Evaluation of the Court Integrated Services Program Final report

Evaluation of the Court Integrated Services Program Final report Dr. Stuart Ross, Melbourne Consulting & Custom Programs December 2009

Table of Contents Table of Contents...2 Executive Summary...5 Program implementation findings...6 Program outcome findings...12 Recommendations...15 Part 1: Program description...20 Chapter 1 Introduction...20 1.1 Description of the Court Integrated Services Program...21 Chapter 2 Case flows & system load...24 2.1 Case flow and system load goals...24 2.2 CISP program flow processes...25 2.3 Referrals to CISP...26 Repeat clients...28 Source of referrals...28 2.4 Assessments and assessment outcomes...30 Assessment outcomes...30 2.5 Case recommendations and outcomes...31 Engagement rates...32 Consistency in allocation to program level...33 Allocation to program level in relation to clients assessed risk...34 2.6 Use of judicial monitoring...35 2.7 Program completions...36 Court outcome at CISP exit...38 Time on CISP...39 2.8 Commentary on CISP case flow and system load issues...40 Meeting program case flow targets...40 Case flow and judicial support...41 Part 2: CISP Case management and intervention model...43 Chapter 3 Client characteristics and service needs...44 3.1 Client demography...45 3.2 Prevalence of drug problems...47 3.3 Prevalence of alcohol problems...50 Melbourne Criminological Research and Evaluation 2

3.4 Prevalence of mental health problems, ABI and intellectual disability...51 Acquired brain injury...53 Intellectual disability...53 3.5 Combinations of problems...54 3.6 SF-12 survey of client physical and mental health...54 Pre-CISP physical health...55 Pre-CISP mental health...56 Comparing pre- and post-cisp scores...57 3.7 Comments on client characteristics and service needs...58 Chapter 4 CISP service model...60 4.1 Clinical and procedural basis of CISP...61 Multidisciplinary team approach...61 Workers roles...62 Team leaders...63 Program venues...64 4.2 Referrals and interventions...65 4.3 Client feedback...69 4.4 Support systems...71 Policy and procedures manual...71 Assessment process...71 Platypus Systems Case Management System...73 4.5 Recommendations:...74 Chapter 5 Drug and alcohol services...76 5.1 Provision of drug and alcohol services...76 5.2 Drug and alcohol service needs of CISP clients...76 5.3 Service output measures...78 5.4 Change in drug and alcohol problems during CISP engagement...80 5.5 Relationship between CISP and drug and alcohol agencies...81 5.6 Recommendations...83 Chapter 6 Mental Health and Acquired Brain Injury Services...84 6.1 Provision of mental health and ABI services...84 6.2 Mental health service outputs for CISP clients...84 6.3 Service relationships for mental health...87 6.4 ABI service outputs for CISP clients...88 Melbourne Criminological Research and Evaluation 3

6.5 Service relationships for ABI...90 6.6 Recommendations...90 Chapter 7 Accommodation support services...92 7.1 Provision of housing services...92 7.2 Housing service needs of CISP clients...93 7.3 Service output measures...94 7.4 Relationship between CISP and housing agencies...97 7.5 Access to housing services for CISP clients...98 Chapter 8 Magistrate perspectives on CISP and Therapeutic Jurisprudence... 100 8.1 Magistrates use of CISP... 100 8.2 Relationship to practice... 102 8.3 Satisfaction with CISP services... 105 8.4 Future directions for CISP... 108 8.5 Recommendations... 109 Chapter 9 Reoffending and compliance outcomes... 110 9.1 Outcome targets for CISP... 110 9.2 Outcome assessment methodology... 110 9.3 Bail compliance... 111 9.4 Order compliance... 111 9.5 Re-offending rates... 113 Proportion of recidivists... 113 Time to re-offence... 114 Melbourne Criminological Research and Evaluation 4

Executive Summary The Court Integrated Services Program offers a coordinated, team based approach to the assessment and treatment to defendants at the pre-trial or bail stage. It links defendants to support services such as drug and alcohol treatment, crisis accommodation, disability services and mental health services. The program commenced at the beginning of 2007 after an establishment period in late 2006. CISP operates at three Victorian Magistrates Court venues: Melbourne, Sunshine and Morwell and is managed within the Court Support and Diversion Services branch of the Magistrates Court of Victoria. This is the final evaluation report on the Court Integrated Services Program (CISP). The evaluation commenced in late 2006 and the findings presented here cover the implementation and operation of the program up to the middle of 2009. The project specification also included an econometric (cost-effectiveness) component. This was conducted by PricewaterhouseCoopers and is reported separately. Overall, the evaluation found that CISP: had achieved or exceeded its targets for the engagement and retention of clients, was able to match the intensity of intervention to the risk and needs of clients, achieved a high rate of referral of clients to treatment and support services. Other key findings were: A study of CISP clients health and well-being showed they had much lower levels of mental health than comparable community groups and that their mental health improved during their period on the program; Magistrates and other stakeholders showed a high level of support for the program and its outcomes; and compared with offenders at other court venues, offenders who completed CISP showed a significantly lower rate of re-offending in the months after they exited the program. This Executive Summary reports on the evaluation findings against each of the questions or issues specified in the project brief. Findings in relation to program implementation issues are presented first, followed by findings in relation to program outcomes. Each Melbourne Criminological Research and Evaluation 5

finding is accompanied by chapter and section references showing where more detailed analysis is presented. The recommendations from the evaluation are presented at the end of this section. Details of the evaluation methods can be found in Appendix 2. Program implementation findings 1. Is the number of defendants being referred to the CISP reflecting those estimated by the demand modelling? If not, what factors are influencing the referral numbers and the apparent variations? There were 2,004 clients referred to CISP in the 2008 year, which is 86% of the target of 2,316 clients set when CISP was established. Referrals at the Latrobe Valley venue exceeded the target by 50% (276 referrals compared with a target of 184) while Sunshine and Melbourne venues achieved 88% and 78% of their referral targets respectively. There was a gradual increase in referrals throughout 2007 and 2008, although these increases were apparent at Sunshine and Latrobe Valley, but not Melbourne (see Chapter 2.8). Generally, the characteristics of the CISP client population reflect the assumption made in the demand modelling for the program. The most significant variation from the demand estimates is in the high rate of clients with possible Acquired Brain Injury. (Chapter 6.4) 2. From which source/s are referrals most commonly originating? Are engagement rates variable according to referral source? Are outcomes linked to referral sources? The majority of referrals (75%) are made by clients legal representatives, with referrals by Magistrates accounting for a further 15% of referrals. Self-referrals make up around 5% of referrals, although it should be noted that some clients who wish to be referred to CISP may ask their legal representative to do this on their behalf. Clients referred by Magistrates have a higher engagement rate than those referred by their legal representative or self-referrals (see Chapter 2.5). There was no variation in program Melbourne Criminological Research and Evaluation 6

completion outcomes across the different referral sources. The key factors in determining program outcomes were the offender s custody status and accommodation stability at program entry, and whether CISP was made a bail condition. 3. Is the number and proportion of referred defendants engaging in the CISP reflecting those estimated by the demand modelling? If not, what factors are influencing the variations in engagement rates? Overall, 64% of those referred to CISP become engaged clients. This is greater than the estimated demand modelling take-up rate for referrals of around 60%. Referrals were most frequently found to be unsuitable because the person failed to attend the assessment (41% of unsuitable outcomes), was referred to Forensicare (26% of unsuitable outcomes), or because the referral was withdrawn (22%). Overall, higher risk clients are less likely to become engaged clients. (Chapter 2.4) 4. Are the deviations in referrals and engagement rates proportional across regions? What factors are influencing the variations? What has this highlighted about region-specific needs in relation to various aspects of the CISP? There is significant variation in engagement rates between the three program venues. Engagement rates are around 40 % higher at Sunshine than at Melbourne and this variation cannot be accounted for by differences in clients risk profile or referral source. Proportionately more referred clients at Melbourne fail to attend their assessment, are referred to Forensicare or withdraw their referral application. However, it is unclear why these factors are more prevalent at this venue. (Chapter 2.5) The rate of referral to the Latrobe Valley program site is affected by transport access issues for defendants from communities away from the main Latrobe Valley, and to some extent within the valley. This is mainly an issue of equity of access rather than a limiting factor on overall referral rates at Latrobe Valley, which are higher than anticipated (see Chapter 4.1). Melbourne Criminological Research and Evaluation 7

5. Is the CISP receiving priority access to treatment and support services? Is this access adequate and timely? Are there any changes to service-level agreements that would be beneficial? In general, CISP clients receive appropriate and timely referral to treatment and support services. Their engagement with these services is constrained by issues of service availability that are common to all clients of forensic programs. The contracted service agreements have resulted in improved access to accommodation, mental health and ABI assessment services. Access to COATS drug assessment and treatment services is consistent with previous service arrangements under the COATS service protocols. Key problem areas for service access include: Limited availability of places in residential drug and alcohol treatment programs (Chapter 5.5) Very limited availability of long-term housing under the Justice Housing Support Program, and limited availability of emergency and temporary housing in the community (Chapter 7.3) Limited access to mainstream mental health service agencies (Chapter 6.3) 6. Are the existing court services that have integrated into the CISP working effectively? What barriers are evident in achieving a cooperative and co-ordinated response? The CISP teams demonstrate a high degree of integration across the service areas of drugs and alcohol, mental health, disability, indigenous support and accommodation support. The primary barriers to effective team operation is the high level of staff turnover which inhibits the development of stable, productive relationships between team members, and the high level of work demand that means that cases are sometimes assigned to staff members on the basis of availability rather than expertise. (Chapter 4.1) One site-specific gap in the integration of contracted services exists at the Latrobe Valley site where there appears to be limited use of mental health screening services (Chapter 6.3) and limited contact with the accommodation support agency (Chapter 7.4). Melbourne Criminological Research and Evaluation 8

