If you’ve had therapy, particularly if you got it through a public healthcare system like the NHS in the UK or Medicare in Australia, there’s a good chance it was cognitive behavioural therapy (CBT). Even with private health insurance, if you want therapy, the one you are most likely to be recommended is CBT.
This is quite strange when you consider how many types of therapy there are. Psychoanalysis is well known, but there’s also humanistic therapy, existential therapy and body therapy, to name a few.
However, research suggests that CBT is not a always a good fit for people from black, Asian and ethnic minority backgrounds, those with learning difficulties and those with complex needs. Yet other types of therapy are usually sidelined – the NHS Talking Therapies programme explicitly aims to make 70-90% of their therapy provision CBT.
I believe the prevalence of CBT is not because it is uncontroversially better than the others. It’s a complicated story about how we evaluate whether something works, what it even means to say it “works”, how medical research is funded, and how decisions about which interventions to fund are made.
The randomised controlled trial
So how do we decide whether a therapy works? The answer, in most health systems, is a type of study called a randomised controlled trial. These involve taking two groups of people who have a particular condition researchers are trying to treat or address in some way and giving only one group the treatment.
The people in the groups are chosen to be similar in terms of age, gender, race and other characteristics, and are randomly allocated to an “active” group that receives the treatment or a control group that doesn’t. The idea is that if you have two similar groups of people and the group that gets the treatment improves, then you can be more confident that it is because of the treatment rather than other factors.
In an article for the British Journal of Psychiatry, I argue that though randomised controlled trials might make sense for certain kinds of things, like testing new drugs, they are not always the right tool for evaluating whether therapy works. Randomised trials work best for treatments that are brief, replicable, measurable and where it is clear what you are targeting.
Media Photos/Shutterstock.com
Medicines meet these conditions: they work on a relatively short time-frame; it’s easy to ensure that you are giving the same treatment to all the people in the active group and to measure the effects. And what you’re trying to target is often quite clear because it is usually a particular illness.
Therapy doesn’t meet these conditions. It often takes place over an unspecified period and is highly responsive to the needs of the patient, so it is difficult to replicate. The problem being targeted is not usually specified in advance, may change, and success is often understood in terms of complex goals.
Therapy researchers have dealt with this problem by adapting therapy in various ways. For example, improvement in trials is measured by numerical scales – some of them the same ones used in drug trials – which focus on narrow symptom changes like better sleep.
But qualitative research with patients shows that they often go to therapy for more nuanced goals like self-understanding. The way therapy has been shoehorned to fit randomised trials has created a gap between how it is practised in trials and how it is practised in the wild, where it is understood very differently by therapists and patients.
When bodies such as the National Institute for Health and Care Excellence (Nice) in the UK and the National Institute for Mental Health in the US decide which therapies to recommend, randomised trials carry more weight than any other form of evidence.
It is also easier to publish in prestigious journals and get research funding for randomised trials than other types of research. So the therapies that are most prevalent now are the ones that have most successfully adapted themselves for randomised trials.
Adapting to what randomised trials require
Research suggests that CBT was particularly quick to make this shift, which may help explain its head start in attracting research funding and its position as a recommended evidence-based treatment. Other types of therapy have been playing catch up ever since.
Randomised trials are simply a tool for evaluating efficacy. They are not an end in themselves. Forcing therapy to conform to the shape of these trials, instead of trying to find more appropriate tools, has diminished the range of available therapies to the ones that are most able to mould themselves to randomised trials. In doing so, we have substantially limited patient choice.
Therapy is not a pill. We should not be testing it like one by relying so heavily on randomised trials. The way forward is to pick the right research tools and take them seriously.
The NHS Talking Therapies programme generates one of the largest routine mental health outcomes datasets in the world. However, this observational data plays a much smaller role in Nice guideline development than evidence from randomised controlled trials, despite research suggesting that many schools of therapy perform broadly on par with each other.
Nor do these institutions take research with patients into account, meaning that groups that struggle with the sorts of therapy adapted for trials like CBT, continue to get inadequate treatment.
