Since 1981, formalized treatment services for sex offenders have been provided at the Regional Psychiatric Centre (Prairies). The Clearwater Program, as it is known locally, has evolved into a six-month cognitive-behavioural package with a focus on relapse prevention.⁽⁸⁾

A comprehensive evaluation strategy, including a wide variety of theoretically relevant psychometric and behavioural measures,⁽⁹⁾ has been part of the program since its inception. These evaluations consistently show that program graduates demonstrate highly significant and positive changes over the course of treatment.

Since 1981, 250 sex offenders have entered treatment. Fifteen percent of these did not complete the program (i.e., completed less than four months of treatment). Most of these (approximately 70%) left after only a few weeks, claiming that they had "changed their mind" or that they "weren't ready for treatment." Most of the others who did not complete the program were asked to leave because of their aggressive and disruptive behaviour.

As of June 1990 when the data were gathered, 169 treated men had been released into the community. The average length of time since release was four years, with a range of four months to nine years. We note that all analyses reported below were repeated using only those men who had been released for at least one year. Findings from this reanalysis did not change any conclusions.

Four types of outcome were defined, based on information from Finger Print Service (FPS) records: no readmission to a correctional institution, revocation of mandatory supervision or parole without any additional charges or convictions, one or more convictions for non-sexual offences, and one or more convictions for sexual offences with or without non-sexual offences.

Outcome was also measured at two different times. In some analyses, we considered only the first event that brought the offender back to prison. In other analyses, we considered all offences (events) recorded after the offender left treatment (post treatment).

The outcome data are contrasted with several admittedly weak comparison groups. More appropriate comparison data are being gathered for a group of 90 men who were interviewed for the Clearwater Program but who did not receive treatment.

The treatment group was first compared to a group of 1,100 sex offenders released from federal institutions over a three-year period and followed up three years later.⁽¹⁰⁾ Note that in gathering this data, Porporino and Robinson reported only the first event after release. No group differences in sexual recidivism rates are apparent between the treated men (7.1%) and the national sample (6.2%).

However, if we consider only offenders who have a previous history of sex offences, the treatment sample shows 37% fewer sexual reconvictions than the non-treated sex offenders (with recidivism rates of 9.2% versus 14.6% respectively). This last finding is consistent with the conclusion of Andrews and colleagues⁽¹¹⁾ that correctional treatment has its greatest impact with higher-risk offenders.

We also looked at the outcome for the 25 released men who did not complete the Clearwater Program. The results suggest that these men represent a greater risk, and more effort should be made to keep them in treatment. Specifically, considering all events during the follow-up period, men who did not complete treatment were much less likely to avoid reincarceration (32% versus 53%). Furthermore, they were almost twice as likely as program completers to be convicted of at least one further sexual offence (24% versus 13.6%).

Our research indicates that all sex offenders are not equally likely to recidivate. As summarized in the following figure, pedophiles were most likely to be reconvicted of sex offences, while incest offenders were least likely to recidivate. Rapists were most likely to be reconvicted for non-sexual offences. Consistent with our earlier conclusions,⁽¹²⁾ recidivism rates for men who had assaulted both adults and children (rapist/pedophile in the figure) resembled those of rapists rather than pedophiles.

Not all pedophiles were found to be equally likely to commit another sex offence. If we look at all events after treatment, we find that more than half (57%) of bisexual pedophiles, compared with less than one fifth (17%) of heterosexual pedophiles and no (0%) homosexual pedophiles, committed further sex offences. This is consistent with the work of Abel and colleagues⁽¹³⁾ who found that, among child molesters, greater variety in the age and gender of victims was associated with higher rates of recidivism.

The present data provide preliminary evidence that treatment is affecting sexual recidivism among high-risk sex offenders. Firmer conclusions await analysis of our no-treatment group.

The Clearwater Program appears to be particularly effective with homosexual pedophiles but shows poorer results for men who have assaulted both boys and girls. Clearly we have to review and improve our approach to the latter group.

