Balancing policy development and research evidence: are we falling short?

Disease sometimes brings increased risk of death, morbidity and distress beyond its primary victims. Most diseases thus attract collective effort from whole communities towards their understanding, prevention and/or amelioration. Governments fund research and relevant services; corporate and individual private wealth may be added, often with a celebrity or two to promote the cause. Only in satire is punishment added to the sufferings of illness (1) Ð that is, unless that illness is a mental disorder. Then, public attitudes are more fearful and hostile, less funds ̄ow, and many of those who fail to form a therapeutic alliance with primary health services, for whatever reason, fall into other systems, including the criminal justice system. When people present with both mental disorder and offending behaviour, confused moral responses tend to outweigh evidence-based approaches. This tendency peaks when the diagnosis is the tautological `anti'or `dis'-social personality disorder. Even though details of response may differ between nations and cultures, in most, politics augments the problem. In the relatively small con®nes of the United Kingdom this is currently well illustrated as frightening, highly publicized cases have provoked government intervention. In Scotland, the relatively pragmatic MacLean Committee was appointed, and will `review various matters relating to the sentencing and treatment of serious violent and sexual offenders, including those with personality disorder'. For England and Wales, however, the government's Home and Health Departments have already proposed preventive detention of `dangerous people', using a concept of `severe personality disorder'. New laws and services may be introduced based on `informal discussions over the past two years'. It is acknowledged that there are no data to support such laws and services, but asserted that they cannot be delayed (2). Scandinavian (e.g. 3, 4) and American (5) studies have established a small but signi®cant association between some mental disorders, most consistently psychosis and substance abuse disorders, and offending, particularly violence. Personality disorder (p.d.) has proved more dif®cult to pin down in this respect. First, Loranger's (6) demonstration of the difference that transition from DSM-II to DSM-III made in apparent point prevalence of p.d. rather discredits a categorical approach to such disorder; secondly, standardized interview techniques, so helpful in cases of psychosis and even substance abuse, have been questioned for p.d. (e.g. 7); then, categorization has rarely been exclusive of antisocial acts Ð in the Swanson study (5) the violence criteria were explicitly integral to the diagnosis of p.d., rendering tests of association impossible. In these respects, dimensional approaches to p.d. have not yet fared much better. Adequate service provision, further, requires explanatory models. Does offending follow from the mental disorder, or mental disorder from offending or its consequences? Are the two rather related by a common cause? Could the frequency of association be a mere accident of the coincidence that offending on the one hand and schizophrenia, personality disorder and substance abuse on the other, peak for onset and effect between similar ages, at least for men? The nature of any relationship between disorder and offending is best described for psychosis which, for these purposes, has the advantage that the disorder can be described wholly without reference to lawbreaking. A small body of work on relative timings of onset of disorder and offending (8) suggests a direct effect of illness on violence. Other work shows not only an overall correlation between the presence of symptoms and violence (e.g. 9, 10) but also that symptoms may directly drive violence (11, 12). So far, so good, but even if violence in the presence of psychosis were invariably driven by delusions (and it is not), such symptoms occur more commonly and for longer without expression in violent acts. Do we even understand as much about any links between elements of psychopathology and offending when the disorder is one of personality? Further clari®cation requires that people with such disorders who are violent are compared with those who suffer and survive without; potential strengths are as important as weaknesses (or risk factors). The study by Miller et al. (13) not only Acta Psychiatr Scand 2000: 102: 1±2 Printed in UK. All rights reserved Copyright # Munksgaard 2000