7. What assumptions have been made in the development of the CISP Service Delivery Model? The CISP service delivery model is one of a range of clinical and support program service approaches currently operating in Victorian courts. The other examples include CREDIT/Bail Support, the Neighbourhood Justice Centre client services function, Drug Court, Koori Court, and the applicant and defendant support functions for the Family Violence Courts. Each is based on different assumptions and service models, and each involves some specific skill sets. However, underpinning all these services is a general body of clinical knowledge and technique and common case management, support and other court-based functions. It is proposed that there should be a general review of these approaches with a view to creating a court support services function that would provide the basis for the delivery of a range of clinical, support, referral, supervision and case management services to court clients. 8. Are clients being assigned to their appropriate level of intervention? Is this being reflected in retention rates or other outcomes? Engagement at the three program levels is matched to defendants risks (higher risk defendants are engaged at higher program levels). Based on the risk assessment component of the CISP Screening Assessment, clients assigned to the Community Referral program stream are low risk, while those assigned to the Intermediate and intensive stream are medium and high risk respectively. (Chapter 2.5) There is consistency in engagement patterns across the three program venues (Chapter 2.5). Clients assigned to the Intensive program level have a lower completion rate than those assigned to the Intermediate level, but this is consistent with their generally higher level of risk. Melbourne Criminological Research and Evaluation 9

9. Do Magistrates Court staff, support staff, stakeholders and clients have an awareness and accurate understanding of the CISP? There is a high level of support by Magistrates for the CISP. This is demonstrated by the high rate of agreement between workers recommendations and case outcomes as determined by the court (Chapter 2.8), and an increasing preparedness by Magistrates to refer defendants directly to the program. Overall, external stakeholder knowledge about the CISP is good, with awareness about the program increasing over time. Lawyers continue to be the primary source for referrals to the program, and their engagement is critical to the program s success. There are real differences in approach between CISP and the treatment and support services, especially in relation to accommodation and mental health programs, and this limits the effectiveness of their engagement with CISP. Regular communication with external stakeholders about the program s goals and achievements is required to ensure that stakeholders are able to respond appropriately to CISP s service needs. (refer stakeholder feedback sections in Chapters 5, 6, 7, and 8) There are significant variations in the way that CISP services are incorporated into court operations. This is primarily the product of the extent to which different Magistrates adopt therapeutic jurisprudence practices. It is particularly important for CISP to communicate its goals and achievements to this audience. However, it must be recognized that the adoption of therapeutic jurisprudence approaches is ultimately the responsibility of the Magistracy, and CISP staff and management should be available to support these developments. (Chapter 8.1, 8.4) The falling rate of judicial monitoring is of concern, although this may represent greater confidence in the capacity of CISP case managers. (Chapter 2.6) 10. Are the Courts satisfied with the quality and timeliness of reporting by the CISP team? Does the CISP meet defendants expectations of service? Melbourne Criminological Research and Evaluation 10

Overall, there is a high level of stakeholder satisfaction with the quality and timeliness of reporting by CISP team members. Where there have been problems, these are attributed by stakeholders to the high level of staff turnover in the program and the consequent inexperience of staff responsible for these reports. Case managers should be provided with more explicit criteria to guide decisions about program entry, and more detailed advice about client assessment and reporting procedures. (Chapters 4.3, 8.4) Feedback from clients in the program was generally very positive but the evaluation design did not provide for any systematic assessment of defendants expectations of service. 11. What is the average and range of duration of client engagements by level of intervention? Do these vary by factors such as region or judicial monitoring? The mean period of engagement for clients who completed CISP (from notification of a program place to exit date) was 110 days (3 months and three weeks). For noncompleting clients, the mean number of days from notification to exit was 62. There were no significant differences in time on the program between the three program venues, the Intermediate and intensive program levels, or between clients who were subject to judicial monitoring and those who were not. (Chapter 2.7) 12. Has the integration of various services resulted in a reduction in duplication of assessments for clients, and a reduction in an overlap of referrals? This question is difficult to assess as there are no clear benchmarks for assessment and referral overlaps prior to the program. There is no evidence of duplication in assessments or referrals in the current program. However, it should be noted that some service referrals (accommodation, mental health) need to be supported by continuing case management by CISP. (see generally Chapters 5, 6, and 7) 13. Does the assessment tool utilised by the CISP team capture adequate information? Is the assessment being administered Melbourne Criminological Research and Evaluation 11

appropriately? What improvements could be made to the assessment tool? A review of the Screening Assessment in 2008 found that CISP staff members were administering it appropriately. The main weakness was in the collection of risk information, but this appears to have been rectified. A number of refinements were made to the instrument following the 2008 review. Staff members give generally positive reports of the screening instrument, but also report that they find assessment one of the most challenging aspects of their work. It is recommended that the CISP Policy and Procedures Manual should be supplemented with training and procedural advice on assessment procedures and working in a court environment. (Chapter 4.3) Program outcome findings The key outcomes for CISP include individual client outcomes in relation to the needs addressed by intervention and support services, program completions, impact on court workload and other elements of the justice system, bail and order compliance, and reoffending. 14. Do clients show an improvement in social and physical health and well-being, including a reduction in drug use and drug-related harms? There are indications across a number of measures that clients showed improved health and social functioning as a result of involvement in CISP. The SF-12 health and wellbeing survey showed significant improvement in clients reported physical and mental well-being. Case manager assessments show reduced levels of problematic alcohol use. However, there are a number of factors that make assessing the impact of CISP on social, health and substance abuse factors problematic. CISP is a gateway to services, but for many clients the main outcome of their engagement with CISP is referral to a service agency, and looking for tangible outcomes is premature. This is most evident in relation to housing issues. While CISP was successful in referring clients to housing services, the waiting times associated with providing more stable accommodation are such that there Melbourne Criminological Research and Evaluation 12

was little change in actual housing status across the period of their engagement. A second problem is that outcomes are only known for those clients who remain engaged with CISP. (Chapters 3.6, 5.4 and 7.5) 15. What factors are important in determining program completion? Around six in ten engaged clients complete CISP successfully. Statistical models to predict completion outcomes were developed. The most important factors predicting non-completion were whether the offender was in custody at the time she/he was assessed for CISP, whether CISP was made a condition of bail, and the offenders level of accommodation stability at the time of CISP entry. Court location was also an important factor in predicting non-completion outcomes but it seems likely that this is an artefact of the generally higher completion rates at Latrobe Valley. 16. What is the impact of CISP on court workloads? Engagement in CISP affects court workloads in a number of ways. In order to be considered for CISP, offenders must go through a screening assessment, and this may involve some delay in hearing their bail application. Judicial supervision involving a return to court may involve additional appearances and may require adjustment to Magistrates work schedules. Positive impacts on court workloads include avoiding adjournments required to arrange services for offenders, and avoiding new court appearances associated with recidivist offending. Overall, Magistrates report that CISP does require additional work on their part, but that this is justified by the better results they achieve. 17. Does involvement in CISP lead to lower rates of non-compliance on bail or Community Corrections orders? For any pre-trial program, bail completion rates are obviously a key outcome. However, bail data is not currently able to be extracted in a manner that allows reliable and consistent analysis of bail outcomes for CISP clients and Victorian offenders generally. Melbourne Criminological Research and Evaluation 13

The absence of bail data constitutes a significant barrier to the evaluation of court programs in Victoria. Community corrections order compliance rates were not available for CISP clients as a group. Again, this represents a significant information gap in evaluating programs of this kind. Order compliance was examined for sample of 200 CISP participants, and 200 comparison offenders matched on gender, age, offence type and offence history. The overall successful completion rates for orders up to June 2009 were 49% for the CISP sample, and 45% for the control sample. While positive, this difference was not statistically significant. The successful completion rate for CCS orders state-wide in 2007/08 was 58%, however the sampling process means that these rates cannot be directly compared with one another. 18. Does involvement in CISP lead to lower rates of re-offending? Re-offending rates were examined for the same 200 CISP and comparison offenders as in the order compliance analysis. They were followed up for between 400 and 900 days and any further offences or new charges were recorded. For the CISP group, around 50% were classed as recidivists, of whom 40% had proven charges against them, and a further 10% had charges that had not been finalized. In the comparison group, 64% were classed as recidivists, with 50% having proven charges recorded and a further 13% having unfinalized charges. After 200 days around 30% of the CISP group and 32.5% of the control group had recidivated. By 400 days the degree of divergence was six percent (37% of the CISP group and 43% of the control group) and by 600 days it was eight percent (40% of the CISP group and 48% of the control group. The extent of divergence in re-offending rates reaches its maximum of ten percent by around 700 days. Melbourne Criminological Research and Evaluation 14

