Research Reports

Institutional Methadone Maintenance Treatment:
Impact on Release Outcome and Institutional Behaviour

2001 N^o R-119

ACKNOWLEDGEMENTS

We would pke to acknowledge several individuals who assisted in this research.First we would pke to acknowledge Sandra Black for her help in identifying theMMT cases for the study. We would also pke to acknowledge the regional HealthCare representatives for their assistance in providing the data for each region.Finally, Jason Wong, Nancy Sutton, and Janice Milpgan were critical in helping toorganize and clarify MMT start dates and related information. Without thecontribution of these individuals, this research would not have been possible.

EXECUTIVE SUMMARY

Heroin is a powerful and extremely addictive drug. Research has shown that oneof the best interventions for heroin addiction is Methadone MaintenanceTreatment (MMT). Positive benefits from the use of MMT include reductions inilpcit opiate use, HIV risk behaviors and drug and property-related crimes (eg.,Marsch, 1998). Correctional research on MMT has demonstrated a lowerprevalence of heroin injection, syringe-sharing (Dolan et al., 1998), increasedpkephood to apply for post-release MMT and other drug abuse treatment, andlower drug use and crime (Magura et al. 1993).

The present study compares post-release outcome and institutional behaviour ofMMT participants to a group of offenders who tested positive for heroin use whileincarcerated and were assessed as having a substance abuse problem (Non-MMT group). Overall, offenders participating in MMT had lower readmissionrates and were readmitted at a slower rate than the Non-MMT group. Within a12 month period, the Non-MMT group were 28% more pkely than the MMT groupto be returned to custody. Furthermore, the MMT group were less pkely to havebeen unlawfully at large (UAL) or in violation of an abstinence condition due toalcohol use while on conditional release than Non-MMT offenders. While theMMT and Non-MMT groups were similar in terms of time to new offence andnumber and type of new offences committed, the trend in the data was towards alower rate of reoffending for the MMT group.

In terms of institutional behaviour, the MMT group had a reduced rate of seriousdrug related institutional charges following initiation of the MMT. This pkelyindicates a decrease in drug seeking and drug taking behaviour among MMToffenders in comparison to Non-MMT offenders after MMT initiation.Compared to other offenders, the MMT offenders were spghtly older and had aspghtly lower criminal history risk. Over 80% of the offenders receiving MMT arein either the Ontario or Pacific Regions. The regional distribution of MMT andNon-MMT offenders indicates that the Pacific Region has the most serious heroinproblem, but that the Prairie Region may be the most in need of increased MMTparticipation.

Overall the study found that participation in an institutional MMT program had abeneficial effect on outcome following release. Additional research is needed toaddress issues such as continuation of treatment in the community and othercommunity safety benefits.

TABLE OF CONTENTS

• ACKNOWLEDGEMENTS
• EXECUTIVE SUMMARY
• TABLE OF CONTENTS
• LIST OF TABLES
• LIST OF FIGURES
• INTRODUCTION
  • Difference Between Heroin and Methadone
  • Effects of Methadone Maintenance Therapy
  • Purpose and Rationale
• METHOD
  • Sample
    • MMT Group
    • Non-MMT Group
  • Sources of Data
    • Identification of MMT Participants
    • Urinalysis
    • Demographic and Assessment Data
    • Outcome
    • Institutional Behaviour
  • Analyses
• RESULTS
  • Release Outcome
    • Survival Analyses
    • Fixed Follow-up
    • Revocations
    • New Offences
    • Summary
  • Group Characteristics
    • Demographics
    • Current offence
    • Summary
    • Pre to Post MMT Changes in Behaviour
• DISCUSSION
• Implications
• REFERENCES

LIST OF TABLES

• Table 1: Outcome for 6 month fixed follow-up period
• Table 2: Outcome for 12 month fixed follow-up period
• Table 3: Type of revocation
• Table 4: Type of new offence
• Table 5: Racial distribution of Non-MMT and MMT groups
• Table 6: Gender distribution of study groups
• Table 7: Age at release for study groups
• Table 8: Regional distribution of study groups
• Table 9: Current offence types for study groups
• Table 10: Risk and need levels for study groups
• Table 11: Needs indicated for Non-MMT and MMT groups
• Table 12: Pre to post measures (rate/month) of institutional behaviour

LIST OF FIGURES

• Figure 1: Proportion of offenders remaining in the community
• (not readmitted) for the MMT and Non-MMT groups
• Figure 2: Proportion of offenders not reoffending for the MMT and
• Non-MMT groups

