it is difficult to know where to begin. When I first looked at the document, I felt so scandalized and embarrassed by it that I didn’t know what to say; it seemed so shoddy, really a travesty of science. What is my justification for such a strong reaction?

1. Let’s start with the cited evidence base for the guideline, which consists of three RCTs (Bedi, 2000; Simpson, 2003; Ward et al., 2000). These three studies are a diverse group: Simpson (2003) used psychodynamic counsellors; Ward (2000; a version of the King study) used person-centred counsellors; and the theoretical orientation of Bedi (2000)’s counsellors was given. When Beth Freire and I were collecting studies for our meta-analysis of Person-centred-experiential (PCE) therapy outcome, we looked at the Bedi (2000) study and dropped it because the theoretical orientation of the counsellors wasn’t specified; and we didn't look Simpson (2003) at all because of their focus was on psychodynamic counselling. (The latter should have been included in the short-term psychodynamic therapy section, but apparently was not.)

However, the theoretical orientation of counselling in the draft guideline is described as person-centred. It is a step in the right direction to specify the theoretical orientation of the counsellors (this was not done in the previous depression guideline). However, only one of the cited studies (Ward, 2000) used counsellors with a person-centred orientation (more on this below). So the first problem is that the guideline is not consistent with its evidence base.

2. Nevertheless, on looking more deeply at the documentation for the guideline, it became clear that the real problem is a significantly flawed search strategy: This is described (in the Full Guideline document, section 6.1.5, p. 132) as using the following treatment inclusion criteria:

Behaviour activation
Cognitive behavioural therapies
Computerised cognitive behaviour therapy
Counselling
Couples-focused therapy
Guided self-help
Interpersonal therapy
Problem solving
Physical activity
Psychodynamic psychotherapy
Rational emotive behaviour therapy

I was startled to note the absence of the following important search terms, all of which have outcome literature:

Person-Centred therapy
Client-Centred therapy
Nondirective therapy
Supportive therapy
Process Experiential therapy
Emotion-Focused therapy

So the second problem is that, by any account, the search strategy used resulted in an under-representation of the existing literature on PCE therapy for depression. When Beth Freire and I (Elliott & Freire, 2008) did our meta-analysis, we identified 17 different comparative outcome studies involving person-centred-experiential (PCE) therapies, 16 of them RCTs. Yet only one of these studies (Ward, 2000) was reviewed in the GL document.

Why weren’t these studies included? Why didn’t the review panel look more carefully for literature? Were they under-resourced? Did they lack expertise on this literature? Or did they just not bother because they didn’t think there was anything there anyway? In any case, this should be, for them, an embarrassing omission.

What studies are missing? In particular, the following RCTs of process-experiential/emotion-focused therapy with clinically depressed clients are not included:

*Goldman, R. N., Greenberg, L. S., & Angus, L. (2006). The effects of adding emotion-focused interventions to the client-centered relationship conditions in the treatment of depression. Psychotherapy Research, 16, 537-549. [Replication of Greenberg & Watson, 1998: Well-designed RCT with clinically depressed clients and medium-sized sample, comparing PE-EFT to Person-Centred therapy; outcomes were strongly and significantly better for PE-EFT.]

*Greenberg, L.S., & Watson, J. (1998). Experiential therapy of depression: Differential effects of client-centered relationship conditions and process experiential interventions. Psychotherapy Research, 8, 210-224. [Well-designed RCT with clinically depressed clients and medium-sized sample, comparing PE-EFT to Person-Centred therapy; outcomes were slightly but not significantly better for PE-EFT.]

*Watson, J.C., Gordon, L.B., Stermac, L., Kalogerakos, F., Steckley, P. (2003). Comparing the effectiveness of process-experiential with cognitive-behavioral psychotherapy in the treatment of depression. Journal of Consulting and Clinical Psychology, 71, 773-781. [Balanced researcher allegiance RCT comparing PE-EFT to CBT in clinically depressed sample; outcomes for PE-EFT were generally at least as good as CBT and on some measures were better.]