Recommendations Recommendation 1: Establish a review of court support programs with the aim of developing a general court support service model that provides statewide services to all the Victorian Magistrates Court at all its venues and across all specialist lists and divisions. The CISP program has achieved its primary output and outcome goals, including: achievement of referral targets; exceeding the target rate for engagements; satisfying the requirement for level of engagement to be matched to need; achieving a high rate of program completions; achieving a high rate of treatment and support program referrals, and; demonstrating a high level of support from Magistrates and other stakeholders. The program has received continuing funding for two years. The primary issue is therefore what should be the future of this program in the period after 2011? The CISP service delivery model is one of a range of clinical and support program service approaches currently operating in Victorian courts. The other programs include CREDIT/Bail Support, the Neighbourhood Justice Centre client services function, Drug Court, Koori Court, and the applicant and defendant support functions for the Family Violence Courts. Each is based on different assumptions and service models, and each involves some specific skill sets. However, underpinning all these services is a general body of clinical knowledge and technique and common case management, support and other court-based functions. It is proposed that there should be a general review of these approaches with a view to creating a court support services function that would provide the basis for the delivery of a range of clinical, support, referral, supervision and case management services to clients of the Magistrates Court of Victoria. This review should consider the following issues: The contribution of court services to the continuing development of therapeutic jurisprudence practices Melbourne Criminological Research and Evaluation 15

The creation of functional specializations that can be applied across a range of court business streams and specialist lists and divisions. These might include: Client assessment Case management Victim counselling The development of client services infrastructure for court services comprising: Information management tools Co-ordinated training and professional development Clinical and case management tools and systems (for example, common assessment tools and procedures), and Professional resource kits (for example, centrally maintained databases of service provider agencies). The development of a mobile service capacity appropriate for regional courts Recommendation 2: Continue to develop and support the connections between CISP and therapeutic jurisprudence practices CISP is an integral element in therapeutic jurisprudence processes that commence at the bail stage and continue throughout the offender s involvement in the court process and through into supervision under court order. For the full value of the interventions and support delivered through CISP to be realized there needs to be continued development of the program s connections with post-sentence support and interventions. This may also include provision for ongoing judicial review in selected cases. Recommendation 3: Legislative or procedural amendments are required to allow defendants charged with serious indictable offences to be subject to judicial monitoring through CISP. Where a defendant is charged with a serious indictable offence and then bailed, there is limited capacity for a Magistrate to exercise judicial supervision with regular reviews of the defendant. Legislative or procedural amendments are required to allow defendants charged with serious indictable offences to be subject to judicial monitoring. This might require establishing reporting dates intermediate between the Filing Hearing and Melbourne Criminological Research and Evaluation 16

Committal Mention appearance that would apply to defendants placed on CISP. These legislative changes might also support the extension of the program to the County Court. Recommendation 4: The CISP Policy and Procedures Manual should be supplemented with training and procedural advice on assessment procedures and working in a court environment. The material on assessment procedures should include more detailed information about the clinical aspects of forensic practice, assessment criteria for the elements of the Screening Assessment, and the interpretation and integration of assessment information across multiple problem domains (especially substance abuse and mental health). The development of more explicit eligibility criteria may also assist staff by making such decisions more externally accountable. The material on working in a court environment should include advice on court processes, reporting in court, and the roles of other professional groups in the court (especially lawyers and police) and the development and management of effective relationships with them. Recommendation 5: The CISP Screening Assessment should be subject to regular review and staff should receive feedback on the aggregated results of assessments. It is important to maintain staff commitment to the structured assessment model for CISP. Two strategies for this are regular reviews of the assessment process (every second year) to ensure that the assessment content and procedures remain up to date and consistent with work practices, and demonstrating the value of the collection of assessment data through communication with staff about the results of assessments. Recommendation 6: Provide CISP staff with more explicit program eligibility and suitability criteria. Both Magistrates and case managers are aware that some referrals are made that are not appropriate for the program. This can be addressed by providing more explicit referral Melbourne Criminological Research and Evaluation 17

criteria that take into account factors such as the defendants previous bail history, any breaches of suspended sentences or Community Corrections orders, and other factors relevant to the court s bail decision. One possibility would be to establish three referral outcomes: unsuitable and rejected referrals (as at present), referrals accepted for assessment, and referrals where a Magistrate s approval should be sought before accepting the defendant for assessment Recommendation 7: CISP program goals for drugs and alcohol should be concerned with effectiveness of the referral process and maintaining clients engagement with treatment programs. Improved drug and alcohol outcomes are part of the CISP goal set, however it seems inappropriate to hold the program responsible for treatment goals that are beyond its direct control. The key service delivery issues for CISP are how effectively it operates as a referral pathway and case management service. The indicators of success in this area should be whether drug and alcohol program referrals are based on a comprehensive assessment of clients risks and needs, and whether clients are provided with the support and supervision that ensures they satisfy the attendance requirements for drug and alcohol services. Recommendation 8: Review the Justice Housing Support program protocol to give greater weight to the needs of pre-trial defendants. It is recommended that the service level agreement for the JHSP should be reviewed to determine whether justice outcomes can be better achieved through greater focus on short-term accommodation. This review process should consider: The likely future availability of long-term housing places for justice clients taking into account patterns of intake and transition to permanent housing for justice clients; and The need for short-term housing places for justice clients and the impact of availability of this form of housing on justice system outcomes, in particular bail and remand decisions. Melbourne Criminological Research and Evaluation 18

Recommendation 9: Review the provision of services to justice system clients with suspected Acquired Brain Injury The rate of suspected Acquired Brain Injury in program clients is much higher than allowed for in the demand modelling for CISP. This points to a high rate of ABI in justice client populations generally, and indicates that a comprehensive strategy to address this issue is required. While arbias 1 provides a high-quality assessment and service response for these defendants, the timescales involved are longer than can be accommodated in a pre-trial program. CISP case managers should receive additional advice on the management of such cases while awaiting full neuropsychological assessments. There should also be consideration of the continuing management of these clients as they progress through other stages of the justice system. 1 Alcohol Related Brain Injury Australian Service Melbourne Criminological Research and Evaluation 19

Part 1: Program description Chapter 1 Introduction This is the final evaluation report on the Court Integrated Services Program (CISP). The evaluation commenced in late 2006 and the findings presented here cover the implementation and operation of the program up to the middle of 2009. This report is the fourth report on the CISP program arising out of the evaluation, and it incorporates some material presented in interim evaluation reports prepared in 2007 and 2008. A special report in 2008 examined the assessment of clients risks and needs. The evaluation approach used for CISP focused initially on the conceptualization of the program (that is, its theoretical and policy basis), followed by examination of program implementation issues. This final report is mainly concerned with the outcomes of the program, but where appropriate it makes reference to process and implementation issues. The evaluation had two primary objectives: To determine the effectiveness of the CISP in meeting its overarching objective to reduce the re-offending rate of defendants To gather objective evidence to support future decision making by the Victorian Government in relation to the cost effectiveness of this initiative, and its expansion state-wide. The project specification also identified six areas of program effectiveness and efficiency to be examined by the evaluation. These included: Stakeholder satisfaction with services Program process efficiencies Reduction in re-offending rates Enhanced order compliance Reductions in drug use and drug-related harms, and improvements in clients health and well-being The impact of the program on justice system load Melbourne Criminological Research and Evaluation 20

The project specification also included an econometric (cost-effectiveness) component. This was conducted by PricewaterhouseCoopers and is reported separately. 1.1 Description of the Court Integrated Services Program The Court Integrated Services Program (CISP) commenced at the beginning of 2007 after an establishment period in late 2006. The program represents a development of existing pre-trial and bail support program models, and in particular the CREDIT (Court Referral and Evaluation for Drug intervention and Treatment) and Bail Support programs, established in Victoria in 1998 and 2001 respectively. The CISP model diverges from traditional pre-trial programs in placing more emphasis on addressing the underlying causes of offending through: - Greater emphasis on individualized case management. Case management in CISP involves more therapeutic interactions between clients and workers, in comparison with the predominantly referral and advocacy approach of many pre-trial programs; - The CISP program teams are multi-disciplinary, and the referral and assessment process is intended to match clients needs with workers skills and expertise. CISP brings together a range of services that were previously available to defendants as separate services; - The case management model provides for three levels of service response (Intensive, Intermediate and Community Referral) and clients are intended to be directed to these levels of service response in a manner consistent with their assessed level of risk and need. - Establishing service agreements with housing agencies as part of the Justice Housing Support Program, with arbias for the provision of Acquired Brain Injury assessment and support services, and with the Community Offenders Advice and Treatment Service (COATS) for drug and alcohol assessment and referral services. CISP operates at three Victorian Magistrates Court venues: Melbourne, Sunshine and Morwell. The first two venues are metropolitan courts servicing inner Melbourne and the western suburbs respectively, while the Morwell venue services the Latrobe Valley, to the east of Melbourne. The program is managed within the Court Support and Melbourne Criminological Research and Evaluation 21

