Drs Bhui & Bhugra (Reference Bhui and Bhugra2002) address the interesting area of explanatory models for mental distress. They do not, however, justify why we should elicit patients' explanatory models. The notion that members of a specific cultural group hold similar ideas about illness and that culture can be distilled into a set of specific ‘beliefs’ is considered outdated and oversimplified by medical anthropologists. Kleinman (Reference Kleinman1980) points out that explanatory models are idiosyncratic and are justifications for actions rather than causes. Bhui & Bhugra themselves cite Williams & Healy (Reference Williams and Healy2001), who point out that it is difficult to distil a single set of causal explanations that might relate to behaviour, diagnosis or adherence to medication treatment.

The assertion by Bhui & Bhugra that shared understanding of illness between patient and healer distinguishes traditional healing systems from Western biomedicine is simply not borne out by the anthropological literature. In many systems of traditional healing, patients have little understanding of how the treatment ‘works’ and it is the healer who holds highly esoteric knowledge. There is little empirical evidence that eliciting explanatory models improves satisfaction. The one study cited (Reference Callan and LittlewoodCallan & Littlewood, 1998) in fact found that 79% of patients with divergent explanatory models (a comparison of the explanatory models of doctors and patients) were satisfied with psychiatric services.

Of course, patients do have cultural understanding of their illness but this may not be very sophisticated and may not directly relate to decisions about treatments. There is a large amount of data from medical anthropological research which suggests that treatment choice is determined primarily by social and political factors rather than by underlying explanatory models (Reference Pelto and PeltoPelto & Pelto, 1997). Even a study using the Explanatory Model Interview Catalogue (Reference Weiss, Doongaji and SiddharthaWeiss et al, 1992) among leprosy patients suggests that those who held theories of humoral imbalance rather than biomedical theories of infection, sanitation and hygiene had the best biomedical clinic attendance records for leprosy treatment. In terms of treatment outcomes patients may not be interested in how a treatment works (Reference LastLast, 1981) as long as it does work. The weight of empirical evidence suggests that people are keen to utilise biomedical treatments regardless of their cultural beliefs without giving up traditional explanations of illness. In fact, as my own data (Reference DeinDein, 2001) suggest, among Asian psychiatric patients, biomedical and traditional models of illness are held concurrently and informants agree that biomedical treatments help symptoms although they do not treat the underlying cause.

Even if we do elicit our patients' explanatory model, how much will it change the treatment we give them? For instance, consider an African patient who, in terms of an ICD-10 diagnosis, is suffering from a hypomanic episode. He is physically violent. Both he and his family hold that he is possessed by a spirit. Are we to accept their explanatory model and enlist an exorcist? Will we withhold pharmacological treatment because the patient holds an alternative view of his illness?

What is needed is an approach in transcultural psychiatry that looks at not just what people believe but what they actually do in practice. A comprehensive approach involving participant observation, not just the administration of questionnaires to patients, will lead to greater understanding.