The Drugs don’t work (and nor does anything else)
The first thing to say is that drugs don’t cure mental illness. That’s not even what people are trying to do with them. They can reduce the severity of symptoms and help people to lead less disrupted lives, but they can’t do anything to address the underlying issues. Medication can make things better, but not people. For practical purposes, that can feel like a very fine distinction – ask someone with diabetes. All other things being equal, reducing symptoms and distress is an uncomplicatedly good thing.
But all other things aren’t equal here: there are serious concerns about side effects, about dependency and problems withdrawing from medication and about long-term health effects. And it’s actually far from clear how effective drugs really are at improving things for people. The most powerful evidence for drugs seems to be that people like them – around 20% of the UK population is prescribed some form of psychiatric medication.[1] But reviews of the research (exhaustive reviews by teams at leading research institutions like Harvard and Oxford) indicate that the effect of antidepressants is only modest compared to a placebo[2] or even no better at all.[3]
This is all well-known and it has contributed to a powerful critique of psychiatry. So powerful and widespread that psychiatry spends a lot of its time feeling under attack – this is from an introduction to a piece of research in the British Journal of Psychiatry: ‘there is a deep mistrust of psychiatry … a vocal antipsychiatry movement. Psychiatrists, patients, caregivers and the press are unsettled … and some may think that psychiatric medication is not worth the bother’.[4]
Tucked away, out of sight, in all this criticism of psychiatry is an idea that never really gets examined. That idea is that there is something better available, something that would really help people and psychiatry is keeping it away from them. And that better alternative is psychotherapy – what’s needed is better access to psychotherapy, reduced waiting lists, more money, more therapists and for GPs to stop prescribing antidepressants.
It is extremely difficult to measure the effectiveness of treatments for mental health. That’s as true for medication as for therapy and it’s very basic: there are serious problems in identifying when people are ill and when they aren’t and identifying what conditions they might be suffering from. It’s also difficult and expensive to maintain experimental conditions for extended periods of time, so most randomized controlled trials last around six to eight weeks which is a very questionable period for assessing the effectiveness of treatment for a condition like depression. Also (defenders of psychiatry sometimes make this point) you can’t really assess a treatment just by accessing scores across a population without taking account of the impact of symptoms and treatment on individuals’ lives and there aren’t any research tools to allow you to do that at the moment.
The clearest picture of the effectiveness of psychotherapy comes from the results of the NHS Improving Access to Psychological Therapies programme. More than a million people a year access IAPT, it’s been going since 2008, some participants will also be on medication, others won’t. Patients are assessed at the outset (and generally only go onto treatment if they score above a clinical threshold) and they’re assessed again, usually half-way through treatment and at the end. If their score has fallen below the clinical threshold at the end, they are treated as recovered. This isn’t just far and away the most robust measure available on how effective therapy is at treating depression and anxiety (actually, anywhere in the world), since the results include people on medication, it’s the best measure for how effective mental health care is as a whole. IAPT is the kitchen sink; these figures show what happens when the system tries it’s best to make you better.
The figures generally show that around half the people who complete a course of therapy are considered to have recovered. But as around half the people who start the therapy drop out before the end and aren’t included in those figures, therapy, in fact, only works for around I in 4.[5] And, in reality, even that figure will give a misleading impression about the true effectiveness of therapy. This is because people tend to come forward for treatment (and receive their first assessment) when they are at their lowest ebb, when their condition is at its most acute. So, if you measure the situation again sometime later (and given the waiting lists for therapy on the NHS it is often quite a while) it is to be expected that, in quite a lot of cases, things will have improved from that low point. Statisticians call this ‘reversion to the mean’ and it’s the reason why your GP’s first response is so often to go away and come back in a couple of weeks if it’s still bad. On top of that, you have the placebo issue here too. There is evidence that the placebo effect is particularly strong in mental health:[6] how do you know it’s really the therapy that’s making the difference rather than just getting treatment? There is evidence out there that therapy is no more effective at treating depression than acupuncture or routine visits to the GP (it’s just being treated or cared for that makes the difference).[7] Therapy may not have the same potential to cause problems and adverse effects as medication but it’s pretty expensive; if it’s not doing any good, that’s a lot of money that could be spent on cancer treatment or housing or education.
Also, those IAPT figures just involve taking a score at the end of the last session and sending people off. Can you really say they’re recovered without knowing how they’re doing, say, six weeks later when they’re no longer receiving treatment or a year later? In fact, the evidence suggests that people who are considered to have recovered from depression are quite likely to suffer a recurrence: 25-40% after two years, 60% after five years and 85% after 15 years.[8] That leaves open questions about how ill those people might be relative to other people, or more likely to go for help, and how reliable the diagnostic test really are, but what it certainly doesn’t indicate is that treatment effects long term cures.
These results are pretty bad: at best, they indicate that therapy doesn’t work for around three quarters of the people who encounter it. And, even where people do seem to get better, it could be reversion to the mean or the placebo effect. And evidence suggests that you might expect nearly half those people to be ill again within two years and more than half after five years. That’s your children, your partner, your parents, your siblings.
Much of the criticism of psychiatry is deserved. The problem with the current attitudes towards mental health care is that the criticism is too focussed on medication and psychiatry alone. Psychotherapy gets an underserved easy ride. That’s because, we all want to believe there’s something that works out there, that there are good guys. It’s a comforting idea that there is a real solution and it just never gets a fair chance because we can’t get the stupid or badly motivated people to get out of the way and fund the system properly.
It's worse than that: psychotherapy has had its chance and (other than peoples’ appetite for antidepressants perhaps[9]) the evidence is that nothing we’re doing to treat mental health works very well. The trouble with our over-indulgent view of psychotherapy is that it encourages people to believe there is something that works.
At the risk of an over-extended metaphor, one problem with the idea that there’s a lifeboat that works when there isn’t, is you don’t get around to building a better one.
Another is that people might not be careful enough not to fall into the water. When you see what’s really there (and what isn’t), maybe it isn’t so OK not to be OK.
[1] www.gov.uk/government/publications/prescribed-medicines-review-report/prescribed-medicines-review-summary
[2] https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)32802-7/fulltext
[3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4172306/
[4] https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/putting-the-efficacy-of-psychiatric-and-general-medicine-medication-into-perspective-review-of-metaanalyses/39C15F3428BDD1F8A4C152B67C06A5A6
[5] https://therapymeetsnumbers.com/iapt-2018-why-do-less-than-1-in-5-referrals-reach-recovery/
[6] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6003660/
[7] https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001518
[8] https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(20)30036-5/fulltext
[9] You can argue this isn’t good evidence, since people also like alcohol and cocaine which can be demonstrated to harm them, but an obvious explanation for the levels of antidepressants use is that they make a lot of people feel better