Diversion Services branch of the Magistrates Court of Victoria, and has a staff complement of 26, including four team leader positions, 18 case manager positions and four administrative positions. Chapter 2 of this report gives more details of the volume and nature of the program s case-flows, and Chapter 4 provides a more detailed description of the clinical and case management processes in the program. Readers of this evaluation report may find it useful to understand how CISP operates in a day-to-day sense. The following case study describes a typical case and illustrates the way the program interacts with court bail and sentencing processes. A CISP case history A 32-year old male on remand was assessed, at his request, for CISP. The person had a very long history of offending, had served a number of gaol and Community Corrections sentences. At the time of his assessment he was facing two separate sets of charges involving burglary and theft, and was in breach of a suspended sentence. The most likely outcome for a defendant with this history would be a further gaol sentence. During the assessment, the CISP case manager noted that the person showed indications of an acquired brain injury (ABI) as well as other psychological issues. As part of his parole conditions, the offender had been attending Turning Point for counselling where he had established a beneficial therapeutic relationship with his counsellor which he wished to continue. However, the case manager ascertained that the offender was ineligible for CISP because he was currently serving a period of parole. Bail was granted with various conditions. The case was adjourned to a date after his parole expired. The CISP case manager liaised with the supervising Community Corrections Officer in relation to the issues identified for the offender, and Corrections worked with him on those issues for the rest of the parole period. On the return date, the offender reiterated his commitment to CISP and his bail was varied to include participation on CISP as a bail condition. While on CISP he continued his treatment at Turning Point, attended all appointments with CISP, found housing, maintained pharmacotherapy, remained abstinent from drugs, remained in a stable relationship, participated on a Personal Support Program through Centrelink and maintained psychological counselling as arranged through CISP. Melbourne Criminological Research and Evaluation 22

At the sentencing hearing a number of reports were tendered to the court from CISP, Turning Point, and the Centrelink PSP case manager, as well as a report on his parole from the Community Corrections Officer. The reports all attested to the effort that the offender had put in whilst on the program and the progress he had made. In all, the offender had been under court supervision for over eight months, including three months on bail while completing parole and five months on CISP. The magistrate imposed a further suspended sentence. It was deemed by the magistrate that it would be counter-productive to impose an immediate custodial sentence given the progress made by the offender and his prospects for long-term rehabilitation. The prosecutor was invited to make a submission in relation to the restoration of the suspended sentence. He indicated that in view of the offender s excellent progress, he declined to make any further submissions. The magistrate made no further order on the breach of the suspended sentence. At the conclusion of the proceedings, the offender asked to address the Court. He thanked the magistrate and his CISP case-manager for the opportunity and indicated that he was planning to attend TAFE to study social work. Melbourne Criminological Research and Evaluation 23

Chapter 2 Case flows & system load 2.1 Case flow and system load goals A number of the goals of the CISP are concerned with the way that clients are recruited for the program, enter and progress through the program, and finally exit. These processes are referred to as case flow, and the case flow goals set for CISP included: The number of clients engaged at the three program venues would meet specified targets. These targets were set on the basis of case flows for the existing CREDIT/Bail Support, Disability Services and Aboriginal Liaison programs, plus an estimated number of new referrals. These case flow targets are shown in Table 2.1 below. Clients would be allocated to the appropriate program level based on their assessed risk and needs. Clients would spend up to four months on the program, with case management to monitor the defendant s progress, address identified needs and provide access to support services, and plan for the client s exit. Judicial monitoring will be used where deemed necessary and appropriate by the judicial officer. Case location Referral targets Latrobe Melbourne Sunshine Total Valley 41 895 330 1266 Existing ALO, CREDIT/BSP and DS referrals 143 581 326 1050 New CISP referrals 184 1476 656 2316 Targets for CISP referrals Table 2.1 CISP case flow targets (annual) This chapter examines the case flow outcomes of CISP, including: Numbers and sources of referrals to the program, Melbourne Criminological Research and Evaluation 24

Engagement rates by program level Judicial monitoring rates Completion rates and non-completion factors Duration of engagement Except where stated, all information presented in this chapter is derived from the CISP Platypus Systems Case Management System (PSCMS). 2.2 CISP program flow processes CISP begins when a person is charged with criminal offences. At this stage, the person may be in custody awaiting a bail hearing, already on bail, or summonsed to appear. The CISP case flow process includes the following stages: Prospective clients may be referred to CISP from a variety of sources, including their legal representative, police, another treatment or support program, court staff including the judiciary (via the court hearing as part of the bail application), family and friends or self-referred. A referral application is completed to determine whether the person should be further assessed for the CISP via the screening assessment process. This referral application collects identifying information on the person and his/her legal representative, offence and legal status, indigenous status, and any issues or problems that might warrant assessment by CISP staff. The information on the referral form is used to determine allocation to a case worker who completes a more detailed assessment. This Screening Assessment includes a more detailed criminal and legal history, an assessment of the person s social and economic support needs, drug and alcohol use, and physical and mental health. At the end of the Screening Assessment a brief assessment of risks is also completed. Where a client is assessed as appropriate for a Level 1 (Community Referral) response, only basic information is collected at this assessment. Melbourne Criminological Research and Evaluation 25

Following the screening assessment, the client may return to court where conditions associated with his or her involvement in CISP may be attached to the Bail Order. There is considerable variability in practice at this stage, with some Magistrates making a Bail Order with conditions relating to CISP engagement, while others recommend CISP engagement but without any court direction. Magistrates may also order the person to re-appear at a later date for a progress review. When a client is engaged with CISP, the case worker prepares a case management plan that may involve referral to a range of treatment and support agencies, as well as continued supervision through CISP. Program exit occurs when the case management plan is completed, the client fails to satisfy the plan s requirements, or other legal matters arise. The CISP pathway is shown in Figure 2.1 Figure 2.1 CISP Process Map (March 2005) 2 2.3 Referrals to CISP 2 From CISP Service Delivery Model (Department of Justice, 16 March 2006) Melbourne Criminological Research and Evaluation 26

The CISP was launched in November 2006 at the Melbourne and Sunshine Magistrates Courts and the program took a small number of clients during the final months of 2006. In the first year of the program (2007), a total of 1,752 clients were referred to the program, and in 2008 this had increased to 2,004 referrals. Approximately 60% of referrals were to the Melbourne venue of the program, 27% were to the Sunshine venue, and 11% were to the Latrobe Valley venue. There was little evidence of a ramp-up phase for the program (a period when referrals increased rapidly over a period of months from an initial low level to a higher, more stable level), probably because the program was seen as being a continuation of the earlier CREDIT/Bail, Disability and ALO programs that were present at these venues. However, there is evidence of a gradual increase in the referral rate over the two years for the program as a whole, and for the Sunshine and Latrobe Valley venues. Figure 2.2 shows the monthly number of clients referred to CISP at each venue, and the linear trend lines for each venue. Monthly referrals to CISP 250 200 150 100 50 0 January 2007 March 2007 May 2007 July 2007 September 2007 November 2007 January 2008 March 2008 May 2008 July 2008 September 2008 November 2008 Latrobe Valley Melbourne Sunshine CISP total Linear (Latrobe Valley) Linear (Sunshine) Linear (Melbourne) Linear (CISP total) Figure 2.2 Monthly referrals at each venue The trend line for Melbourne is stable (that is, the slope is not significantly different from zero), but the trends for Sunshine, Latrobe Valley and the program as a whole all show significant positive slopes. There is some evidence of a levelling off in referrals in the last Melbourne Criminological Research and Evaluation 27

six months of 2008 at each of the venues but it is too early to say whether this indicates stabilization in the rate of program referrals. It should be noted that the number of referrals to CISP reported in internal monthly reports cannot be validated from PSCMS case management system records. Over the 2007 and 2008 years an additional 499 additional referrals were reported in monthly reports, but these cannot be reliably identified from PSCMS referral records. Around half of these additional referrals were not assessed for the program, and none became program clients at program levels 1, 2 or 3. Repeat clients A substantial number of CISP referrals involve clients who have been previously referred to this program or its predecessor (CREDIT/Bail Support). Of the 3,756 CISP referrals in 2007 and 2008, there were 2,710 defendants who were only referred once over the period, 384 who were referred twice, 73 who were referred three times, and 14 who were referred four or more times. Overall, 28% of CISP referrals involve defendants who have had previous contact with the program. Referrals are also cross-matched with client records from the CREDIT /Bail Support (CBS) program. In the first six months of the program, 19% of referrals were defendants who had previously been engaged with the CBS program, but by the second half of 2007 this had fallen to 12% and by 2008 less than 10% of referrals had prior involvement with CBS. Persons with significant criminal behaviour frequently go through repeated court episodes in a given period, and hence some level of multiple engagement in a program like CISP is to be expected. Since each referral constitutes a separate assessment and case management process, the remainder of this analysis will use referral episodes ( cases ) as the primary unit of analysis. Source of referrals The most common source of referrals to CISP was the defendant s legal representative. Eight in every ten referrals were from this source. Magistrates accounted for the next largest group (12.5%), and self-referrals accounted for a further 5%. Referrals from Magistrates were less frequently made at Latrobe Valley than at the other two sites, and while referrals by police were uncommon at all three sites, they were least likely at Melbourne. Within the Other category are nine referrals that were Crimes Family Melbourne Criminological Research and Evaluation 28

