FORUM on Corrections Research

Effective sex offender treatment: The Warkworth Sexual Behaviour Clinic

The Warkworth Sexual Behaviour Clinic opened in 1989 and has since provided treatment to about 75 offenders per year. The treatment given in this program is intended to reduce the likelihood of offender recidivism -- especially violent or sexual recidivism.

The program uses a group therapy format, and is designed to fit into institutional work and job-site organization. Participants report to work five days per week throughout the five-month program.

This article sets out the preliminary results of a recent research project that attempted to evaluate both the clinic's risk assessment process and the effectiveness of the clinic's approach to treatment.

Program characteristics

The Warkworth Sexual Behaviour Clinic uses a relatively novel process for assessing offender recidivism risk. The pre-treatment assessment looks for a history of sexual offending, signs of deviant sexual arousal, a history of antisocial behaviour, and other indicators of antisocial personality and social competence problems. Research indicates that these four characteristics are predictive of recidivism among sex offenders.

The assessment then considers offender motivation for treatment and their degree of behaviour change during treatment, to incorporate the offender's progress into the risk assessment. Each offender's pretreatment risk assessment is ultimately combined with consideration of these two dynamic factors and overall clinical impression to determine an offender's overall post-treatment risk rating (ranging from low to high).

This risk assessment is used to identify offender treatment needs. It establishes treatment targets relevant to the offender's risk of recidivism, and helps indicate how long the course of treatment should be -- the greater the risk, the longer treatment should last.

The program tries to address the specific needs of each offender by providing treatment when it is most likely to produce behaviour change and to lead to a safe release, that varies according to changes in offender needs, and that provides a variety of treatment opportunities to allow offenders to engage the programming interactively.

Treatment groups meet for about three hours a day, five days a week. Group therapy is supervised by one senior therapist, with regular visits by the program director. Groups typically include 10 offenders who are in treatment at the clinic for the first time, but may also include additional offenders who are being treated for the second or third time.

After treatment is completed, a report is prepared detailing the offender's risk assessment, progress during treatment, offence cycle and relapse-prevention plan. The report also recommends assistance to meet offender post-release needs, discusses treatment targets that raise concerns, and refers the offender to a community-based treatment program (where warranted). This report is eventually sent to the National Parole Board.

Study sample

The 250 offenders who had received treatment at the clinic by the time of this study included 123 rapists, 56 incest offenders, 56 extra-familial child (younger than 14) molesters and 15 offenders convicted of a sex-related homicide. All offenders consented to the use of their information for research purposes as part of their consent to assessment and treatment at the clinic.

Information was drawn from institutional files, semi-structured interviews with the offenders, psychological and phallometric testing, and pre- and post-treatment reports. The earliest offender releases after treatment were in 1989, allowing for a maximum follow-up period of six years.

Of the original 250 offenders, 193 completed treatment. There was no significant variation among offender types as to the proportion completing treatment.

Risk assessment

Incest offenders were assessed as presenting significantly less pretreatment risk than the other sex offender groups. However, there was no significant variation among offender types as to their overall post-treatment risk scores. The average risk scores of the incest offenders did not change during treatment, while the average scores of the other groups decreased slightly.

Conditional release outcome

Information from National Parole Board files was obtained for 215 of the sample offenders. Files for the remaining 35 offenders were unavailable for a variety of reasons. Of the 215 offenders in this follow-up group, 17 were ineligible for conditional release during the time-frame of this study. Therefore, only 198 offenders were eligible for conditional release.

Two thirds of these offenders received some form of conditional release, while the rest were detained in custody on their statutory release date. There was no difference between offender types as to the proportion who were detained. However, offenders assessed as more likely to reoffend after treatment were more likely to be detained. There was also a significant relationship between the recommended level of post-release management and detention.

Of the 132 offenders who were released, 32% failed on conditional release in some way -- a relapse for which no official action was taken, suspension of conditional release for breach of a condition, or complete revocation of conditional release. Rapists were more likely than the two groups of child molesters to fail on conditional release, although this difference was not statistically significant.