In addition, the following studies of person-centred counselling with perinatal depression are also missing:

*Cooper, P. J., Murray, L., Wilson, A., & Romaniuk, H. (2003). Controlled trial of the short- and long-term effect of psychological treatment of post-partum depression. British Journal of Psychiatry,182, 412-419 . [Well-designed RCT; Person-centred comparable outcome to CBT and Psychodynamic]

*Holden, J.M., Sagovsky, R., & Cox, J.L. (1989). Counselling in a general practice
setting: Controlled study of health visitor intervention in treatment of postnatal depression. British Medical Journal, 298, 223-226. [Medium-sized RCT: Clients in Person-centred therapy had better outcomes than control clients]

*Morrell CJ, Warner R, Slade P, Paley G, Dixon S, Walters SJ, Brugha T, Barkham M, Parry G, Nicholl J. (In Press). Clinical effectiveness of health visitor training in psychological interventions for postnatal women – a pragmatic cluster-randomised trial in primary care. BMJ. [RCT; Person-centred similar outcome to CBT, better than treatment as usual]

*Wickberg, B., & Hwang, C. P. (1996). Counselling of postnatal depression: A controlled study on a population based Swedish sample. Journal of Affective Disorders, 39, 209-216. [RCT; Person-centred much better than treatment as usual]

The Elliott & Freire meta-analysis also includes RCTs by Beutler et al., 1991; Brent et al. 1997; Fleming, 1980; Lerner, 1990; Maynard, 1993; McNamara, 1986; Propst et al, 1992; Shaw, 1977; Stice et al., 2006; Tyson et al, 1987; Wilson et al., 1990.

3. What are the consequences of omitting these studies? Across 33 comparisons between PCE therapies and other therapies (in 22 cases, CBT) in Elliott & Freire’s (2008) data set, the mean comparative effect size was -.07, a statistical equivalence finding (meaning that it is not statistically different from zero but is significantly different from -.4 standard deviation units, a useful cut-off for demarcating small from medium effect sizes). A robust literature supports the proposition that PCE therapies are equivalent in effectiveness to CBT and other therapies. The third problem with the revised draft guideline is that the reading of the evidence base is simply inaccurate.

The implications: The Revised Draft NICE Guideline for Depression is in significant error on its recommendation against offering counselling to depressed clients. The guideline should be revised by deleting the prejudicial language on counselling, cited at the beginning of this entry, and the Treatment options section should at minimum be revised as follows:

6.5.4 Treatment options
6.5.4.1 Discuss the relative merits of different interventions with the person with depression and offer:
• antidepressant drugs (normally SSRIs)
• psychological interventions (normally CBT, interpersonal therapy, or person-centred-experiential [PCE] therapy)
• a combination of antidepressants and CBT.

The choice should be based on patient preference, the likelihood of adherence to the treatment, and the likely side effects.

Furthermore, the next section should also be revised, at minimum, as follows:

6.5.5 Choice of psychological treatment
6.5.5.1 For people with moderate depression who are offered psychological interventions the choice of treatment should include:
• Individual cognitive behavioural therapy
• IPT if the person expresses a preference for it or if, in the view of the healthcare professional, the person with depression may benefit from it.
• PCE therapy if the person expresses a preference for it or if, in the view of the healthcare professional, the person with depression may benefit from it.

I say these guidelines should be revised at least as far as I’ve suggested, because to my mind they still display way too much bias in favour of SSRIs and CBT, but this at least would be less of a travesty of science than the current version.

When the BACP Research Committee examined these draft guidelines last month, we noted various places where evidence was interpreted in favour of CBT and against other therapies, and also how difficult it was to really pin down the evidence base for the conclusions made. I’m sorry that the report isn’t more careful, thorough or balanced, given the amount of time and effort that the review panel obviously put into the review process (there are a couple of thousand pages of documentation). Maybe I'm wrong, but it looks like they just didn’t care what they did with the PCE therapy or counselling evidence base. Not caring might be convenient when it favours your expectations or preferences, but it is neither good science nor good health care policy. When the consequences of this carelessness have real world consequences, risking thousands of counsellors losing their jobs or having to retrain in CBT, and hundreds of thousands of clients having their access to the full range of effective treatments curtailed, then carelessness can begin to take on a moral or even an ethical dimension.