Violence cases: eight of these were referrals from the Sunshine venue. Note that provision to record these CFV referrals on the CMS was only made in 2008. Prior to this these were recorded as Magistrate referrals, and the real number in this group is somewhat higher. Referral source Legal representative Case location Latrobe Valley Melbourne Sunshine Total N 368 1831 782 2981 % 86.4% 79.5% 76.1% 79.4% Judiciary N 19 291 163 473 % 4.5% 12.6% 15.9% 12.6% Self referral/ N 17 122 42 181 family or friend % 4.0% 5.3% 4.1% 4.8% Treatment or support N 11 34 9 54 agency/service % 2.6% 1.5%.9% 1.4% Police member N 8 4 13 25 % 1.9%.2% 1.3%.7% Other N 3 21 18 42 %.7%.9 1.8% 1.1% Total N 426 2303 1027 3756 % 100.0% 100.0% 100.0% 100.0% Table 2.2 Source of Referrals by Venue The source of referrals altered somewhat over the two years, with referrals by legal representatives falling from 84% of all referrals in the first six months of the program, to 75% in the second half of 2008. Referrals by Magistrates increased from 10% to 14%, and self-referrals increased from 3% to 6% over the same period. Magistrate referrals were more likely where the defendant was female, and female defendants were also more likely to self-refer. Melbourne Criminological Research and Evaluation 29

2.4 Assessments and assessment outcomes There were 3,736 assessments of referred defendants made in 2007 and 2008, of which around 60% involved defendants who were in custody at the time. There were substantial differences between the three venues in the likelihood that a referred defendant was in custody. Two-thirds of defendants at Melbourne were in custody when they were assessed, but only half of those at Sunshine and a third of those at Latrobe Valley. This difference reflects the role of Melbourne as the central Magistrates Court, with custody facilities for holding defendants who have been refused bail. Assessed in custody? Yes Case location Latrobe Valley Melbourne Sunshine Total N 141 1523 511 2175 % 33.1% 66.6% 50.0% 58.2% No N 285 765 511 1561 % 66.9% 33.4% 50.0% 41.8% Total N 426 2288 1022 3736 % 100.0% 100.0% 100.0% 100.0% Table 2.3 Assessments in Custody by Venue Assessment outcomes Approximately 70% of those referred to CISP were assessed as suitable for the program and recommended for one of the three intervention levels, and one-quarter were assessed as not suitable for the program, with a further 3% found not suitable but offered consultancy services (provided with advice and referral to support or treatment services). Referrals were most frequently found to be unsuitable because the person failed to attend the assessment (417 cases, or 41% of unsuitable outcomes), was referred to Forensicare 3 (262 cases, or 26% of unsuitable outcomes), or because the referral was withdrawn (220 3 Offenders with a serious mental illness or who have other significant forensic issues are referred to Forensicare for assessment and specialist case management. Melbourne Criminological Research and Evaluation 30

cases or 22%). A further 40 cases were not accepted onto CISP because they were adjourned to another jurisdiction or to the venue where CISP was not available. Case recommendation outcome Case location Suitable, recommended for Latrobe Valley Melbourne Sunshine Total N 334 1489 874 2697 CISP engagement % 78.4% 65.2% 85.8% 72.4% Not suitable - Consultancy Not suitable Total N 22 69 24 115 % 5.2% 3.0% 2.4% 3.1% N 70 725 120 915 % 16.4% 31.8% 11.8% 24.6% N 426 2283 1018 3727 % 100.0% 100.0% 100.0% 100.0% 2.4 Case Recommendations by Venue The proportion of referrals found unsuitable was relatively stable over the two years, and the reasons for unsuitability were similarly stable over time. However, the proportion of those assessed as unsuitable was significantly higher at Melbourne (32%) than the other two venues (16% and 12% respectively). It is unclear why a greater proportion of referrals were found unsuitable at Melbourne, as there was little site-to-site variability in the reasons for assessments of unsuitability. 2.5 Case recommendations and outcomes Following assessment, a report is made to the court on the defendant s suitability for CISP, and if assessed as suitable, recommending engagement in one of the three program levels (Community Referral, Intermediate, Intensive). The Magistrate may accept or vary this recommendation. Overall, nearly 80% of workers recommendations were accepted by Magistrates. Over 95% of recommendations that a defendant was unsuitable resulted in a case outcome of unsuitability and over 85% of recommendations for the Community Referral level of CISP were accepted. The likelihood that a recommendation for engagement in CISP would be rejected by a Magistrate was higher for more intensive levels of the program. Twelve percent of recommendations for level 1 were ultimately found to be unsuitable, compared with 19% of recommendations for program level 2 and 26% of recommendations for program level 3. The level of agreement between Melbourne Criminological Research and Evaluation 31

workers recommendations and case outcomes has remained stable over the two years 2007 and 2008. Case outcome Level 1 Level 2 Level 3 Not suitable Not Total CISP worker Community Intermediate Intensive Consultancy suitable recommendation Referral Level 1 N 289 3 3 14 28 337 Community referral % 85.8%.9%.9% 4.2% 8.3% 100.0% Level 2 N 13 1700 47 28 372 2160 Intermediate %.6% 78.7% 2.2% 1.3% 17.2% 100.0% Level 3 N 5 29 602 11 209 856 Intensive %.6% 3.4% 70.3% 1.3% 24.4% 100.0% Not suitable N 2 1 1 52 14 70 Consultancy % 2.9% 1.4% 1.4% 74.3% 20.0% 100.0% Not suitable N 1 0 1 10 292 304 %.3%.0%.3% 3.3% 96.1% 100.0% Total N 310 1733 654 115 915 3727 % 8.3% 46.5% 17.5% 3.1% 24.6% 100.0% Table 2.5 CISP Recommendations and outcomes Engagement rates In the remainder of this chapter, clients who had a case outcome of Level 2 or Level 3 and who received case management through CISP are referred to as engaged clients. In the period July 2007 to June 2008, there were 1,833 clients who were assessed for CISP, of whom 1,140 were accepted onto Level 2 or 3 of the program, 175 went to Level 1, and 518 were found unsuitable, giving an engagement rate of 62.2%. In the six months to December 2008 there were 651 clients engaged at Level 2 or 3 out of a total of 1000 assessed, giving an engagement rate of 65.1%. The proportion of referred clients who became engaged clients was highest at Sunshine (77.3%) and lowest at Melbourne (56.5%). Clients referred by Magistrates were more likely to become engaged clients than those referred by legal representatives or self-referrals. Melbourne Criminological Research and Evaluation 32

Engagement rates by referral source 80.0% Proportion of referrred clients who engage 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0%.0% Self referral Legal representative Judiciary Figure 2.2 Engagement rates by source of referral Consistency in allocation to program level In the early stages of the evaluation substantial variations between the three venues were identified in the allocation of clients to program levels. Clients at the Latrobe Valley venue were much more likely than clients at the other two venues to be engaged at the Intensive program level, while clients at Sunshine were more likely to be engaged at the Intermediate program level. These variations could not be attributed to differences in the characteristics of clients. Program guidelines about allocation were reviewed in consultation with team leaders and by 2008 the level of site-to-site variation had greatly diminished. Across the two years the proportion of clients engaged at each level showed a high level of consistency between the three venues. Melbourne Criminological Research and Evaluation 33

Proportion of CISP clients at each program level: 2007 and 2008 70.00% 60.00% 50.00% 40.00% 30.00% Latrobe Valley Melbourne Sunshine 20.00% 10.00% 0.00% Community Referral (Level 1) Intermediate (Level 2) Intensive (Level 3) Figure 2.3 CISP clients by program level Allocation to program level in relation to clients assessed risk The CISP assessment requires case mangers to assess clients risk in relation to a series of risk criteria (see Screening Assessment at Appendix 3). These risk criteria were modified slightly following the assessment review in 2007. The total number of items identified provides an index of the client s risk, and this index can be used to examine whether clients are allocated to program levels in relation to their assessed level of risk. Table 2.6 shows the mean risk index scores for 2007 and 2008 clients at each of the three program levels 4. It can be seen that clients allocated to the Community Referral level program were on average low risk in that less than one risk item was identified for each client. Clients allocated to the Intermediate and Intensive program levels were of relatively higher risk, with an average of three and over four risk items identified respectively. 4 There were 277 clients who had no risk items identified. Without examination of the original assessment forms it cannot be determined whether these were blank risk assessments or clients who demonstrated no risks. Melbourne Criminological Research and Evaluation 34