INTRODUCTION

Heroin is one of the most addictive and damaging illicit drugs. Addiction toheroin can occur after only a few uses and once addicted, it is extremely difficultto stop using the drug. The damage associated with an addiction to heroinincludes loss of family and friends, dependence on a criminal lifestyle to supportthe purchase of heroin, and serious deterioration in health. The most serioushealth consequences include HIV and Hepatitis C infections from sharingneedles and other drug taking paraphernalia. These medical conditions lead toserious health consequences that may ultimately result in death.Treatment for a heroin addiction is extremely difficult, but efforts to control theeffects of heroin by substituting the drug methadone and stabilizing the behaviourof the addict has resulted in successful interventions. Methadone has been usedas an intervention in the community since the mid 1960s by Dole and Nyswander(1965). One of the earliest reports of the use of methadone in correctionalsettings was in Lexington, Kentucky, in the 1950s where methadone was used asan experimental medication to treat heroin addiction (Parrino, 2000). Since theearly 1970s, incarcerated heroin addicts in New York city's Rikers Island havebeen receiving methadone (Magura, Rosenblum, Lewis, & Joseph, 1993). It hasbeen used in other correctional settings since that time (Darke, Kaye, & Finlay-Jones, 1998; Dolan, Wodak, & Hall, 1998; Motiuk, Dowden, & Nafekh, 1999).In January 1998, the Correctional Service of Canada (CSC) implemented Phase1 of a National Methadone Maintenance Treatment (MMT) Program for federaloffenders with heroin or other opioid addictions (Correctional Service Canada,1999). Phase 1 was designed to continue methadone treatment that began inthe community. The initial eligibility criterion was expanded from continuity ofparticipation to recent participation in a community MMT program to adjust fordelays in processing cases through the judicial system. CSC's MMT programintegrates the medical management of the offender with existing institutional andcommunity based substance abuse treatment programs.

In March 1999, Phase 1 of the National MMT Program was modified to allow, inexceptional circumstances, the option of providing methadone treatment toseverely heroin-addicted offenders presently not eligible for MMT. To be eligible,the following criteria must be met: all available treatments and programs havefailed; the health of the offender continues to be seriously compromised byaddiction; and there is a dire need for immediate intervention (CorrectionalService Canada, 1999).

The goal of the National MMT Program in CSC is to minimize the adversephysical, psychological, social, and criminal effects associated with opioid use,including the spread of HIV and other infectious diseases in CSC operationalunits (Correctional Service Canada, 1999).

Research on the impact of MMT will identify the possible benefits of MMT and itspotential contribution to community safety. MMT is an expensive program thatrequires considerable economic and human investment. Knowledge about theoutcome of the program will provide decision-makers with the information theyneed to evaluate the potential impacts of an expanded program that couldaddress the needs of offenders who have not previously been on an MMTprogram. The current study is the first step in developing the requiredinformation.

Difference Between Heroin and Methadone

One of the conflicts surrounding the use of methadone concerns the issue ofsubstituting one addictive drug for another. Contrary to the longstandingmisconception that methadone produces many of the same negative effects asheroin, methadone has been shown to be medically safe (Parrino, 2000).Heroin, considered a short-acting narcotic, has an immediate onset of action withfour to six hours duration. The route of administration of heroin is typicallyinjection, snorting, or smoking. Conversely, methadone has a duration of actionlasting from 24 to 36 hours and is administered orally (O'Brien, Cohen, Evans, &Fine, 1992).

Unlike heroin, methadone does not produce a narcotic effect once an individualis stabilized with the appropriate dosage. Indeed, the benefit of using methadonefor heroin addicts is that methadone acts to eliminate withdrawal symptoms whenthe concentration of the opiate (heroin) in the body drops below a specific point.It has been noted that withdrawal from methadone is less severe than withdrawalfrom heroin, it is more extended and can be controlled by the slow reduction inmethadone dosage over time (Parrino, 2000). This occurs because methadoneand heroin both act upon the same opiate receptors. Therefore, with a relativelysteady concentration of methadone in the blood, the opiate receptors remaincontinuously occupied and the patient becomes functionally normal.

Effects of Methadone Maintenance Therapy

Opioid substitution with methadone is the most widely implemented treatment forheroin addiction (Marsden, Gossop, Farrell, & Strang, 1998). There has beensome controversy surrounding whether methadone should be applied using anabstinence-oriented or a maintenance approach. Caplehorn (1994) examinedthis issue with a sample of participants in an alcohol and drug program. Resultsshowed that participants who were assigned to an abstinence-oriented programwere significantly more likely than those assigned to indefinite maintenance touse heroin and amphetamines during the first 2 years of methadone treatment(Caplehorn, 1994). Thus, it would appear that there are merits to using a longtermmaintenance approach, as opposed to an abstinence approach, with theadministration of methadone.

Research has also been conducted to determine the effects of dosage andlength of time in treatment on outcome. Overall, 60 milligrams of methadone perday has been identified as the lowest daily dose that is efficacious. Bellin,Wesson, Tomasino, Nolan, Glick and Oquendo (1999) compared inmatesaccepting high dose (60 mg. or more) and low dose methadone whileincarcerated on the time from release from prison to the community untilreincarceration. They found that inmates discharged on high dose methadonewere less likely to be reincarcerated than those on low doses, with the mediantime to reincarceration of 253 and 187 days respectively. Furthermore, use ofhigher doses of methadone has been found to be associated with lengthierretention in treatment, while maintenance on lower doses of methadone (oftendue to policy rather than dose being determined on an individual basis) tended tolead to increased drop out from treatment (Ward, Mattick & Hall, 1992). In termsof the effect of treatment length on outcome, Ward, Mattick and Hall (1992)concluded that length of time spent in methadone maintenance is related to postmethadonemaintenance behaviour. In addition, results from studies suggestthat more than two to three years of methadone maintenance is necessarybefore significant behavioural changes will occur, but that arbitrarily limiting theduration of methadone maintenance to such time periods has been found to havenegative consequences (Ward, Mattick & Hall, 1992).

