When I was a medical student, we were in the halcyon days of therapeutics in psychiatry. Each new treatment, whether it was a new antidepressant or antipsychotic drug, or a new psychological approach such as ‘implosion’ for phobic anxiety, was greeted with enthusiasm and excitement. The snake pits of mental hospitals, where people had been placed because they were incurable, were on the way out, carried away into the distance by this wave of therapeutic optimism, where everything that came into the ambit of psychiatry was potentially treatable. Indeed, such was this optimism that when a wise old bird like Desmond Curran, doyen of psychiatry at St George's Hospital, argued against this in favour of ‘Psychiatry Ltd’ (1952) he was regarded as in his dotage by the young therapeutic turks and ignored.

My, how times have changed. None of Adshead's (2001, this issue) seven factors of treatability ever entered our heads. The nature and severity of the pathology we were treating was immaterial. William Sargant, who taught generations of medical students that they would be on to a winner if they entered psychiatry, shouted at us that it did not matter if we had only the foggiest idea of the changes in the brain in depression; what did matter was that antidepressants worked, and we should use them in the same way that quinine was used to treat malaria before the protozoan parasite was discovered. The extent of involvement of other bodily systems and coincidental morbidity was also immaterial; this was the age of empiricism, when experiment was everything. As for diagnosis, this was easy. You found out whether your treatment was effective and, if it was, the condition could be labelled accordingly. Thus, when the wide spectrum of efficacy of antidepressants became known, we had masked, atypical, hysteroid and hypochondriacal depression, all of which were trumped by the Italian ‘depressione sine depressione’, which allowed all conditions to be treated logically with antidepressant drugs. The questions as to whether sufficient staff or facilities were available to treat the disorders concerned and the cultural implications of so doing also fell by the wayside. If there was a treatment that worked, it was up to the front-line staff to provide it; nothing more needed to be said.

Although these views now seem utterly outdated and stereotypical, they have some relevance to Adshead's article. Many of the obstacles to successful treatment outlined so eloquently by her would disappear overnight if unequivocal treatment success could be demonstrated. Doctors love to be in the treatment business and, once they have good evidence of efficacy, they generously treat not only these conditions but also many others not formerly considered to be in the frame. The continued penetration of the tentacles of cognitive therapy to almost all parts of psychiatry is witness to this. The real problem with most forms of personality disorder is that no one particularly wants to treat them and those who do are hard put to demonstrate efficacy in a way that satisfies the proponents of evidence-based medicine. It is therefore very convenient to use complex arguments and weasel words to avoid taking on the responsibility for caring for a population that yields few rewards and many brickbats. The question doctors and other health professionals have to ask is, if we do not treat, or at least manage, the care of people with personality disorders, who else will do it more successfully? Currently, the answer is a deafening silence and I personally feel, although I recognise the counterarguments, that we are best placed to carry out these responsibilities. John Gunn (1992) has made the observation that “if psychiatry gives up all its difficult patients, society will give up psychiatry” and there is no doubt that society is looking towards us, perhaps rather desperately and despairingly, for a lead.

I also suspect that we have treatments for personality disorder that are effective but have not yet researched them adequately. When Adshead comments that very few cancer treatments have been exposed to randomised controlled trials, she is not being entirely accurate. As anyone will know who has sat on the boards of the main funding bodies in medicine, it is the cancer trials that are invariably funded and psychiatry is usually an also-ran at the rear of the field. This is partly because of the greater intrinsic difficulties of carrying out such studies in psychiatric patients in general and patients with personality disorder in particular, but it is not an argument for giving up on high-quality research in these disorders, or for accepting some sort of lower standard for the evaluation of efficacy.

In any case, dealing with people who have personality disorders can be rewarding and is certainly challenging, and once successful treatments are available the diagnosis will lose its stigmatising label, just as depression did all those years ago. Adshead starts her article with a quotation from one of Belloc's cautionary tales, gently suggesting that some of our enterprises may be fruitless. Another cautionary tale may be equally apt, in which Belloc generously describes a young woman with a clear personality disorder:

I think this nicely strikes the balance between the angry despair and mindless optimism that Adshead is so keen to avoid. Constructive dialogue should continue.