Case recommendation outcome Mean N Std. Deviation Level 1 Community Referral.4871 310 1.59027 Level 2 Intermediate 3.0537 1732 2.06145 Level 3 Intensive 4.3645 653 2.61697 Total 3.0761 2695 2.41910 Table 2.6 Risk scores by program levels 2.6 Use of judicial monitoring Where a magistrate directed that a defendant should take part in either Level 2 or Level 3 of CISP, he or she could also direct that this should involve monitoring of progress by the court (judicial monitoring). Information on judicial monitoring was recorded for approximately three-quarters of all engaged CISP clients. Judicial monitoring was most often used at the Sunshine venue, where 80% of cases had this recorded, and least used at Latrobe Valley where 26% of cases had judicial monitoring recorded 5. Judicial Monitoring Case location Latrobe Valley Melbourne Sunshine Total Yes N 51 537 488 1076 % 25.8% 53.3% 80.1% 59.3% No N 147 470 121 738 % 74.2% 46.7% 19.9% 40.7% Total N 198 1007 609 1814 % 100.0% 100.0% 100.0% 100.0% Table 2.7 Use of judicial monitoring by CISP venue Judicial monitoring was more likely to be required where the defendant had been referred to CISP by a Magistrate (77% of magistrate-referred cases received judicial monitoring) and least likely where the defendant was self-referred (30% of cases). There were no evident differences in the likelihood that judicial monitoring would be imposed on male versus female defendants, or Indigenous versus non-indigenous defendants. However, there was a strong downward trend in the use of judicial monitoring throughout the life 5 The level of missing information about judicial monitoring was also highest a Latrobe Valley. Melbourne Criminological Research and Evaluation 35

of the program (Figure 2.4). In the first six months after commencement, over threequarters (77%) of all engaged clients were subject to judicial monitoring but by the second half of 2008 this had declined to less than half (43%). Magistrates were asked about this trend and their comments are reported in Chapter 8. Proportion of engaged CISP clients receiving judicial monitoring 90% 80% 70% 60% 50% % 40% 30% 20% 10% 0% First half 2007 Second half 2007 First half 2008 Second half 2008 Figure 2.4 Proportion of engaged clients receiving judicial monitoring 2.7 Program completions Of the 2,387 defendants who became engaged CISP clients in 2007 and 2008, over 85% had a program exit record 6. Overall, around six in ten engaged clients completed CISP satisfactorily, while 17% were unable to complete because they were remanded in custody (either as a result of breaches of bail conditions or further offending), and 18% were terminated for non-attendance. A small number terminated voluntarily. Included in the Other category were ten defendants who died while on the program. The 6 In theory, almost all clients who were engaged in 2007 and 2008 should have exited by the date of the CMS extract on which these results are based (1 April 2009). The proportion of engaged clients with exit records did not vary greatly between venues. Melbourne Criminological Research and Evaluation 36

completion rate was highest at Latrobe Valley, where three-quarters of all engaged clients completed. Completion rates were higher for clients on the Intermediate program level than for those on the intensive level (61% versus 52%) and Indigenous clients were less likely to complete than non-indigenous (46% versus 60%). However, there were no differences in completion rates for men versus women, clients with judicial monitoring versus those without, or across the main referral sources. Reason for exit Case location Latrobe Valley Melbourne Sunshine Total Completed program N 197 618 386 1201 % 74.6% 56.1% 56.5% 58.6% Did not complete program - N 19 157 166 342 remanded % 7.2% 14.3% 24.3% 16.7% Did not complete program- N 35 234 106 375 non attendance % 13.3% 21.3% 15.5% 18.3% No longer wanted to N 9 38 9 56 participate % 3.4% 3.5% 1.3% 2.7% Other N 4 54 16 74 % 1.5% 4.9% 2.3% 3.6% Total N 264 1101 683 2048 % 100.0% 100.0% 100.0% 100.0% Table 2.7 Reasons for exit by CISP venue Statistical models to predict completion outcomes were developed. These models and the development process are described in detail in Appendix 1. The key findings of this modelling process were that: Prediction of completion / non-completion outcomes yielded a more balanced model than one designed to predict non-completion as a result of nonattendance; The most important factors predicting non-completion were whether the offender was in custody at the time she/he was assessed for CISP, whether CISP was made a condition of bail, and the offenders level of accommodation stability at the time of CISP entry; Court location was also an important factor in predicting non-completion outcomes but this is mainly related to the generally higher completion rates at Latrobe Valley. Melbourne Criminological Research and Evaluation 37

Court outcome at CISP exit Over 60% of CISP clients had not had their court matters resolved at the time they exited the program. The majority of these were cases that had not yet been determined (980 cases or 48% of exiting clients), cases where a warrant had been issued (182 cases or 9% of exiting clients) and cases where the defendant had been committed for trial to a higher court (87 cases or 4% of exiting clients). There were significant differences in exit outcomes depending on whether clients exited after completing CISP. Clients who completed CISP were more likely to have their matters resolved at exit, less likely to be given a custodial sentence, and more likely to receive a community order, suspended sentence, fine or bond. Court outcome at CISP exit Completion Completed Not completed Total Custody N 17 79 96 % 1.4% 9.3% 4.7% Community order N 257 24 281 % 21.4% 2.8% 13.7% Suspended sentence N 139 9 148 % 11.6% 1.1% 7.2% Fine, bond N 136 11 147 % 11.3% 1.3% 7.2% Other penalty N 69 55 124 % 5.8% 6.5% 6.1% Matters not yet determined, warrant issued or N 581 668 1249 committed for trial % 48.5% 79.0% 61.1% Total N 1199 846 2045 % 100.0% 100.0% 100.0% Table 2.8 Court outcome at CISP exit by completion status Melbourne Criminological Research and Evaluation 38

Time on CISP The mean period of engagement for clients who completed CISP (from notification of a program place to exit date) was 110 days (3 months and three weeks) 7. For noncompleting clients, the mean number of days from notification to exit was 62. Clients who were exited as a result of being remanded in custody tended to exit faster than those who were exited for non-attendance or other reasons (50 days versus 64 days). There were no significant differences in time on the program between the three program venues, the Intermediate and intensive program levels, or between clients who were subject to judicial monitoring and those who were not. Mean number of days on CISP 120 Days from notification to exit 100 80 60 40 20 0 Completed program Did not complete program - remanded Did not complete programnon attendance No longer wanted to participate Client death Adjourned to non CISP location Matters adjourned to higher court Figure 2.5 Mean number of days to completion by completion status 7 There are problems with accurate calculation of time on the program as some clients are placed on the program but do not take up this place immediately. The 2% of cases with the longest durations (over 200 days) were deleted from the calculations of elapsed times. Melbourne Criminological Research and Evaluation 39

2.8 Commentary on CISP case flow and system load issues Meeting program case flow targets CISP was established on the basis that there was a very large unmet need for specialized treatment and support services for defendants at court. The case flow targets for the program were set by taking the levels of activity across the existing CREDIT/Bail Support, Disability Services and ALO programs and adding an estimated flow of new business arising from the expanded scope of the program. In all of the three venues, this new business target was substantial when compared with the cases dealt with by the existing programs. Prior to CISP, the Latrobe Valley venue had been dealing with around 40 CREDIT/Bail Support, disability support and ALO cases a year, and it was estimated that the establishment of CISP would increase this to 180 cases per year (that is, a 450% increase). The estimated increase in case flow at Sunshine was around 100% (from 330 per year before CISP to 656), and at Melbourne 65% (895 pre CISP to 1476). Thus, when we ask whether the CISP program was able to meet these targets, we are also asking whether the proposition that there was a large unmet service need in the courts was valid. Case location Referral targets Latrobe Melbourne Sunshine Total Valley 184 1476 656 2316 Targets for CISP referrals 150 1154 446 1750 Annual referrals 2007 276 1147 581 2004 Annual referrals 2008 150% 78% 88% 86% 2008 referrals as a % of target Table 2.9 CISP targets and actual referrals 2007 and 2008 The analysis presented here shows that CISP was able to generate large increases in the number of referred defendants over pre-cisp levels (Table 2.9). Overall, there were 2004 referrals in the 2008 year, equivalent to 86% of the target figure. The increase in case Melbourne Criminological Research and Evaluation 40

flow at the Latrobe Valley venue was very large. The 276 referrals in 2008 are half as great again as the target number of 184, and represent a seven-fold increase in referrals over pre-cisp levels. Increases at Sunshine and Melbourne were relatively smaller but nevertheless still represent large increases over pre-cisp levels. Both Sunshine and Latrobe Valley show significant upwards trends in referral rates. Case flow numbers in isolation provide only limited information about the effectiveness of a program. In order to make a judgement about whether CISP has achieved its goals it is also necessary to consider: Whether the risks and needs of clients are appropriate? That is, were the additional clients who were referred and engaged the kind of people likely to benefit from the services and supports offered through the program? This issue is examined in detail in the next chapter of this report. Is the case-flow consistent with the level of resourcing for the program? While this evaluation did not include an examination of staffing and workload issues 8, it is relevant to note that the translation of referrals into engaged clients was relatively high (out of every ten referrals, seven became engaged clients), and that the period of involvement with clients was around four months for completing clients and two months for those who did not complete. Case flow and judicial support The CISP model depends on achieving a high level of judicial support. There are two measures that indicate that this was achieved (agreement between worker recommendations and the referral of clients directly by magistrates) and one that is more difficult to interpret (the falling rate of judicial monitoring). In general, Magistrates endorsed the recommendations of CISP workers about program engagement. On the question of whether or not a defendant was suitable for the program, Magistrates endorsed around 90% of worker recommendation, and on the more specific issue of the level of program engagement, Magistrates confirmed workers recommendation in around 80% of cases. Over the period of the program, Magistrates showed an increasing preparedness to refer defendants directly. 8 These issues were the subject of a separate review carried out in 2008. Melbourne Criminological Research and Evaluation 41

It is unclear why there are such large differences between the three program sites in the rate of judicial monitoring, or why the rate of judicial monitoring has declined. Judicial monitoring is central to the idea of therapeutic jurisprudence, and the large fall in its use (from 77% of cases to less than half) is of concern. The decline in monitoring cannot be attributed to changing referral patterns, as the proportion of magistrate-initiated referrals increased from 2007 to 2008, and while self-referrals also increased they only accounted for 4% of all engaged clients. Other possible reasons include greater confidence by magistrates in CISP case management, increased work pressure on magistrates, and the difficulty in organizing review hearings experienced by magistrates who work at several court venues. It should be noted that there appears to be little if any relationship between the use of monitoring and program completion. These issues are taken up in more detail in Chapter 8 of this report. Melbourne Criminological Research and Evaluation 42