Researchers have also found that ancillary services in addition to MMT such asprovision of medical services, frequent and high quality counseling and financialservices were related to better outcomes including increased retention intreatment, lower rates of drug use and reduced crime (Ward, Mattick & Hall,1992). For example, Simpson, Joe, Dansereau and Chatham (1997) found thatenhanced counseling and length of time spent in treatment in addition to MMTwere related to treatment outcomes.

Research has shown that being prescribed methadone in the community, thendiscontinuing that prescription on entry to prison was associated with anincreased likelihood of needle sharing (Shewan, Gemmell, & Davies, 1994). Thisfinding supports the need for MMT programs for incarcerated individuals as amethod for dealing with the negative health consequences of needle sharing.Many researchers have demonstrated the beneficial effects of participation inmethadone maintenance programs. For example, Maddux and Desmond (1997)found that among participants on methadone, days of intravenous drug use,crime, and needle sharing decreased markedly and days of productive activityincreased from the month preceding admission to methadone maintenance to themonth preceding the first anniversary of admission. Parker and Kirby (1996) alsofound substantial reductions in illicit drug use and acquisitive crime (e.g., theft,burglary) when comparing a sample of MMT patients to a community sample ofheroin only and poly-drug users.

Magura et al. (1993) compared a sample of offenders from New York city whowere enrolled in the MMT program with a control group of similar addicts whoreceived seven-day heroin detoxification in jail. Results demonstrated that MMTparticipants were more likely than controls to apply for methadone or other drugabuse post-release treatment and to be participating in treatment after a 5-monthfollow-up. Moreover, being in treatment at follow-up was associated with lowerdrug use and crime. It should be noted that the in-jail MMT program was mosteffective in maintaining post-release continuity of methadone treatment foroffenders previously enrolled in methadone at arrest (Magura et al. 1993).Dolan et al. (1998) conducted a study with a prison population in New SouthWales. They found that participants who had been maintained on methadonereported a significantly lower prevalence of heroin injection, syringe sharing, andscored lower on an HIV Risk-taking Behavioural Scale than participants whoreceived standard drug treatment and time-limited methadone treatment.In a comprehensive study, Coid and his colleagues (2000) found thatparticipation in a community Methadone Treatment Programme contributed todecreases in self reported illicit drug consumption and criminal activity. Overall,use of heroin decreased by 50% from intake to 6 months after beginningtreatment, although there was no change in the levels of illicit methadone,amphetamines, barbiturates, crack cocaine, powder cocaine, cannabis, orbenzodiazepine misuse. Methadone treatment was associated with a fall in thelevel of financial gain from criminal activities, and a decrease in the number ofreported arrests by the police during the treatment period. However, differentialimpact of MMT on different categories of crime were observed: burglaries andthefts were reduced by half and the effect on drug dealing was even greater. Incontrast, there were no changes reported in the level of fraudulent activities,acquisitive crime (such as mugging), sex work, or obtaining benefits illegallywhile working. Furthermore, MMT also appeared to have had the greatestimpact on those individuals who were most heavily involved in crime, withdramatic reductions observed in the crimes committed by those who had beenthe most criminally active.

Marsch (1998) conducted a meta-analysis examining the efficacy of methadonemaintenance interventions. The results demonstrated a consistent, statisticallysignificant relationship between MMT and the reduction of illicit opiate use, HIVrisk behaviors and drug and property-related crimes. Specifically, MMT had amoderate effect in reducing opiate use and drug and property crime, and a smallto moderate effect in reducing HIV risk behaviours (Marsch, 1998).

It should be noted that there is a paucity of research on the effectiveness of MMTon behavioural outcomes in Canada and among correctional populations.Fischer, Gliksman, Rehm, and Medved (1999), who followed-up communityopiate users over one year, conducted one of the few Canadian studies. Theyfound that non-MMT participants, in comparison to MMT participants, weresignificantly more likely to be involved in: illegal activities as an income source(52% versus 2%), illicit drug market activities (56% versus 17%), heroin use(65% versus 34%), other opiate use (65% versus 38%), alcohol use (70% versus45%), and benzodiazepine use (45% versus 28%). Although there was a trendfor MMT participants to have decreased health care utilization (e.g., emergencyservices, hospitalization) in contrast to non-MMT participants, it was notstatistically significant. Similarly, incarceration rates were comparable for the twogroups, as were arrests for drug and property offences during the past year.These results should be interpreted in light of the study limitations. First, thesample was composed of admitted drug users, but not offenders per se; thus, theresults involving criminal justice system involvement may be different using ahigher risk sample. Second, the sample size in the two groups was small (29MMT, 40 non-MMT), which may have attenuated the relationships examined.Stones (1999) examined a group of 37 Canadian federal offenders who initiatedMMT upon release to the community. The results demonstrated that 57%(N = 21) of MMT cases were revoked during the course of treatment, while 43%(N = 16) succeeded on parole or statutory release; however, the length of followupwas not specified. More importantly, the results showed that 65% ofoffenders on MMT in the community markedly reduced or altogether ceased theirchronic heroin use. While these findings are encouraging, it should be noted thatthis group of offenders commenced methadone while in the community; hence, itis possible that they were not stabilized on the proper dosage during the firstportion of their release.

