Peter Gøtzsche, a Danish physician and researcher, has written a book arguing that 97% of psychiatric drugs cause more harm than good. Allen Frances, emeritus professor of psychiatry at Duke University and chair of the committee that produced DSM IV, says that 70% of Gøtzsche’s book is right but the big problem is that patients with serious psychiatric illness are being abandoned while many people with minor problems are overtreated. Both were speaking at a meeting in Leiden last week to celebrate 50 years of Geneesmiddelenbulletin, a journal that provides independent advice on prescribing.

The Dutch national organisation of psychiatrists boycotted the meeting and emailed their members saying they should not attend because the meeting was antipsychiatry, antiscience, and bad for patients. Yet there were some 40 Dutch psychiatrists in an audience of 400, and two of the speakers were Dutch psychiatrists.

Dutch psychiatrists, it was explained to me, are feeling vulnerable because there are too many of them. They have two treatments to offer–drugs and psychotherapy. But the Netherlands has many clinical psychologists, and they have taken over the psychotherapy. Psychiatrists are left with drugs and anxiety about their future.

Gøtzsche’s attack is profound. The research base of psychiatry is unreliable and corrupted by the pharmaceutical industry. There is little evidence that psychiatric drugs, particularly antidepressants, actually work, and yet the drugs kill through making suicide more likely. The clinical expertise of psychiatrists is unreliable, with psychiatrists not agreeing on diagnoses; and the patients are not listened to–worse they may be locked up against their will. Psychiatry is also guilty of creating “fictitious diseases” like ADHD (Attention Deficit and Hyperactivity Disorder).

Jim van Os, a Dutch professor of psychiatric epidemiology, summarised the evidence base of psychiatry in 25 minutes, beginning by pointing out the absence of any established link between a brain abnormality and a psychiatric diagnosis. Frances later said that neuroscience and neurogenetics are exciting areas of science but have not benefited even one psychiatric patient. What we do know, said van Os, is that many people are mentally vulnerable, that there is real need. (Trudy Dehue, a historian, sociologist, and philosopher of science who has been booed by psychiatrists, pointed out that those who criticise the diagnoses and treatments of psychiatry are not dismissing the suffering of the patients.)

Problems arise, continued van Os, when we try to classify the vulnerability. Frances said how when producing DSM IV the task force had adopted a very high evidence hurdle for agreeing a new diagnosis. They had 94 proposals but accepted only two, both of which, he added, turned out to be “disasters.” The DSM V task force started with a blank sheet, which could have been useful if they’d agreed a high evidence hurdle for a diagnosis. Instead, they have created many new diagnoses driven not by the pharmaceutical industry, said Frances, but by the “intellectual conflicts of interest” of the members of task force. By intellectual conflict of interest he means commitment to their own research, theories, ideas, and experience. But once the diagnoses are there the industry moves in fast.

We know, continued van Os, that “one size does not fit all” and there is great heterogeneity among the patients within a diagnosis and in how patients respond to treatment. At a group level almost all psychiatric treatments have a small positive effect, but the group response hides the heterogeneity.

One consistent finding with the evidence is that over some 40 years the placebo response has been getting stronger and the effect of the psychoactive drug weaker, meaning the difference has diminished. This may be because less sick people are being treated and included in trials. Another consistent finding is that drugs plus psychotherapy works better than either treatment alone. The intensity of treatment is also important, with, for example, 12 sessions of psychotherapy being more effective if delivered twice a week rather than once a week. The quality of the relationship between the doctor and the patient is also crucial, with a good relationship meaning not only that talking treatments are more effective but also drug treatments. A placebo delivered in a good relationship may be as effective as an active drug delivered in a poor relationship.

Gøtzsche agrees that psychotherapy can be effective, even, he believes, with severely psychotic patients. Perhaps psychiatry has been diverted by being too concerned with drug treatments, particularly when, as we heard in the morning’s session of the conference, drug companies have pursued a mission of increasing the number of people taking their drugs rather than producing more effective drugs.

Now, said Frances, the worried well are being treated on a huge scale while this is the worst time ever to be severely mentally ill in the US. About a fifth of the population in the US is taking a psychoactive drug but perhaps 4-5% are really sick. The privatisation of services for the severely ill has meant that many, some 300 000 patients, have ended up in prison or on the streets; and in prison it is the psychotic who end up in solitary confinement (the worst thing possible for a seriously ill patient) or being raped. The UK, warned Frances, is going the same way.

Psychiatrists may have forgotten, he suggested, that the brain is the most complicated thing in the universe. Schizophrenia may be a 1000 different conditions, as may Alzheimer’s disease and other psychiatric conditions. Changing diagnostic criteria means, said Frances, that they are getting closer and closer to normal. Some 105 genes have been associated with schizophrenia, but these findings are unlikely to lead to better treatments. Personalised medicine, he said, is a marketing term.

Hippocrates, said Frances, made his statement of “first do no harm” to counter overtreatment and cruel treatments happening elsewhere in Greece. The history of medicine and psychiatry is filled with ineffective and often cruel treatments. A third of patients, he said, will recover without treatment, and another third will not respond to treatment. Doctors must be prognostications and treat the middle third, avoid treating those who will recover untreated, and help as much as possible, perhaps through social support, those who do not respond.

The audience was left with the impression that Gøtzsche may have overdone his condemnation of psychiatry but that his attack needed to be taken seriously. Psychiatry seems to have lost its way, as, I believe, has all of medicine. The response should not be to refuse to listen to criticisms, as the Dutch organisation for psychiatrists advised, but to listen to and use them as a stimulus to deep examination of current practice.

Richard Smith was the editor of The BMJ until 2004. 

Competing interest: RS chaired the day’s meeting and had his expenses paid (at least he hopes he will).