It would also seem that the program is particularly successful with incest offenders. However, this apparent success must be viewed in context: incest offenders in general, even if left untreated, have very low rates of recidivism.⁽¹⁴⁾

The recidivism rates reported by Porporino and Robinson⁽¹⁵⁾ for a largely untreated sample of sex offenders are lower than have been previously reported in the literature.⁽¹⁶⁾ These relatively low recidivism rates suggest that not all sex offenders may "need" treatment. Our task should be to improve identification of higher-risk offenders and give them priority for treatment. We recognize that imposing treatment on low-risk and low-need offenders can be counter-productive.⁽¹⁷⁾

The data from the Clearwater Program indicate that pedophiles, even if treated, present a greater risk to the community than other sex offenders. These men should receive the most extensive treatment available while still incarcerated. Following release, continued treatment under close supervision is also warranted.

At the other extreme, incest offenders present the least risk to the community. It might, therefore, be more appropriate to offer these offenders community-based treatment following release, which costs less than institutional treatment. Currently, though, incest offenders are as likely as pedophiles to receive treatment during incarceration.⁽¹⁸⁾

The rapists in our sample present a moderate risk of committing new sex offences but are more likely to commit additional non-sexual crimes. This pattern is consistent with the conclusion that, as a group, rapists are most similar to the general offender population.⁽¹⁹⁾ Sufficient treatment for most of these men may be provided through programs, such as anger management, already being offered in all federal institutions. Relapse-prevention training might well be useful for both rapists and non-sex offenders. The more violent and recidivistic rapist may require more intensive and specialized services, and his risk and needs should be carefully evaluated.

A final comment relates to postrelease follow-up. The relapse-prevention model of treatment does not attempt to "cure" sex offenders. Rather, offenders are helped to gain control over their urges and behaviours and to avoid future sexual deviance. It follows that offenders' ability to generalize and apply skills learned through treatment in the institution should be monitored and reinforced in the community by knowledgeable personnel.

In the Clearwater experience, the availability of treatment in the community has been very uneven. Some cities on the Canadian prairies are rich in resources to treat sex offenders. Other communities have few service providers who will deal with these people. Moreover, community therapists may adopt a variety of treatment models that may be at odds with treatment already completed. Thus, rather than being helped to refine and implement self-management skills, many offenders find themselves starting treatment over again.

Ideally, we must integrate institutional and community services to provide continuity and consistency of care. We believe that the recidivism rates found with the Clearwater Program can be reduced through more systematic and consistent community follow-up.

We turn now to a recent programming initiative in the Prairie region of the Correctional Service of Canada. The Parkland Wellness Program was born out of the need to increase services for sex offenders at Bowden Institution, a medium-security penitentiary with approximately 450 inmates.

The need for such services was obvious. In March 1991, there were 208 men in Bowden Institution whose current conviction was a sex offence. (This did not include men who were in Bowden for other crimes but had previous sex offences on record.) Of these 208 sex offenders, 167 had never been involved in any sex-offender treatment program. Bowden inmates do have access to specialized programs at the Regional Psychiatric Centre, but the demand for admission was such that only about 20 inmates per year could be treated there.

In brief, the Parkland Wellness Program is designed to improve services for sex offenders by coordinating the efforts of Bowden Institution, the Regional Psychiatric Centre (Prairies), parole officials and the community.

Sex offenders will be screened and assessed upon their arrival at Bowden to determine their treatment needs and risk level. The highest-risk offenders will be referred to the Regional Psychiatric Centre. Low-risk offenders will be geared toward earlier release with treatment being provided in the community. Moderate-risk offenders will receive most of their treatment at Bowden with an option for transfer to the Regional Psychiatric Centre if necessary. In all cases, community follow-up will be provided to ensure that the positive effects of treatment are not lost.

Repeated assessment of psychological and risk status will provide important information for decisions on treatment and release throughout the offender's sentence. All involved centres are committed to adopting compatible treatment and assessment procedures to ensure continuity of care. Such continuity and consistency will also facilitate information sharing, while providing a data base to allow for clinical and evaluative research.