Building upon the work of Stones (1999), Motiuk et al. (1999) conducted apreliminary investigation of the post-release outcome of offenders who initiatedCSC's MMT program while incarcerated. They found that 8.6% (N = 3) of MMTparticipants returned to federal custody; one for a new offence and the other twofor technical violations. The authors concluded that "the very low rate ofrevocation for this group of higher-risk offenders is encouraging" (p. 4).However, these results were based on a small sample size (N = 35), a limitedfollow-up period, and no comparison group of offenders who were notparticipating in an MMT program but were heroin users.

Purpose and Rationale

The purpose of the current study is to examine the release outcome of offenderswho have participated in the institutional MMT program. The MMT offenders arecompared to a group who tested positive for heroin use while incarcerated andwho were identified as having a substance abuse problem, but who did notparticipate in the MMT program. It is predicted that offenders who participated inMMT will survive longer in the community than the comparison offenders withrespect to readmission following release.

The current study also examines the effect of MMT participation on institutionalbehaviour. Specifically, institutional misconducts and time spent in segregationis examined before and after MMT initiation. It is predicted that there will be areduction in misconducts and time spent in segregation after MMT.

METHOD

Sample

MMT Group

The MMT group consists of all 303 offenders identified as having received MMTin a federal institution from November 20, 1996 to October 20, 1999. Among these offenders, approximately 62% (187 offenders) were released from custody before May 15, 2000 and these offenders are used for the follow-up analyses.The earliest release date for these offenders was November 19, 1997 and the latest was May 8, 2000. On average, these offenders were at risk to be readmitted (the number of days from the release date until the cut-off date) of approximately 415.3 days (SD = 215.7), and a median of 397 days.

Non-MMT Group

Good research requires that treated groups be compared to similar groups thatdo not receive treatment. The most direct way to do this is through random assignment of participants to either the treated or untreated group. However, this is not considered ethical in most treatment research and alternatives must be used. The challenge in creating an appropriate comparison group is to identify people who match the treated groups in most characteristics, except that theyhave not received treatment. In many cases an ideal comparison group cannot be identified so compromises must be made between scientific rigor (perfectly matched groups) and the value a comparison can bring to a study. Other studies of methadone maintenance treatment have used comparison groups consisting of offenders who have received alternative therapies (Dolan et al., 1998; Maguraet al., 1993) but this type of group was not available in current study. It was decided that the key characteristic for members of the comparison group was that they be known heroin users and have a substance abuse problem. To identify heroin users, urine alysis data were examined and to identify a substance abuse problem offender in take assessment data were reviewed.

To be included in the Non-MMT group, an offender had to have at least one positive urine alysis result for opiates or opiates A (heroin metabolites) in random and systemic testing from January 1, 1998 to October 20, 1999. Offenders in theMMT group were excluded from the Non-MMT group. To confirm a drug problem, the offender intake assessment (OIA), and correctional plans were examined. If offenders did not have a substance abuse problem indicated, or if the urine alysis report specified that the offender was receiving Tylenol with codeine for pain relief, they were excluded from the Non-MMT group. There were 215 offenders in the Non-MMT group and approximately 52% (112offenders) were released from the institution prior to May 15, 2000 and could be used in the follow-up analyses. The earliest release date for these offenders wasFebruary 26, 1998, and the latest was April 3, 2000. On average, these offenders were at risk to be readmitted of approximately 383 days (SD = 202.7),and a median of 364 days. The average time at risk was not significantly different between the MMT and Non-MMT groups (t (1,297) = 1.28, ns).

Sources of Data

A number of different types of data are used in the study.

Identification of MMT Participants

Lists of offenders involved in MMT were forwarded to the researchers by regional health care representatives from each of the five CSC regions (Atlantic, Quebec, Ontario, Prairie, Pacific). Additional information such as dose level and methadone initiation date was also collected in two regions.

Urinalysis

To identify offenders for the comparison group, urine alysis data were examined.Under the urine alysis program, offenders may be tested for illicit drug use by screening their urine for the drug metabolites. The testing may be part of a random drug testing program used for program monitoring or for "reasonable grounds" where it is believed that the offender is intoxicated. Approximately 5% of all incarcerated offenders are tested in each month for the random testing program.

Demographic and Assessment Data

Demographic data such as age, race, number and type of current offences, and offender in take assessment information were obtained from the OffenderManagement System (OMS). The OMS is an electronic file system used to track all offenders in custody or under supervision of the Correctional Service ofCanada. Overall criminal history risk and criminogenic need are assessed as part of the the Offender Intake Assessment (OIA) process that is completed shortly after admission to prison by a parole officer at the receiving institution.Criminal history risk and criminogenic need are rated on a three point scale (high,medium or low). Furthermore, offenders are assessed in terms of seven criminogenic need domains: associates, attitude, community functioning,employment, marital/family, personal/emotional and substance abuse.Sentence and offence history information was also obtained from the OMS as well as important dates, such as release and admission dates.To determine how the MMT and Non-MMT groups compared with the general inmate population, a data set containing all offenders released in 1998 was created and used to calculate comparative statistics. A release data set was used because only MMT and Non-MMT cases that were released are included in this report. Characteristics of released offenders frequently differ from those for the institutional population because more offenders serving shorter sentences are in the release population while more offenders serving longer sentences are in the institutional population.

