Choose the odd one out: Philippe Pinel, the DSM-IV, Hannibal Lector, and the Home Secretary, Jack Straw. It's Hannibal Lector — the other three have all attempted to define a mental condition manifested by antisocial behaviour. Pinel based his attempt on careful observation, DSM-IV did it by committee and Jack Straw responded to media images. All three would have done better if they had been able to read Paul Moran's carefully constructed review of the prevalence, causes, risk factors and social burden of this ancient disorder, currently enjoying a rediscovery.

The author's conclusions, supported by balanced appraisal of the literature, demonstrate the huge gaps in our knowledge of antisocial personality disorder. We have various prevalence figures but know little of its incidence. Good long-term studies on the natural history of the disorder are still awaited. These need to be based on psychosocial outcome measures and not simply conviction data. We know something of risk factors but little of protective factors. We are not yet in a position to suggest preventive strategies with great confidence. In all these fields, the diagnostic criteria of antisocial personality disorder remain as elusive today as they were to Pinel two centuries ago. Moran's argument for a combined dimensional and categorical approach to personality disorder is compelling. Until we have a clearer understanding of the measurable personality characteristics that render a person dissocial (as distinct from his or her behaviour) we will continue, like our politicians, to make statements based on belief rather than science.

The author was commissioned to conduct this review by a government agency to help underpin a new research programme. Ministers need to look at Moran's findings before they leap into legislation on severe personality disorder. Sadly, we know that tabloid headlines, and not reviews of current research findings, are likely to inform government thinking on this matter.