Vernon L. Quinsey, PhD
Professor of Psychology and Psychiatry, Queen's University at Kingston, Ontario
Marnie E. Rice, PhD
Associate Professor of Psychology, Queen's University
Professor of Psychiatry and Behavioural Neuroscience, McMaster University
Professor of Psychiatry, University of Toronto
and
Grant T. Harris, PhD
Associate Professor of Psychology, Queen's University
Associate Professor of Psychiatry, University of Toronto
Regehr, Edwardh, and Bradford (1) discussed ethical issues in the conduct of research involving forensic patients. We agree that adhering to ethical priciples is important in forensic research, the clinical management of forensic patients, and achieving an appropriate balance between public safety and the civil liberties of offenders. However, research that is entirely archival and does not involve experimentation or even participation by subjects is quite different, especially with respect to informed consent, from the "research" considered in the Nuremberg Trial as raised by Regehr et al. The Canadian Tri-Council Policy Statement (2) appropriately allows Research Ethics Boards to waive requirements for informed consent provided that they are satisfied the research involves minimal risk; the waiver or alteration is unlikely to adversely affect the rights and welfare of the subjects; the research could not practicably by carried out without the waiver or alteration; whenever possible and appropriate, the subjects are provided with additional pertinent information after participation; and the waivered or altered consent does not involve a therapeutic intervention. In this context, we also note that the Mental Health Act of Ontario does not require patients' consent when archival and clinical record material are used for research as long as clinical information is not disclosed.
In order to clarify the issues raised by Regehr et al., we provide some context to and a critique of the case of the forensic psychiatric client they used to introduce their discussion of the ethical issues in research and clinical application.
The first point to clarify is the distinction between research projects resulting in the development of an actuarial risk assessment technique (the SRAG, now called the Violence Risk Appraisal Guide, VRAG), on the one hand, and the subsequent application of that instrument to individual forensic patients, on the other. The research leading to the VRAG was entirely archival. With respect to all our research subjects, research data have never been disclosed. Moreover, Regehr et al. stated that their client had not been released; by definition then, their client has never been a subject of any our follow-up research on the development of the VRAG.
Regehr et al. critically noted that consent was not obtained for the scoring (i.e., application) of the actuarial instrument from their client's clinical file. However, as a matter of law, neither the Ministry of Health nor its hospitals require consent for hospital staff to examine a patient's file, make judgments or score instruments from that file, use this information for program development, or communicate this information to the Ontario Criminal Code Review Board. Whatever materials appear on the 'record' of Review Board proceedings, including all exhibits, are also available to the general public, absent a specific order to the contrary made during the hearing in question. In this and all related matters, we have met our ethical and legal obligations to maintain the confidentiality of clinical information.
Although a patient's consent might be required for an interview, it is not true, as stated by Regehr et al., that an interview is required to score the Revised Psychopathy Checklist (PCL-R). The manual (3) states on page six that "Valid PCL-R ratings may be made solely on the basis of collateral information if there is sufficient high-quality information available." Indeed, PCL-R data from hundreds of cases obtained without interview are included in the manual. More importantly, the excellent predictive accuracy of the SRAG/VRAG assessment technique has been repeatedly demonstrated using, as one of its items, the PCL-R scored without interview (4).
Regehr et al. accurately stated the SRAG/VRAG comprises static, historical items and is not designed to measure fluctuations in risk. Because actuarial instruments based on static information provide the most accurate estimates of long-term risk currently known (5), we do not believe that it is sound practice to disregard an actuarial estimate of long-term risk of committing a new violent or sexual offense on the basis of clinical judgment. Forensic clinicians must manage the short-term risk of patients and use some dynamic short-term indicators to accomplish this. By definition, however, short-term, dynamic indicators cannot sensibly be used to modify or set aside estimates of long-term risk. Moreover, the accuracy of dynamic indicators has not been established even for short-term prediction. These observations are relevant to the interpretation of the "more advanced" neuro-diagnostic assessment results introduced on page 893 of Regehr et al.
The neuro-diagnostic assessment results seem to be adduced by Regehr et al. to set aside the actuarially derived estimate of risk. That is, the argument seemed to be that, if the offender had a neurological impairment, then the actuarial estimate of risk would be irrelevant. This argument fails on two counts: First, there are no empirically established estimates of risk based on neurological impairments or impairments of particular kinds (such as "personality change due to neonatal asphyxia"). Second, the relationship of neurological impairments to either actuarially estimated risk or violent recidivism itself is unknown.
Structural abnormalities of the brain are another sort of static variable. It is, however, possible to provide treatment for some symptoms caused by some structural abnormalities. Response to this sort of therapy can be treated as a dynamic variable. The fatal difficulty with Regehr et al.'s argument here is that the degree to which therapy lowers the likelihood of violent or sexual recidivism is unknown. Many clinical activities might affect the risk of violent adult offenders, but there are no scientific data about the direction, strength, or duration of such effects.
In conclusion, Regehr et al. (1) raised a number of ethical problems pertaining to research concerning forensic psychiatric patients. Although some of their concerns were valid, neither their client's case nor any of our research studies illustrate them.
References
1. Regehr C, Edwardh M, Bradford, J. Research ethics and forensic patients. Can J Psychiatry 2000;45:892-8.
2. Medical Research Council of Canada, Natural Sciences and Engineering Research Council of Canada, and Social Sciences and Humanities Research Council of Canada. Tri-Council policy statement; ethical conduct for research involving humans. Ottawa: Medical Research of Canada: 1998.
3. Hare RD. Manual for the Psychopathy Checklist (Second Edition). Toronto: Multi-Health Systems: 2003.
4. Quinsey VL, Harris GT, Rice ME, Cormier C. Violent offenders: Appraising and managing risk (Second Edition). Washington, DC: American Psychological Association: 2006.
5. Swets JA, Dawes RM, Monahan J. Psychological science can improve diagnostic decisions. Psych Sci in the Public Interest 2000;1:1-26.
Return