The average time at risk in the community before conditional release failure was approximately 43 months, with a range from one week to 5.2 years. Survival analysis revealed that 29% of the rapists had failed on conditional release within one year, but only 14.4% of the child molesters did the same. A similar pattern appeared over longer follow-up periods, indicating that the rapists failed at roughly twice the rate of the child molesters.

However, given that more than 50% of the offenders in each group were still at risk of failing because they had not completed their sentences, these results should be interpreted with caution.

In general, highly antisocial offenders who behaved poorly in group treatment were more likely to fail on conditional release.

Recidivism

A total of 218 of the offenders had been released from custody at the time of this study -- 132 on conditional release and 86 because their sentence expired. However, one offender died and 15 others were deported, so the following observations are based on 202 offenders.

Table 1

Recidivism and Sex Offender Type (202 offenders)
Sex offender type	Any type of recidivism	Sexual recidivism	Violent recidivism
Rapists	26	8	3
Sex Killers	1	1	0
Incest offenders	5	2	1
Extra familial child molesters	5	2	0

Of these offenders, 36 committed a new offence after release -- 13 committed a sex offence and four committed a violent offence (see Table 1). Rapists were most likely to commit a new offence of any kind and to commit a new sex offence.

However, no association was found between completion of treatment and recidivism. These rates are comparable to those of other large sex offender treatment programs.

Comparing treatment acceptors and refusers

National Parole Board data were available for a comparison group of 74 offenders who were offered treatment at the clinic but refused.

Of these offenders, 65 were released from prison -- 39 on some form of conditional release and 26 at the end of their sentence. Not surprisingly, those who refused treatment were 60% less likely to be granted conditional release.

Fifteen of the 39 treatment refusers who received conditional release failed in some manner. This proportion did not differ from the results of those who accepted treatment.

However, the average time these offenders were at risk was just 582 days (with a range of four months to 3.3 years), which was significantly less than the at-risk time of those who received treatment. The treatment refusers, therefore, may well have a higher failure rate than the treatment acceptors after an equivalent follow-up period.

A similar trend was revealed by survival analysis. While 77.8% of the treatment sample survived the first year of follow-up, just 61.1% of those who refused treatment did the same. A similar result was present over a two-year follow-up.

However, these results must be again interpreted with caution given the small sample of treatment refusers and the fact that more than half of each group of offenders were still at risk of failing on conditional release.

Approximately 18.5% of the 65 treatment refusers who were released committed a new offence after release. This proportion was the same in the group of offenders who received treatment, and there was no significant difference between the two groups as to the commission of a new sex offence.

What does it all mean?

These preliminary data indicate that the risk assessment completed at the Warkworth Sexual Behaviour Clinic is predictive both of decisions made by the National Parole Board and of failure on conditional release.

This preliminary evaluation also suggests that this treatment program is effective in reducing recidivism and helping offenders complete conditional release successfully.

Two other findings also deserve emphasis. First, there are sensible relationships between decisions made at different stages of this treatment process and offender case management, indicating that offender information is used systematically. Initial risk scores are based on historical factors drawn from sources such as file reviews and, although post-treatment risk scores are conservative because they are heavily influenced by these initial scores, they also reflect the offender's performance during treatment.

Similarly, recommendations for post-treatment management are informed by the post-treatment risk ratings, and National Parole Board decisions are influenced by these recommendations.

Finally, the relatively prominent role of treatment-process factors in predicting post-treatment outcome suggests that it could be important to consider treatment responsivity throughout the treatment process.

---

⁽¹⁾ 250 College Street, Toronto, Ontario M5T 1R8. For more detailed information, please see H. E. Barbaree, M. C. Seto and A. Maric, Working Papers in Impulsivity Research: Sex Offender Characteristics, Response to Treatment and Correctional Release Decisions at the Warkworth Sexual Behaviour Clinic (Toronto: Clarke Institute of Psychiatry, 1996).