Outcome

For all offenders released prior to May 15, 2000, outcome following release was measured as any re admission to a federal correctional institution. For the study group, release date was the first date following the methadone initiation date. If the methadone initiation date was not available, the first release date afterJanuary 1, 1998 was used. For the comparison group, release date was the first release following the date of the positive urinalysis result. If multiple urinalysis dates were identified, one was randomly chosen to be the start point.Readmission includes both readmissions due to technical violations and readmissions due to the commission of a new offence. Additional information such as reason for readmission (new offence and/or revocation), and type of revocation were also coded from the Offender Management System.

Institutional Behaviour

Measures of institutional behaviour, such as number and type of institutional misconducts and time spent in segregation, were collected from the OMS. Forthe study group, this information was collected for a maximum of six months prior to and six months after the date of the initial methadone dose. Similarly, for the comparison group, this information was collected for a maximum of six months before and six months after the date on which the offender had a positive urinalysis result for opiates or opiates A. Not all offenders had a full six month pre and post treatment so results were converted to an average monthly number.Institutional charges were examined in terms of three general types: drug, violent and other. Drug disciplinary offences included possession of alcohol/drugs/drug paraphernalia, refusing to provide urine sample, failing urine sample, taking intoxicants into the body, involvement in drug trade. Violent charges included disrespectful/abusive to staff, fights/assault/threatens staff/inmates, and creates/participates in disturbance to jeopardize security. All other charges such as disobeys written rule/direct order, possession/deals in contraband, possession of unauthorized/stolen property, damages/destroys property were contained in the other charges category.

Analyses

The data were analysed using the Statistical Analyses System (SAS) Version6.12 (SAS, 1997) or Version 8.01 (SAS, 1999). Statistical analyses consisted off requency distributions and tests of statistical significance using the Chi square test. Survival analysis was used to evaluate outcome following release. Survival analysis evaluates the rate of survival (i.e. avoidance of negative outcome) at several points in time while taking into consideration time at risk.In addition, continuous variables such as age and number of offences were compared between the two groups using t-tests. Institutional measures were measured at two points in time and therefore a repeated measures analysis was employed.

RESULTS

The results of the analyses are presented in three sections. First, outcome following release is examined, followed by comparisons between released offenders in the MMT and Non-MMT groups in terms of criminal history risk and criminogenic need, offence type, and demographics. Finally, pre to post changes in institutional behaviour is examined.

Release Outcome

Survival Analyses

Survival analysis was used to compare MMT and Non-MMT groups in terms ofthe length of time they were able to remain in the community. Survival analysis makes more effective use of the available data and allows for longer follow-up periods than other methods when releases are spread over a long period of time thereby creating variable follow-up periods.Figure 1 shows the results of the survival analysis by presenting the proportion of offenders remaining in the community at each month following release for up to a28 month period. The differences in the curves for the two groups are statistically reliable (÷2(1) = 6.02, p < 0.05).

Overall, the MMT group was readmitted at a lower rate and more gradually than the Non-MMT group as shown by the slope of the curves. For example, at approximately 12 months after release, 59% of the MMT group had not been readmitted, compared to only 42% of the Non-MMT group. At 24 months, 47% of the MMT group had not been readmitted, while 34% of the Non-MMT group had not been readmitted. The success rate for the Non-MMT group was 13percentage points lower than for the MMT group, or in other words, the Non-MMT group were 28% more likely to be returned to custody than the MMT group.

A survival curve was also prepared for new offences. As can be seen in Figure2, there is only a small difference in the rate of new offending between the MMTand Non-MMT groups. The observed differences are not statistically reliable(÷2(1) = 1.81, ns), however, there are interesting patterns in the results. During the first 12 months after release, the MMT and Non-MMT groups are relatively similar in the proportion who commit a new offence, although the MMT group was more likely to be offence free at 12 months (82% for the MMT group vs. 76%for the Non-MMT group). The difference between the two groups in the rate of reoffending increases in later months, with approximately 74% of MMT off enders and 61% of Non-MMT offenders not reoffending 24 months after release.

Fixed Follow-up

Analyses using fixed follow-up periods provide opportunities for more detailed examination of the reasons for return to custody. The next set of results presents data for both six and twelve months fixed follow-up periods. It is not possible to extend these results beyond twelve months due to the small number of cases who had more than twelve months available between their release date and the end of data collection period. Readmissions were divided into three types,revocation without a new offence, revocation with a new offence and Warrant ofCommittal admission. Revocations occur while the offender is in the community under parole supervision and the conditional release is revoked by the NationalParole Board. Readmissions under a Warrant of Committal occur when the offender has completed a sentence, but commits a new offence resulting in anew period of incarceration in a federal institution.

Among the offenders in the two groups who could be followed for at least six months, 71% of the MMT group remained in the community, while only 56% of those in the non-MMT group were not readmitted (see Table 1). The MMT group was also less likely to have a revocation for a technical violation and less likely to have committed a new offence while in the community (revocation with a new offence or warrant of committal). The difference between the two groups in terms of outcome following release for a fixed six month period was statistically reliable (÷2(3) = 8.03, p < 0.05).

As was evident in the survival analysis, the MMT group was much more likely to remain in the community after 12 months (62%) than the Non-MMT group (39%).Consistent with the six month fixed follow-up results, the MMT group was less likely to be readmitted and was less likely to have committed a new offence(revocation with a new offence or warrant of committal) after release (÷2(3) =8.39, p < 0.05) (see Table 2).