Part 2: CISP Case management and intervention model A fundamental principle in therapeutic jurisprudence is that how courts respond to offenders should be targeted at the problems and disadvantages that give rise to criminal behaviour. For the CISP to work effectively it must identify these problems and disadvantages, provide the court with information about offenders that assists in decisions about bail, judicial monitoring and sentencing, and establish case management plans that respond to these issues. The CISP model is based on multi-disciplinary teams offering a range of specializations, the members of which work together to screen and assess defendants and respond their identified needs in a way that matches the level of intervention received to the level of risk and need of the defendant. A key element in the process is improving defendants access to, and the co-ordinated delivery of social and health services. This part of the evaluation report examines the case management and intervention processes within CISP. The analysis presented includes: A description of the demographic and problem characteristics of CISP clients A description and critical analysis of the CISP assessment and case management model, and Detailed examination of the referral processes for three main forms of intervention: drug and alcohol programs, mental health programs and accommodation support. Melbourne Criminological Research and Evaluation 43

Chapter 3 Client characteristics and service needs Relatively little is known about the problems and disadvantages associated with defendants at court. While the health and associated problems of custodial population have been extensively and systematically studied, information about court defendants is patchy. Studies of arrestee populations (most notably, the DUMA research conducted by the Australian Institute of Criminology) show high rates of prevalence of drug, alcohol and mental health problems, but these studies have not (until recently) included samples of Victorian arrestees. This chapter reports on the demographic, substance abuse, health and mental health characteristics of CISP clients in order to address the following evaluation issues: The demand for programs and services The relationship between client needs and risks and program case flows and outcomes The demand modelling and service delivery model for CISP was based on assumptions about the characteristics of defendants, and this chapter also provides commentary on how the CISP model should evolve to take account of the knowledge of client characteristics that has accrued since the beginning of the program. Unless otherwise stated, the material presented here relates to the cohort of engaged CISP clients who took part in the program from its commencement in December 2006 until March 2009. The information presented here is derived mainly from the screening assessments carried out by case managers when clients are first referred to CISP. Most of the information collected is self-report (although case managers may be able to validate some information from other sources) and should not be regarded as diagnostic. The references in this report to drug, alcohol and mental or physical health problems should be read only as identification that these problems were regarded as appropriate as a basis for engagement in CISP. Where these problems are identified by CISP, offenders may be referred to clinical or other specialists who are able to make a more authoritative diagnosis or assessment. However, this information was not available to the evaluation team. Melbourne Criminological Research and Evaluation 44

Note that only clients who were engaged with program levels 2 or 3 (Intermediate or Intensive) received a full screening assessment. Where appropriate, advice from agencies providing services to CISP clients is also reported. The assessment data is also compared with the results of a structured health and mental health survey conducted on a sample of 197 CISP clients. 3.1 Client demography The prevalence of health problems and social service needs varies greatly within the population, and any assessment of these issues must begin with some consideration of which segment of the population forms the CISP client group. Around four in five CISP clients (81.4%) were male, and there was a high level of consistency in the gender mix of clients across the three venues (Figure 4.1). Female clients comprised 16.4% of clients on the Intermediate program level, and 24.3% of those on the Intensive program level. Clients had a mean age of 32.7 years, and again there were no differences in the age distribution of clients between the program venues or program levels. Half of all clients were aged between 26 and 37 years, with around one-quarter of clients aged under 26 years, and one-quarter aged more than 37 years. In some later analyses these groups are referred to as young (under 26), middle (26 to 37) and older (more than 37). Sex of CISP clients by program venue 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% Male Female 30.0% 20.0% 10.0%.0% Latrobe Valley Melbourne Sunshine Figure 3.1 Sex of CISP clients Melbourne Criminological Research and Evaluation 45

Clients who identified as indigenous (Aboriginal, Torres Strait Islander or both) comprised 8.1% of all CISP clients, but the proportion of indigenous clients was significantly greater at Latrobe Valley and Melbourne than at Sunshine (12.3%, 10.5% and 2.6% respectively see Figure 3.2). Relatively more female clients were indigenous than male clients (11.7% female versus 7.3% male), and indigenous clients were also relatively more likely to be on the Intensive program level than the Intermediate level (11.4% engaged as Intensive versus 6.7% engaged as Intermediate). Proportion of indigenous (Aboriginal or Torres Strait Islander) clients 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0%.0% Latrobe Valley Melbourne Sunshine Figure 3.2 Indigenous clients by program venue It is useful to compare these demographic characteristics with those of two reference groups: arrestees and prisoners. The arrestee data is the first Victorian data from the Drug Use Monitoring Australia research (Adams, Sandy, Smith, & Triglone, 2008) and the prisoner data is from the national prisoner census for 2008 (Australian Bureau of Statistics, 2008). Characteristic CISP clients Victorian arrestees Victorian prisoners % Female 18.6% 25.2% 5.6% % Aged 26 or less 26.1% 29.0% 16.4% % Indigenous 8.1% n.a. 5.8% Table 3.1 CISP client characteristics compared with arrestees and prisoners The CISP population has an age distribution that is generally similar to that of arrestees, and younger than prisoners. There are more indigenous and female offenders in the Melbourne Criminological Research and Evaluation 46

CISP client group than in the prison population. While there appear to be fewer females in the CISP population than for Victorian arrestees, it should be noted that the Victorian DUMA sample is small (two quarters of data only) and amongst Australian arrestees generally around 16% were female. 3.2 Prevalence of drug problems Drug problems are assessed according to a four-level classification that distinguishes between no use, use, abuse and dependence. The criteria for the most serious level of the scale (dependence) include: tolerance (the need for larger amounts of the drug over time), withdrawal (physical or psychological effects associated with cessation of use), and an inability to voluntarily cease or restrict use of the drug. The criteria for the second level of the scale (abuse) include recurrent use resulting in social, inter-personal or legal problems or physical hazards. Drug use involves irregular or regular use that does not satisfy the criteria for abuse or dependence. Assessments of drug problems are made for each type of illicit drug that the person reports using, and for any individual the severity of drug problems may vary greatly from one drug to another. Case location Drug problems Latrobe Melbourne Sunshine Total Valley Current use of drugs reported N 190 868 639 1697 % 65.1% 68.5% 80.9% 72.2% Past or present IV drug use N 114 709 503 1326 % 39.0% 56.0% 63.7% 56.4% Current pharmacotherapy N 27 299 268 594 % 9.2% 23.6% 33.9% 25.3% Overdose history N 45 261 143 449 % 15.4% 20.6% 18.1% 19.1% Total N 292 1267 790 2349 % 100.0% 100.0% 100.0% 100.0% Table 3.2 Clients with identified drug problems (2007 & 2008) by CISP venue Overall, around 70% of all engaged CISP clients reported some level of illicit drug use (Table 3.2). Drug use was more prevalent at Sunshine than at Melbourne, and Melbourne in turn recorded higher rates than Latrobe Valley. This pattern was evident Melbourne Criminological Research and Evaluation 47

across a range of measures of drug problems: rates of IV drug use; reported current engagement in pharmacotherapy; and overdose history. Where drug use was identified, clients were assessed in relation to each drug where use was reported. On average, clients reported using around 1.5 drug types. The most commonly reported drug was cannabis, and over half of those who reported using it were assessed as dependent. Heroin was the next most commonly reported drug, and again over half of those assessed were recorded as dependent. Amphetamines, methamphetamine and benzodiazepine use were also commonly reported and in each case between half and two-thirds of users were assessed as either dependent or abusers of the drug. This general pattern of drug use closely matches the pattern observed for arrestees in the DUMA research, where cannabis was found to be the most common drug detected through urinanalysis, followed by heroin and amphetamine/methylamphetamine and benzodiazepines. Melbourne Criminological Research and Evaluation 48

Drug type Cannabis Level of Drug Use Use Abuse Dependence None Unknown Total N 266 144 475 7 14 906 % 29.4% 15.9% 52.4%.8% 1.5% 100.0% Heroin N 116 134 352 4 5 611 % 19.0% 21.9% 57.6%.7%.8% 100.0% Amphetamines N 150 144 116 6 7 423 % 35.5% 34.0% 27.4% 1.4% 1.7% 100.0% Methylamphetamine N 55 70 96 1 2 224 ( Ice ) % 24.6% 31.2% 42.9%.4%.9% 100.0% Benzodiazepines N 29 69 46 0 1 145 % 20.0% 47.6% 31.7%.0%.7% 100.0% Ecstasy N 55 26 5 5 0 91 % 60.4% 28.6% 5.5% 5.5%.0% 100.0% Other N 27 52 52 0 4 135 % 20.0% 38.5% 38.5% 0% 3.0% 100.0% Total N 698 639 1142 23 33 2535 % 27.5% 25.2% 45.0%.9% 1.3% 100.0% Table 3.3 severity of drug problems by drug type Age was the most important demographic characteristic related to drug problems. Clients in the older age group (37 years or more) were much less likely to report drug problems than those in the middle (26 to 37) or young (under 26) groups. Women clients reported somewhat more drug problems than male clients. Women were more likely to report heroin use, IV drug use and a history of overdose, however these differences were small. In the case of indigenous clients, around one-quarter had an unknown status in relation to drug problems and it is therefore difficult to draw any conclusions about the prevalence or severity of drug problems in this client group. Melbourne Criminological Research and Evaluation 49

