background   Possible ways of treating personality disorders vary widely in terms of setting, length of treatment and intensity, leading to differences in cost-effectiveness. From the point of view of both patients and society it is therefore important that therapy choices have an empirical basis. In the health-care sector in the Netherlands we have a model for decision-making known as the 'Funnel of Dunning'; this could serve as a useful guideline for research.
aim To review and integrate empirical evidence concerning psychotherapy for personality disorders, according to the four sections of the Funnel of Dunning: necessity, effectiveness, costeffectiveness, and the responsibility of patients.
method A literature search was conducted in Medline and Psycinfo, using the search terms 'personality disorder', 'psychotherapy', 'effectiveness', 'cost-effectiveness', 'cost-benefit analysis', 'dose-effect relations' and 'efficacy'. In addition we searched the literature manually.
results The urgent need for treatment is reinforced by the first results which demonstrate the heavy burden of disease in patients with personality disorders. There is now convincing empirical evidence that psychotherapy is very effective. However, although preliminary results suggest that psychotherapy is also cost-effective, there has still been no state-of-the-art research into the costeffectiveness of such treatment.
discussion and conclusion Psychotherapy seems to be an effective intervention for treating patients with personality disorders and there are strong indications that the treatment is both necessary and cost-effective. This conclusion contrasts sharply with the recent decision in the Netherlands to cut back expenditure on psychotherapy. Future policy-making will need to be based on research that integrates the data that are collected relating to the various sections of the Funnel of Dunning.