Revocations

There are a number of possible reasons for offenders' conditional release to be revoked by the National Parole Board. For the entire sample of released offenders, Table 3 presents the reasons for revocation. It is important to note that offenders may have more than one reason for revocation of a conditional release so numbers and percentages in the table cannot be summed. Being unlawfully at large (UAL) and violation of an abstinence condition due to drug use were the most common reasons for revocation for both groups. However, theMMT group was less likely to have a revocation for either of these reasons. TheMMT group was also less likely to have a revocation for violation of the abstinence condition due to alcohol use (2% vs. 10%). Approximately 10% of both the MMT and non-MMT groups had their conditional release revoked for committing a new offence.

New Offences

Released offenders were also compared on the number and types of offences committed after release. The average number of new offences for offenders who committed a new offence was not significantly different between the MMT (3.1)and the Non-MMT groups (2.5) (t(1,48.7) = 0.83, ns).The types of offences committed by the MMT and Non-MMT groups are presented in Table 4. Overall, no significant differences were observed between the two groups in terms of the types of offences committed. However, the percentage of offenders with a new offence was lower for the MMT group (17% vs. 23%) than for the Non-MMT group.

Summary

Offenders in the MMT group were less likely to be readmitted and were readmitted at a slower rate than offenders in the Non-MMT group. Offenders in the MMT group were less likely to have their conditional release revoked because they were UAL or in violation of the abstinence condition (alcohol). In terms of new offences, the MMT group was less likely to have committed a new offence, but offenders in the two groups did not have significantly different survival curves for new offences, and were similar in the number of new offences and types of new offences committed.

Group Characteristics

The purpose of the following section is to provide a description of the offenders in the MMT and Non-MMT groups who were released during the study period.Comparisons between the groups will help to define differences which may impact the interpretation of the results and to build a profile of those offenders participating in MMT and those choosing not to participate, or who are unable to participate because of restrictions in Phase 1. Where possible, the MMT groups are also compared to all offenders released in 1998 to provide an indication of how they differ from the general offender population.

Demographics

There were half as many Aboriginal offenders in the MMT group (10%) than in the Non-MMT group (20%) as can be seen in Table 5. In 1998, 19% of the offenders released were Aboriginal indicating that the Non-MMT group is representative of Aboriginal population. However, the MMT group under represents the general population of Aboriginal offenders.

In terms of gender, results in Table 6 show that the Non-MMT and MMT groups did not significantly differ. Between 3% and 6% of offenders in the groups were women. In comparison, approximately 3% of offenders released in 1998 were women.

Offenders participating in MMT were, on average, 38 years old as compared to those in the Non-MMT group who were younger, at 34 years of age. As shown in Table 7, the MMT group was more likely to be in the age category above 35years of age (68%), while the majority of the Non-MMT group (59%) were 35years of age or younger. Offenders in the MMT group are also older than the general population of releases who had an average age at release of 34 years,and only 40% were over the age of 35.

The regional distribution of the MMT and Non-MMT cases is presented in Table 8.The Pacific region, had approximately 12% of releases in 1998, but 40% of theMMT offenders and 34% of the offenders in the Non-MMT group were in thePacific region. These data clearly show that the Pacific Region has the largest problem with heroin addiction, having almost three times the number of cases in the Non-MMT group as would be expected given the proportion of releases who were in the Pacific region in 1998. The Pacific region also has the second largest number of offenders participating in the MMT program in the country even though they have fewer offenders than three other regions. Ontario has the largest number of offenders participating in the MMT program, accounting for 42% (79) of the MMT cases. The Ontario and Pacific regions account for 81% of the MMTcases in the country.

While the Atlantic and Quebec regions have relatively small percentages of theMMT cases, they also have a small percentage of cases in the Non-MMT group indicating that heroin addiction is not a major problem in these two regions.However, the Prairie Region, with 26% of inmate releases in 1998, had approximately 29% of the cases in the Non-MMT group, but had only 10% of theMMT cases. The higher percentage of cases in the Non-MMT group suggests that heroin addiction may be an important problem in this region, but that access to MMT is limited. However, by relying on these data alone, it is not possible to determine if the access to MMT is more limited in the Prairie Region, or if offenders are not applying or being approved for the program. Phase 1 MMTrequires community availability of services to provide MMT prior to incarceration,and this may not be accessible in the Prairie Region.

Current offence

The MMT and Non-MMT groups did not differ in terms of the number of current offences with an average of six offences each on the current sentence. Table 9presents the types of offences for which offenders were sentenced. Given that offenders generally have more than one offence when admitted to prison, the percentages in this table will sum to more than 100. While the types of offences for which offenders in the two groups have been convicted do not differ a great deal, there are two important differences evident in the results. First, the MMTgroup is less likely to have been convicted of a sexual offence (0% vs. 8%) and they are less likely to have committed a violent offence (66% vs. 85%).

Criminogenic risk and need levels at admission are compared in Table 10.Offenders in the MMT group appear to have a slightly lower criminal history risk than those in the Non-MMT group. However, in terms of criminogenic need level,the two groups do not differ significantly. In comparison to all releases in 1998,the MMT group is similar in the level of criminal history risk, but the Non-MMTgroup is higher risk. Both study groups are higher need than the general population of releases.

More detailed results for each of the seven criminogenic need domains are presented in Table 11. The MMT and Non-MMT groups are very similar in terms of the needs identified, with the exception of the Marital/Family need domain where fewer offenders in the MMT group (58%) had a need indicated compared to the Non-MMT group (71%).