Some of the issues and problems that the Parkland Wellness Program attempts to address include:

Providing all sex offenders with the type of intensive treatment available at the Regional Psychiatric Centre would be financially prohibitive as well as unnecessary. The triage approach to classifying risk and need (according to level of priority) will allow for more efficient use of resources. In general, we expect that pedophiles and more violent rapists will be prime candidates for the Regional Psychiatric Centre. Most rapists will be treated at Bowden, while most incest offenders will receive primary services in the community.

Treatment settings that deal with the same offender type often develop their own assessment and treatment procedures. Typically, there is little or no integration of approach across these various settings. This lack of consistency can be particularly frustrating for the offender, who must complete yet another set of psychological tests at each setting or who must start treatment anew from a different perspective.

The Parkland Wellness Program will address this problem in two ways. First, there will be consistent assessment and treatment provided to the inmate throughout incarceration and following release. Ideally, this treatment will start as soon as the inmate has been identified as a sex offender. From that point on, there will be a co-ordinated intervention throughout the offender's sentence. As 95% of those released Clearwater patients who reoffended did so before the end of their sentences, it was decided that follow-up should extend at least this long (until end of sentence). We recognize that even longer follow-ups would be beneficial for some offenders.⁽²⁰⁾
Second, assessment and treatment procedures must be complementary. All agencies involved in assessment and treatment of offenders in the Parkland Wellness Program will adopt the same core battery of psychometric, phallometric and related assessment instruments (e.g., risk assessments). Moreover, all centres will adopt a cognitive-behavioural relapse-prevention model of treatment.

To be successful, the Parkland Wellness Program must deal with at least three areas of accountability. 1) The inmate to the program Participants must be accountable for their effort in the program. It is not sufficient merely to attend groups. Rather, participants must demonstrate that they have absorbed the information provided and that they have begun to alter their thinking and behaviour patterns.

2) Program providers to funding sources It is not sufficient just to operate a program. Rather, we must be able to demonstrate that our efforts are both cost-efficient and effective. Because halfway-house placements are a fraction of the cost of institutional placements, treatment should be provided in the community whenever possible. Similarly, providing quality services in a main institution is considerably less expensive than transferring and treating an inmate in a regional psychiatric centre. As well, to increase timely releases to the community, we must work closely with parole authorities. In particular, to increase the National Parole Board's confidence in our recommendations, we must ensure that our assessments and treatment can be shown to be valid and effective.

3) The program to the inmate The triage method of delivering services, where different levels of risk and need are matched with different levels of programming intensity, ensures that offenders will have access to the most appropriate resources. Providing complementary services in the community should ensure that offenders who complete the necessary program while incarcerated will not have to duplicate their efforts on release. Rather, once inmates have adequately demonstrated competence in institutionally based programs, they can expect to receive community support in maintaining and implementing these skills.

The Parkland Wellness Program must demonstrate that its components are actually achieving their goals. The assessment and triage processes will be closely monitored to ensure that valid decisions are made in a timely manner. Each component of the program (e.g., anger management) will be evaluated to ensure that the expected psychological and behavioural changes are being achieved. Finally, the impact of the program will be assessed to determine whether recidivism rates are dropping.

Fundamental to the success of the Parkland Wellness Program is its commitment to both "pure" and treatment-outcome research. As just described, research on program effectiveness is an essential part of ensuring that the Parkland Wellness Program remains accountable. However, we also must conduct basic research on sex offenders to improve our treatment efforts. A comprehensive data base, now being implemented at Bowden Institution, will complement data bases at the Regional Psychiatric Centre and, ultimately, at community treatment sites. Although maintaining such a data base is resource intensive, it is indeed necessary for effective research in the future.

We believe that we can improve the delivery of services to offenders by ensuring that assessment and treatment strategies are consistent. Although this project is being piloted only on a limited scale at present, we will move toward more uniformity across the Prairie region.

The obvious advantages of this approach - for the sharing of information, non-duplication of services and research - are entirely consistent with the recommendations of the recent report of the Correctional Service of Canada's Task Force on Mental Health.