Before NICE (the National Institute for Health and Care Excellence) finalised and published its first depression guidance in 2004, it made a draft publicly available for comment. NICE will make a draft of a new version of this guidance available for comment in July 2017. Anticipating this new draft, what are the implications of NICE following the same flawed approach to evaluating evidence as they did 13 years ago? Will their suggested approach continue to leave people with long term depression struggling with the limited choices of psychotherapy on offer?
The late Professor Phil Richardson, a psychotherapy researcher, made one notable comment on the 2004 draft guideline: ”where depression is concerned, there is not a single demonstration of the specific efficacy of any psychological treatment covered by the Guideline over any other psychological treatment covered by the Guideline… for this reason it is inappropriate to single out certain psychological treatments for special mention in the Guideline recommendations”. (This quote is from a personal copy of Phil’s response).
Richardson suggested that NICE’s guidance was based on a misguided understanding that:
1) it is useful and necessary to decide if one psychological therapy is better than another;
2) the key means of informing this decision is by using the treatment has the most randomised controlled trials (RCTs), which is not the same thing as the best evidence. RCTs compare treatment A to no-treatment. They do not directly compare treatment A to treatment B.
Richardson was not alone in drawing attention to this problem, yet NICE recommended cognitive behavioural treatment (CBT) for depression because it had the most RCTs. Not long afterwards, Lord Layard (an economist and peer) used the guideline as a basis to lobby the government to fund a new Improving Access to Psychological Therapy (IAPT) service to provide CBT en masse for working age adults with depression or anxiety across the country. This was to the exclusion of many other forms of psychological therapy, which had not managed to gather RCT evidence in time.
RCTs have been criticised since they were first used around the 1950s, particularly in terms of mental health – and particularly for psychological therapies. RCTs favour certain types of therapy such as CBT, because CBT is brief (usually 6 sessions) and structured – it follows a manual. Other therapies like psychodynamic psychotherapy are more difficult to fit around an RCT because they last a lot longer, and don’t follow a strict manual – therapy is more tailored.
In developing its guidelines, NICE reviews all RCTs for each condition and summarises the results. It looks to see which treatments make the biggest impact on symptoms when compared to having no treatment. NICE reviews of psychotherapy to date have tended to look only at the impact of each treatment at the time the treatment finishes. Some RCTs look at the impact of treatment at a follow-up period, which could be 1 or 2 years after the treatment finished – but most don’t, and so NICE has tended to ignore follow-up periods.
Most psychological issues like depression are long-lasting, and any good treatment should have a longer-lasting impact, beyond the end of treatment. If the effects of the treatment wear off as soon as (or soon after) the treatment finished then the treatment can at best be considered a good sticking plaster. For example, NICE reviews of psychotherapies for depression focus on their impact on ‘symptoms’ – things like sadness, pessimism, worthlessness, agitation, frequency of crying. Psychiatrists consider these expressions of emotion to be equivalent to medical symptoms. Things like quality of life, ability to deal with problems, being able to take care of oneself, feeling independent, feeling able to communicate with others, feeling able to get about, ability to attend education, work or social events are seen as secondary outcomes. If in 2004, NICE had examined the RCTs in terms of these secondary outcomes, CBT would not have been found to be superior. The efficacy of CBT is in large part determined by an RCT model that focusses on medical symptoms that are easily measurable within an RCT framework.
There are several other problems with RCTs in psychotherapy. These include the forced catagorisation of complex problems into simple uni-form diagnoses; using self-report questionnaires which are potentially unreliable and disliked by service users; ignoring important elements of therapy such as characteristics of individual therapists and individual clients.
The reification of CBT fails to take account of the well-established ‘Dodo Bird Verdict’ prevalent in psychotherapy since the 1930s: an idea that ‘all have won and all shall have prizes’ (in other words, all psychotherapies are roughly equally effective). Variation is down to individual therapists and individual clients. Moreover, therapy type may be more a matter for patient choice than which has the larger impact on group change when compared to no-treatment.
The continued professional resistance to abandoning a failing approach to evaluating psychotherapies finds NICE in the midst of yet another revision to the depression guidelines due in July. Its review work is likely to have stuck with the traditional hierarchy approach, in which having more RCTs makes a therapy better.
In the case of long-term depression, perhaps NICE will follow in the footsteps of the 2016 European Psychiatric Association (EPA) guidelines . In their assessment, the EPA decided to bunch together any type of depression lasting two or more years, and look at RCTs fitting this criterion. The result was a hierarchy of treatments with a clear winner called Cognitive Behavioural Analysis System of Psychotherapy (CBASP). There had been 5 RCTs of CBASP at the time, carried out in the USA, Germany and the Netherlands. CBASP is not common in the UK, and is a hybrid of other techniques. Interpersonal Therapy (IPT) came 2nd because there is one RCT. A range of therapies were given 3rd equal place, including CBT because it had some RCTs but not very good ones. Psychodynamic psychotherapy came 3rd equal because it had one RCT, which was published after the taskforce had done their shortlisting, but before giving out prizes, and as such they were not able to incorporate it fully. They did find space to review it separately and concluded that it demonstrated some usefulness of the treatment, so slotted it in at 3rd place for good (but late) effort. This RCT of psychodynamic psychotherapy included a 2 year follow up which showed long lasting effects of the treatment whereas the CBASP RCTs did not report follow-up results: this should perhaps have counted for more, since this guideline is for depression that has persisted at least two years already. The clients treated in the RCT of psychodynamic psychotherapy were also generally more severely depressed at the start of therapy than those in the CBASP trials. Here, it seems that good evidence is constituted by the timeframe of the publication of the guidelines, rather than an assessment of all available good evidence.
The EPA also noted that the type of psychotherapy should be individually chosen in consideration of early versus late onset; type of depression; number of episodes; early trauma; symptom severity; patient preference; and comorbid personality disorder. These are ‘good practice points’ rather than firm recommendations, yet arguably these should trump the recommendations which are based on a poor set of RCTs which is generally understood to be a poor means of judging the value of psychotherapy. Another good practice point is for a personalized approach based on patient preferences and needs, such as medication or psychotherapy; group or individual psychotherapy; in- or outpatient treatment. Again, patient preference, which should come foremost, is trumped by an inadequate methodology.
It is at best disappointing that NICE seems set to remain stuck in this approach to making guidelines concerning psychotherapies. This will not help to improve the lives of people suffering from what is usually a lifelong set of severe psychological difficulties. Long-term depression is often linked to childhood and/or adult experiences of trauma: it affects patients’ ability to function in the world and to form good relationships, with consequent impact on social and economic isolation and vulnerability.
NICE needs to break away from the faulty paradigm of RCTs and prioritise patient preference, need and individual circumstances in a way that properly democratises patient involvement. This currently seems unlikely, but we await the draft’s contents with interest.