Summary

The comparisons between the MMT and Non-MMT groups indicated that offenders in the MMT group were older, less likely to be Aboriginal offenders, and had slightly lower criminogenic risk than offenders in the Non-MMT group. All other analyses indicated the groups were relatively similar and therefore appropriate for comparisons. In addition, the results indicated that the MMTprogram is largely concentrated in the Pacific and Ontario regions, while the data for the Non-MMT group suggest that heroin is a problem in the Ontario, Prairie and Pacific regions. The number of heroin users identified by positive urinalysis results in the Atlantic and Quebec Regions was relatively low.

Pre to Post MMT Changes in Behaviour

Offenders' institutional behaviour before and after the start of MMT was measured using misconducts and segregation time. The pre- and post-MMTperiods varied in length, to a maximum of six months, so data were converted toa rate or incidence per month in order to insure uniformity of measurement.Repeated measures analysis of variance (ANOVA) was used to test for differences between the MMT and Non-MMT groups. This analysis also tested for differences across the two time periods (pre and post) and the interaction of group by time factors. Results are presented in Table 12.

Differences between the MMT and Non-MMT groups were observed on a number of variables, but the only change associated with participation in theMMT was for serious drug charges. While, the MMT group had significantly fewer total institutional charges, fewer serious institutional charges and fewer periods of involuntary segregation than the Non-MMT group, none of these variables showed any differences in pre- post-comparisons of MMT initiation.The one exception was for serious drug charges. Not only did the MMT group have fewer serious drug charges than the Non-MMT group, there was also a statistically reliable decrease in serious drug charges from the pre-MMT period(0.11) to the post-MMT period (0.08). In contrast, the Non-MMT group had an increase in serious drug charges (0.16 to 0.41) over time.

These results may indicate that MMT participants have already begun to change their behaviour prior to starting MMT or that offenders applying for and receiving MMT have fewer behaviour problems while incarcerated. Behaviour change prior to participation in MMT could be part of the process of choosing to pursueMMT.

DISCUSSION

Previous research has indicated that participation in MMT has numerousbeneficial effects, both for the individual and society at large (Fischer et al. 1999;Maddux & Desmond, 1997; Marsch, 1998; Stones, 1999). The current studyfurther examined the impact of institutional MMT participation by comparing twogroups of offenders on outcome following release and institutional behaviourbefore and after MMT initiation. As predicted, offenders who had participated inMMT were less likely to be readmitted and were readmitted at a slower rate thanthe Non-MMT group. These results provide support for the need to initiate MMTin the institutional setting.

Among offenders who had a revocation (with or without offence), offenders in theNon-MMT group were significantly more likely to have a UAL or a violation of the abstinence condition due to alcohol use. In terms of new offences, the offenders in the two groups were similar in their survival curves, the number of new offences committed and the types of new offences committed.In order to ensure that group comparisons were valid, offenders in the twogroups were evaluated in terms of demographic characteristics, current offence,and risk and need. Overall, few notable differences were found, and therefore it is suggested that the outcome findings are valid.

Finally, the MMT and Non-MMT groups were compared in terms of institutional behaviour at two time periods (i.e. pre to post MMT) in order to determine if there were any positive effects of MMT on behaviour while incarcerated. Overall, few differences were observed, with the exception of the number of serious drug charges per month. Offenders involved in MMT showed a decrease in drug charges over time while offenders in the comparison group showed an increase.This finding suggests that offenders participating in MMT are less involved in the drug subculture, such as drug taking, drug seeking, and drug trafficking behaviours post MMT initiation. In regards to segregation periods, the MMTgroup spent less time in involuntary segregation than the Non-MMT group both before and after the initiation of MMT.

Although these results do not provide conclusive evidence that methadone serves to calm disruptive institutional behaviour, they do suggest that negative behaviours are at least maintained at a low level while offenders are on the MMTprogram. More importantly, it was demonstrated that there was a decrease forMMT offenders, relative to non-MMT offenders, in behaviours related to activity in the drug subculture.

The impact of MMT on offenders is most likely underestimated in this study.Offenders were included in the MMT group without regard for how long they remained in the program, how close to release they were in the program, and without any measure of their participation in other related program activities. In addition, participation in an MMT program after release was not monitored. TheMMT group was basically offenders who, for some period time while incarcerated, participated in an MMT program. In addition, the follow-up period was longer for the MMT group than the Non-MMT group.

Even with these limitations the MMT offenders did better than similar offenders who did not participate in the MMT program. If it is possible to demonstrate an effect with these minimal requirements for inclusion in the MMT group, more complete data on level of participation and participation in a community MMTprogram would likely yield a larger effect. Future research will address these issues.

Implications

The results of the present study suggest that MMT participation has a beneficial effect on post-release outcome in terms of readmission to a federal penitentiary. An important implication of these findings is that CSC may spend less money on these offenders in the long term. The cost of the institutional MMT program maybe offset by the cost savings of offenders successfully remaining in the community for a longer period of time than equivalent offenders not receivingMMT. In addition, health related costs such as treatment for HIV or Hepatitis C infection could be affected by MMT availability in prisons.

It should be mentioned that other measures of post-release outcome need to be examined in order to get a more complete picture of the effect of MMTparticipation on offenders' behaviour after release. These outcome measures may include such things as health care use, employment status and measures of substance abuse, as well as other measures. This will be assessed in a future study.

