Whatever Treatment We Provide, We Must Measure Results Scientifically

Dr. Grant Harris

Lately I’ve written about some of our research and theories about psychopathy. Psychopaths are very different from other people even when very young. Aggressive antisocial behavior in early childhood is a key aspect of psychopathy.

As adults, psychopaths are sexually promiscuous, impulsive, dishonest, callous, selfish and much more violent than other people. Psychopathy is usually thought to be a mental disorder. But our research is showing that psychopaths do not seem to be disordered. Psychopathy does not seem to be caused by something going wrong in development.

But what treatment would change psychopaths? As you probably know, there was a special therapeutic community for mentally disordered offenders from the mid-1960's until 1978. The program was unique and many worldwide experts were impressed with its methods for psychopaths.

On the other hand, when we measured the results with a follow-up study, we found something different. The program seemed to make the psychopaths more dangerous rather than less. Psychopaths behaved much worse in the program but were just as likely to be made program leaders and get recommended for release. Experts’ impressions and the measurement of results gave exactly opposite conclusions.

New research by Michael Seto and Howard Barbaree adds to these data. They did a follow-up study of an up-to-date therapy for sex offenders in a Canadian prison. They compared psychopaths to other offenders and studied therapists’ ratings of how well the offenders had progressed in treatment. Clinicians’ ratings of progress were inversely related to recidivism -- offenders the clinicians rated as having done well were more likely to commit serious new offenses. And this was especially true among the psychopaths. Clincians’ impressions and measurement of outcome gave exactly opposite conclusions.

This research says very important things. First, our impressions about whether clients benefit from therapy cannot be trusted. There is no doubt that clinicians’ impressions are an inaccurate way to tell whether treatment works. The only accurate way is with formal, systematic measurement. Though this is now widely known, very few hospitals do good measurement. They almost always rely on informal impressions instead. Think about most of the clinical meetings or review board hearings you attend. How is the patient’s progress usually assessed and reported?

Second, it is possible to do harm. Well-intentioned interventions can actually make symptoms worse or make clients more dangerous. Sometimes this happens even though everyone is trying to give the best treatment. It can also happen though when clinicians do not know about the best treatments. And sometimes clinicians lack the skills to provide therapies properly. Because treatment effects are not systematically measured, the harm done often goes unnoticed.

Research such as this certainly can be discouraging. It’s understandable that clinicians are distressed by published findings showing their impressions are inaccurate and treatments harmful. But we hope this doesn’t mean the results will be rejected out of hand because that would make it likely that harmful practices and avoidable mistakes would continue. We hope our work encourages clinicians at this facility to seek new treatments for psychopaths and to measure progress objectively.

We suggest that insight-oriented, psychodynamic therapies should not be provided for psychopaths — doing nothing might be safer. We recommend that clinicians’ impressions should not be the main index of treatment progress because that leads to unnecessary errors.

But what should be done for psychopaths? We think that the right intervention might be different from things that work for persons who do have disorders. That is, our research suggests psychopaths don’t have damage that needs repairing. Perhaps the best program would make it very much harder for psychopaths to avoid the detection of antisocial acts. Maybe this could include intensive long-term supervision with very clearly specified behavioral requirements.

There should probably also be a system of well-understood incentives for complying with the terms of supervision plus penalties for noncompliance. Of course, it would be important to carefully train program staff to minimize the ease with which psychopaths could use manipulation and deception.

Another part of dealing with psychopathy could be public education — information and skills for everyone else on who might be a psychopath and what to do about a person who behaves that way. Social policies might help. Perhaps governments should have programs to increase social cohesion, build informal personal networks, strengthen neighborhoods, and enhance social stability. In theory, all this could reduce anonymity making it more difficult for psychopaths to exploit the rest of us, and make it easier for us to protect ourselves. As well, policies to limit inequity in the distribution of wealth might also shrink the "niche" for psychopathy.

No matter what is tried, however, all the available scientific knowledge says not to rely on impressions of effectiveness. We must all measure the results of our efforts systematically.