There are two problems. One is that there is good reason to believe that other drugs, and in particular the antipsychotics, in low doses, can be beneficial in obsessional disorders. Until clomipramine, however, no company had any incentive to develop the OCD market: there just did not seem to be enough people with the condition to warrant the development costs. Until the mid-1980s, estimates of the prevalence of OCD stood at something like 0.05%; now it is thought that up to 3% of the population may have it. Recognition of OCD has gone up partly because of company support for its recognition. Following the success of clomipramine, the selective serotonin-reuptake inhibitor (SSRI) companies had an incentive to sell OCD in the belief that, if they increased its recognition, the sales of their compounds would follow. This, it should be noted, is not necessarily a bad thing, as it can be argued that until recently OCD was under-recognised, but this selling of diseases does change the way we all view clinical problems.

The example of clomipramine and the SSRIs used in the treatment of OCD shows drug companies listening to some rather minor research, the outcomes of which suited their interest. In this and other cases they then cultivate the germinating seedling. At present, an increasing proportion of clinical research is closely tied to the marketing of compounds. As non-commercial research becomes increasingly difficult to fund, particularly for relatively uncommon conditions such as obsessional disorders, the funding that might come from drug companies becomes ever more attractive. Where OCD was concerned, effectively from the early 1970s to the mid-1980s the only research being funded was by Ciba-Geigy.

The question of who is listening to the outcomes of research, what resources they have, and what interests they might have in promulgating the results of that research, is becoming an increasingly important one in science generally but perhaps in psychiatry in particular. For example, exposure therapy has had in comparison far fewer resources pushed its way, even though it is probably more beneficial for obsessional disorders than the SSRIs. The field must inevitably become distorted if information about SSRIs is facilitated but that about exposure therapy is not.

Up to the mid-1960s, the MAOIs had been the most popular antidepressants. Then the ‘cheese effect’ was discovered, and in a large clinical trial it appeared that the MAOIs didn’t work. There was a dramatic slump in the sales of the MAOI antidepressants from which they have never recovered. However, many of the clinicians who used MAOIs regularly were not prepared to accept that these drugs were not antidepressants. If they were not as effective for conventional depression, the argument was that they must be suitable for other forms of depression. Conveniently, from the early 1960s, there had been suggestions that MAOIs might be specifically beneficial for a variety of atypical depressions, usually those with prominent anxiety features.

The concept of atypical depressions flourished during the 1970s and 1980s, even though no form of atypical depression with a specific response to MAOIs was ever substantiated. The concept should have vanished down the plughole of interesting but irrelevant concepts but it did not, in great part because it provided a marketing niche, used for the advertising of MAOIs.

What advertising? An increasingly large proportion of the scientific literature supplied to prescribers and other mental-health professionals is supplied to them by drug companies. In many instances, this is an uncomplicated provision of information, in the form of free literature searches and other facilities, but this trend has its downside. Companies will rarely spontaneously provide information that might not be supportive of their product or that is too supportive of competing treatment modalities. This lends a bias to the clinical enterprise. It also leads to concepts such as atypical depression surviving when they might otherwise have vanished into oblivion.