Positive features

Many critics of the DSPD Programme on civil libertarian grounds appear to forget that before its arrival (and indeed after it) many patients were already detained in secure psychiatric facilities – often for very long periods – under the legal designation of ‘psychopathic disorder’. As this categorisation does not equate to ‘any scientifically validated category’, ^{Reference Blackburn8} the DSPD proposal was an advance in as much as it attempted to anchor the criteria for detention to an acceptable scientific nosology (see below).

The DSPD initiative was also founded on the correct assumption that a small number of offenders are responsible for a disproportionate number of offences, ^{Reference Moffitt9} with the corollary that if one could identify such offenders and either incarcerate them and/or target their needs with effective interventions, this ought to be the basis of a sensible policy. However, it then assumed (incorrectly) that severe personality disorder was a way of identifying such a group. This belief was widespread as evidenced by its appearance in the Labour Party's manifesto in 2001, which proposed to deal with ‘the most dangerous offenders of all – those with a severe personality disorder’. ¹⁰

The DSPD initiative then went on to identify explicitly the characteristics of individuals that would make it appropriate to detain them for ‘treatment’. To be detained, the individual needed to satisfy the following triad:

1. (a) being dangerous, defined as posing a ‘significant risk’ (>50%) on two risk assessment tools (the duration of this risk being unspecified);

2. (b) having a severe personality disorder (SPD) as evidenced by a Psychopathy Checklist – Revised (PCL–R) ^{Reference Hare11} score of >30 or PCL–R score of 25–30 plus a personality disorder (other than antisocial personality disorder) or two ICD–10 ¹² /DSM–IV ¹³ personality disorders;

3. (c) for there to be a functional link between dangerousness and personality disorder (in two or more offences/prison behaviour).

What was also implied, although not stated explicitly, was that treating the severe personality disorder would reduce the risk of dangerousness.

In retrospect, anchoring these admission criteria to established psychiatric nosologies was probably a hostage to fortune as it would allow critics to scrutinise the criteria against scientific standards – a scrutiny precluded by the vagueness of its legal predecessor, psychopathic disorder. Although some might see this as unfair, Popper ^{Reference Popper1} would see this advance as, at least on scientific grounds, ‘A theory which is not refutable by any conceivable event is non-scientific. Irrefutability is not a virtue of a theory (as people often think) but a vice. Every genuine test of a theory is an attempt to falsify it, or refute it. Testability is falsifiability’. Hence, the fact that these criteria come up short (as I shall argue below) is therefore not a weakness of the DSPD initiative, rather a strength.

The second positive feature of the proposal was that the needs of a marginalised group that were hitherto largely ignored by psychiatrists in particular, were now centre stage in service providers’ agendas. To improve their assessment and treatment, the government provided specialised units with generous (some might say overgenerous) funding. (The sheer scale and generosity of this provision probably backfired as some saw it as excessive, focusing on the few, rather than the many that languished in prison or in the community with little or no input.) On a more positive note, however, this new interest in personality disorder had effects beyond DSPD, highlighting the many non-forensic individuals with personality disorder in the community, whose needs again had hitherto been largely ignored. The publication of the influential National Institute for Mental Health in England document Personality Disorder: No Longer a Diagnosis of Exclusion, ¹⁴ for example, spurred some imaginative initiatives to provide for this group. In addition, there were broader policy initiatives including an educational initiative funded by the Department of Health that focused on training staff with a range of expertise on personality disorder, ¹⁵ together with the publication of National Institute of Health and Clinical Excellence guidelines on borderline personality disorder ¹⁶ and antisocial personality disorder. ¹⁷ It is difficult to believe that these initiatives (many of which were in mainstream psychiatry) would have occurred without the DSPD initiative.

The third positive feature of the DSPD proposal was that the government funded several separate research initiatives, including embedding research units within the services themselves, so that the programme ‘would be firmly grounded in evidence from research and capable of adapting over time as new research comes forward’. ¹⁸ These included not only a commendable evaluation of the programme by independent researchers, but a number of ancillary projects whose benefits are already apparent with several theoretical and empirical publications in mainstream scientific journals.

Negative features

Despite its fine aspirations to base the proposal on evidence, the 1999 Home Office report acknowledged that the process of incarceration could not ‘be delayed until the outcomes of research are known’. ¹⁸ The entry criteria to the units reflected this haste. For instance, the DSM and ICD criteria, set out to categorise those with a severe personality disorder, not only lacked any scientific credibility but also ignored other empirical work that was already available (e.g. Tyrer & Johnson ^{Reference Tyrer and Johnson19} ). Even a high score on the PCL – perhaps the most acceptable marker of severity in this group – has not been found to be suitable; it is especially unreliable regarding the crucial cut-offs of 25 and 30, as the measure is not normally distributed. ^{Reference Cooke and Michie20} Hence, neither of the criteria to define severity of personality disorder in the DSPD proposal are satisfactory.