In terms of offender behaviour while incarcerated, few differences were observed. This may be due to a true lack of effect, or the inability of the current measures of institutional behaviour to detect changes over a 6-month period.Furthermore, it may be that offenders who apply for MMT have a waiting period prior to initiation of MMT that may be affecting the results. There was a difference, however, in terms of serious drug charges (e.g., possession ofalcohol/drugs/drug paraphernalia, takes intoxicant into body). Specifically, heroin-addicted offenders on methadone showed a decrease in drug activity behaviours, while Non-MMT offenders showed an increase over time. This finding has implications regarding the functioning of the institution. In particular,if offenders who are on methadone are not as involved in the drug subculture,there may be less danger to themselves and to the staff around them.It should be noted that the current study only examined institutional charges and segregation time, and other measures such as program participation were not addressed. For example, offenders who are involved in MMT may have increased ability to concentrate, which would in turn affect their ability to participate in useful activities such as employment, education and treatment. A future study will examine the effect of MMT participation on these other institutional behaviours as perceived by staff and offenders.

Most importantly, this research demonstrated that there are definite positive effects of MMT on offender behaviour, with the largest benefit existing following release. Future research is needed to further assess the degree to which institutional MMT is maintained upon release, and how this affects the long term functioning of heroin-addicted individuals.

REFERENCES

Bellin, B., Wesson, J, Tomasino, V., Nolan, J., Glick, A., & Oquendo, S. (1999). High dose methadone reduces criminal recidivism in opiate addicts. Addiction Research, 7(1), 19-29.

Caplehorn, J. R. M. (1994). A comparison of abstinence-oriented and indefinite methadone maintenance treatment. The International Journal of the Addictions, 29, 1361-1375.

Coid, J., Carvell, A., Kittler, Z., Healey, A. & Henderson, J. (2000). Opiates, Criminal Behaviour, and Methadone Treatment. Website: http://www.homeoffice.gov.uk/rds/index.htm.

Correctional Service Canada (1999). National Methadone Maintenance Treatment Program Phase 1. Resource and Information Package.

Darke, S., Kaye, S. & Finlay-Jones, R. (1998). Drug use and injection risk-taking among prison methadone maintenance patients. Addiction, 93(8), 1169-1175.

Dole, V. P., & Nyswander, M. (1965). A medical treatment for diacetylmorphine (heroin) addiction. Journal of the American Medical Association, 193, 80-84.

Dolan, K. A., Wodak, A. D., & Hall, W. D. (1998). Methadone maintenance treatment reduces heroin injection in New South Wales prisons. Drug and Alcohol Review, 17, 153-158.

Fischer, B., Gliksman, L, Rehm, J., Daniel, N., & Medved, W. (1999). Comparing opiate users in methadone treatment with untreated opiate users: Results of a follow-up study with a Toronto opiate user cohort. Canadian Journal of Public Health, 90, 299-303.

Maddux, J. F., & Desmond, D. P. (1997). Outcomes of methadone maintenance 1 year after admission. Journal of Drug Issues, 27, 225-238.

Magura, S., Rosenblum, A., Lewis, C., & Joseph, H. (1993). The effectiveness of in-jail methadone maintenance. The Journal of Drug Issues, 23, 75-99.

Marsch, L. A. (1998). The efficacy of methadone maintenance interventions in reducing illicit opiate use, HIV risk behaviour and criminality: A metaanalysis.Addiction, 93, 515-532.

Marsden, J., Gossop, M., Farrell, M., & Strang, J. (1998). Opioid substitution: Critical issues and future directions. Journal of Drug Issues, 28, 243-264.

Motiuk, L., Dowden, C., & Nafekh, M. (1999). Methadone Maintenance Treatment (MMT) programming for federal prisoners: A preliminary investigation. Ottawa, ON: Correctional Service Canada.

O'Brien, R., Cohen, S., Evans, G., & Fine, J. (1992). The encyclopedia of drug abuse (2nd ed.). New York, NY: Facts on File.

Parker, H., & Kirby, P. (1996). In B. Webb (Ed.),Methadone maintenance and crime reduction on Meyserside. London, UK: Crown.

Parrino, M. W. (2000). Methadone treatment in jail. American Jails, May/June, 9-12.

SAS Institute Inc. (1997). Statistical Analysis System (SAS), Version 6.12. Cary,NC.

SAS Institute Inc. (1999). SAS Procedures Guide, Version 8. Cary, NC.

Shewan, D., Gemmell, M., & Davies, J. B. (1994). Behavioural change among drug injectors in Scottish prisons. Social Science and Medicine, 39, 1585-1586.

Simpson, D. D., Joe, G. W., Dansereau, D. F., & Chatham, L. R. (1997). Strategies for Improving Methadone Treatment Process and Outcomes. Journal of Drug Issues, 27(2), 239-260.

Stones, G. (1999). Dynamic and static predictors of methadone maintenance treatment compliance among forensic clients. A Master's Research Paper, submitted to the Department of Applied Psychology, University of Toronto, Ontario Institute for Studies in Education, Toronto, ON.

Ward, J., Mattick, R. P. & Hall, W. (1992). Key Issues in Methadone Maintenance Treatment.New South Wales: University of New South Wales Press.