It is heartening to see trials of therapy for anorexia nervosa, the most lethal of psychiatric illnesses and the cinderella of research. Dare et al (Reference Dare, Eisler and Russell2001) have shown that over a year it is possible to effectively treat a severely ill group of young adults with poor prognostic features, and to do so on an out-patient basis.

I am surprised, though, that they feel able to conclude that “ specialised psychotherapies are more effective than routine treatment”. The two therapies which came out ‘top’ (family therapy and focal psychoanalytic psychotherapy) were given by the same three highly experienced therapists, and the next best therapy (cognitive—analytic therapy) was given by trained specialists in eating disorders, whereas ‘routine treatment’ was provided by junior psychiatrists on 6-month rotations who had to hand over to colleagues during the year of patient contract.

Certainly, confidence in at least one model of therapy is an important component of an experienced therapist's effectiveness, but for me the clearest implication of the study is that patients who suffer from this chronic condition do best with continuity of care from the most experienced therapists. I am not sure we can say anything at all yet about choice of specific therapy.