Second, there are also problems with the other two entry criteria. With regard to the risk (or dangerousness) criterion, although research over the past decade has shown that many of the risk measures have a moderate ability to predict future risk of violence for groups with little to choose between them, ¹⁷ it has also shown – and this is critical for the DSPD initiative – that what predicts future dangerous behaviour is not the personality features of the scales, rather (for men at least) it is their criminological and impulsive characteristics. ^{Reference Yang, Wong and Coid21} In addition, although extrapolating group data to an individual case presents problems in every area of medicine, the recent work of Cooke & Michie ^{Reference Cooke and Michie20} has shown just how much of a concern this is, as the width of the confidence interval surrounding such decisions in individual cases based on the PCL–R is very large (i.e. the estimates are highly likely to be inaccurate). This is especially important, as the decision-making in DSPD (as in every clinical decision) concerns applying evidence from groups to an individual case. Finally, demonstrating that there is a functional (or evidential) link between the severe personality disorder and the dangerousness is a demanding criterion and one that hitherto has proved to be elusive. ^{Reference Duggan, Howard, McMurran and Howard22}

Third, some have argued that the ‘treatments’ employed to reduce severe personality disorder and dangerousness were very heterogeneous, often with little theoretical rationale for their application. ^{Reference Tyrer, Cooper, Rutter, Seivewright, Duggan and Maden23} Consequently, it was often difficult to determine when an individual ought to move from the DSPD service to a lower level of security on the basis that their treatment needs had been met and their risk reduced. This asymmetry between the criteria of entry to the units (that were partly based on an individual's mental health needs) and those for their exit (that were largely risk based) will inevitably distort and slow the future movement through the units. The worry therefore is that the process of identifying individuals as being at high risk will lead to the units silting up with individuals that other services are reluctant to accept. It is only fair to say that this is my personal view and that some of those who worked within the programme argue the opposite (i.e. that the treatments are very similar, with a cognitive–behavioural therapy focus) and that this clarity facilitates the movement of individuals to lower levels of security.

Finally, what about the overall outcomes from the project? Again, views are mixed, and without long-term follow-up data it is difficult to judge whether or not the initiative is successful. It is certainly the case that those admitted to the DSPD services were those who in the past were resistant to any intervention and were therefore a challenging group to manage. That this has been largely achieved is no mean feat. Nonetheless, the emergent findings from independent and internal research evaluations, so commendably funded by the government, are not especially flattering in either effecting therapeutic change, ^{Reference Draycott, Kirkpatrick and Askari24} the organisation of the services ^{Reference Tyrer, Cooper, Rutter, Seivewright, Duggan and Maden23} or to their cost-effectiveness. ^{Reference Barrett, Byford, Seivewright, Cooper, Duggan and Tyrer25}

Conclusions

As the dust begins to settle on the many controversies surrounding DSPD, what is one to make of the initiative as a whole? Was it, as many of its critics allege, a knee-jerk response by government and policy makers to the ‘something must be done’ brigade when faced with a serious untoward event? Or was it an initiative that, despite its flaws, had several unexpected benefits? I believe that both are true. The assumption that the severity of the personality disorder is causally linked to violence and that treatment would reduce this link now appears to be particularly weak, as is the wide margin of error at an individual level on deciding on the risk that the individual might pose in the future. On the other hand, the initiative has directed attention to a hitherto poorly provided group and extended it beyond mentally disordered offenders. It also funded some badly needed research in this area.

The question remains, however, that in agreeing to implement the DSPD Programme, did psychiatrists and other mental health professionals form a Faustian pact with the government, gaining significant resources by promising to protect the public but at the price of departing from acceptable clinical practice? Or is this portrayal an exaggeration, as, after all, is this not what forensic psychiatry is really about (i.e. relatively well-resourced services that are increasingly predicated on protecting the public)? Before one dismisses this proposition out of hand, consider the following: (a) the virtual disappearance of patient confidentiality where information-sharing with Multi-Agency Public Protection Panel (MAPPP) and other criminal justice agencies is now the norm, rather than the exception; and (b) are we sufficiently conscious of the very wide margin of error when we detain someone against their will on the basis of risk that they might pose to themselves or to others? Although the DSPD initiative has been criticised (rightly) in riding roughshod over many aspects that professionals regard as important, is there not a nagging suspicion that in many respects it was doing explicitly what many psychiatrists practise implicitly? If it draws our attention to some of the important dilemmas that psychiatrists (and other mental health professionals) wrestle with, this will be an important legacy of the initiative.

The DSPD initiative therefore focuses particularly the minds of those who work at the interface of the criminal justice and mental health services on how they manage the conflicting demands of satisfying these two ‘cultures’ (i.e. being an agent of the state to safeguard public safety and/or a provider of services to those with mental distress) Currently, the accepted wisdom is that one is able to achieve both. If, however, one of the consequences of the DSPD Programme is that severe personality disorder is shown to contribute only a small proportion of the variance to violent behaviour, will this result in resources again being diverted away from the health needs of a very marginalised and poorly provided for group. And will anyone care?