3.3 Prevalence of alcohol problems The assessment process for alcohol problems is similar to that for drug problems. Engaged clients are asked whether they currently use alcohol, and those who report any alcohol use are assessed to determine whether this constitutes abuse or dependence on alcohol. The screening process for alcohol problems examines self-reported problems with alcohol, the presence of alcohol-related offences (eg. drink-driving) and the involvement of alcohol in offending. There were very large differences between the program sites in the proportion of clients who were recorded as having any current use of alcohol (and who then went on to the next assessment stage), and in the severity of alcohol problems. Case location Latrobe Melbourne Sunshine Total Valley Any alcohol 191 507 320 1018 use? 65.4% 40.0% 40.5% 43.3% Severity of alcohol problems Use N 62 182 90 334 % 32.5% 35.9% 28.1% 32.8% Abuse N 46 201 68 315 % 24.1% 39.6% 21.2% 30.9% Dependence N 81 97 158 336 % 42.4% 19.1% 49.4% 33.0% None N 1 17 1 19 %.5% 3.4%.3% 1.9% Unknown N 1 10 3 14 %.5% 2.0%.9% 1.4% Total N 191 507 320 1018 % 100.0% 100.0% 100.0% 100.0% Table 3.4 Alcohol problems by venue Nearly two-thirds of Latrobe Valley clients (65.4%) were recorded as having any current alcohol use compared with around 40% at the other two venues (Table 3.4). However, there were no consistent differences between the three sites in the severity of alcohol problems. Overall a third of those assessed for alcohol problems were classed as dependent, but this varied from nearly half of those assessed at Sunshine to less than one-fifth at Melbourne. In our interviews with service providers, magistrates and case Melbourne Criminological Research and Evaluation 50

managers, there was a consensus that alcohol problems are more prevalent at the Latrobe valley venue. Alcohol use was more prevalent amongst male clients and where present alcohol problems were more likely to be assessed as dependent. As with drug problems, the proportion of Indigenous clients with unknown alcohol use is high (around 20%) and this makes any comparison problematic. 3.4 Prevalence of mental health problems, ABI and intellectual disability The screening assessment asks case managers to identify whether the client has any indications of mental health problems, acquired brain injury or intellectual disability. While the assessment questions include a range of known indicators of these conditions, their purpose is solely to identify clients who may require more detailed clinical assessment by a psychiatric nurse, disability worker or ABI clinician. Where serious mental illness is identified, responsibility for the client may be transferred to Forensicare or in the case of intellectually disabled persons charged with serious crimes, to Disability Forensic assessment and Treatment Services. Across the program, just over one-third of all clients were identified as having a possible mental health problem, and there was little variation between the program venues in the prevalence of these conditions. Where mental health issues were identified, the majority of these clients were receiving some form of treatment (39.8%), had received treatment in the past (14.2%) or had a current diagnosis of their condition (20.7%) (Table 4.5). Around one in ten clients with an identified mental health problem had never received treatment, been assessed but not treated, or were currently being assessed. Melbourne Criminological Research and Evaluation 51

Prevalence of mental health problems 50.0% 45.0% 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0%.0% Latrobe Valley Melbourne Sunshine Figure 3.3 Mental health problems by CISP venue Mental health problems were much more common in women than in men (48.4% versus 34.7%) and became more prevalent as clients got older: 30.5% of young clients had an identified mental health problem compared with 38.0% of clients in the middle age group and 42.2% of clients in the older age group. Mental health problem status was unknown for around one quarter of Indigenous clients. Mental health treatment history N % Currently receiving treatment 496 39.8% Has a current diagnosis 258 20.7% Received treatment in the past 177 14.2% Requires assessment 138 11.1% Current client of Mental Health Service 104 8.3% Currently undergoing assessment 43 3.5% Never received assessment or treatment 48 3.9% Received assessment but no treatment 27 2.2% Previously completed assessment 79 6.3% No treatment required at this time 28 2.2% Current client of Disability Services 4.3% Total clients 1246 100.0% Table 3.5 Treatment status for mental health problems at CISP assessment Melbourne Criminological Research and Evaluation 52

Acquired brain injury The screening assessment included several items that are indicators of acquired brain injury (head injury, periods of unconsciousness, inhalant use, long term drug or alcohol use). A total of 174 clients or about one in ten (8.9%) was identified as having indicators of ABI. Full assessment for ABI is the responsibility of arbias and client ABI status was updated at the time of program exit, when a total of 141 clients were recorded with ABI status. The proportion of clients with ABI status was significantly higher at Latrobe Valley than at the other CISP venues (see Figure 4.4), possibly as a result of the higher level of alcohol abuse at that site. There was no difference in the prevalence of ABI indicators between men and women, but there was a significant relationship with client age. Only 3.2% of young clients were recorded as having ABI status compared with 6.8% of those in the middle age group and 10.7% of those in the older age group. Acquired brain injury at CISP exit 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0%.0% Latrobe Valley Melbourne Sunshine Figure 3.4 Acquired brain injury at CISP exit Intellectual disability A total of 87 clients were recorded as having intellectual disability, of whom 67 were clients of Disability Services. Melbourne Criminological Research and Evaluation 53

3.5 Combinations of problems Around one third of CISP clients were recorded as having more than one offending related drug, alcohol or mental health problem 9. The most common combination was drug and mental health problems, present in 16% of CISP clients. Drug and alcohol problems in combination accounted for a further seven percent of clients. Around five percent of clients were assessed as having all three problems. Frequency Percent Drug and alcohol problems 172 7.2% Drug and mental health problems 383 16.0% Alcohol and mental health problems 137 5.7% Drug, alcohol and mental health 126 5.3% problems Total 2387 100.0% Table 3.6 Combinations of offending-related problems 3.6 SF-12 survey of client physical and mental health The SF-12 is a short (12 item) psychometrically valid instrument for measuring health status. The SF-12 allows the self-reported physical and mental health status of the surveyed group to be compared with values for other groups or for the same group over time. The items that comprise the SF-12 are shown in the table below. The instrument measures functional health status: that is, it asks about the impacts and consequences of health issues rather than clinical or diagnostic details. While the instrument is not itself diagnostic, it has been shown that it can reliably distinguish between groups that differ in the severity of their health problems. The primary output measures are summary physical and mental health scores, however the instrument also allows a health profile of eight scales to be measured. 9 Drug problems were defined as current drug use plus any of the following: current involvement in a drug treatment program, a history of IV drug use or an overdose history. Alcohol problems were defined as alcohol use assessed as abuse or dependence. Mental health problems were defined as any identified mental health problem. Melbourne Criminological Research and Evaluation 54

For the CISP evaluation, the SF-12 was administered to 197 clients at the time they commenced the program, and 67 were re-tested at or near their exit from the program. Participation in the survey was voluntary, and the survey questions were administered verbally by case managers who also recorded the clients responses. SF-12 scales Number of questions Physical functioning 2 Role limitations because of physical health problems 2 Bodily pain 1 General health perceptions 1 Vitality 1 Social functioning 1 Role limitations because of emotional problems 2 General mental health 2 Table 3.7 SF-12 scales Pre-CISP physical health The first (pre-cisp) administration of the SF-12 provides a measure of how the physical health status of clients compared with other groups in the community. The instrument is constructed so that the general population mean for physical health is a score of 50 with a standard deviation of around 10. The mean Physical Component Score (PCS) for the CISP sample at program entry was 49.8 (with a standard deviation of 10.8). Figure 4.1 shows the CISP PCS score together with scores for comparison general population samples in Australia and the USA, and a sample of US homeless people. It can be seen that the CISP sample mean is essentially the same as that for the general population samples, and higher than that of the homeless group. Melbourne Criminological Research and Evaluation 55

Mean Value 70 60 50 40 30 Physical Component Score 1 CISP Australian Community U.S. Community Australian Clinical U.S. Homeless Figure 3.5 SF-12 mean physical component scores for CISP and comparison groups Women typically report slightly lower CS scores than men (around two points lower) and this was also the case with the pre-cisp samples, where men recorded a mean PCS score of 49.0 and women a PCS score of 46.6. Pre-CISP mental health The SF-12 is constructed so that the general population mean for mental health is a score of 50 with a standard deviation of around 10. The mean Mental Component Score (MCS) for the CISP sample at program entry was 37.4 (with a standard deviation of 12.5). Figure 4.2 shows the CISP MCS score together with scores for comparison general population samples in Australia and the USA, and a sample of US homeless people. It can be seen that the CISP sample mean is much lower than the general population samples, and somewhat lower than that of the homeless group. MCS scores for women entering CISP were even lower (at 31.2), compared with 38 for men entering CISP. Melbourne Criminological Research and Evaluation 56

Mental Component Score CISP Mean Value 70 60 50 40 30 1 Australian Community U.S. Community Australian Clinical U.S. Homeless 3.6 SF-12 mean mental component scores for CISP and comparison groups Comparing pre- and post-cisp scores Clients who completed CISP were also asked to do the SF-12 again at or near the time of their exit from the program. Around 40% of those who are engaged as CISP clients do not complete, and of those who do a proportion exit before a final meeting with their case manager is scheduled (usually when they are sentenced). As a result, only 67 second round surveys were completed. A comparison of the pre- and post-cisp SF-12 Physical Component Scores showed an increase in the mean score from 50 to 54, and mean Mental Component Scores also increased from 38 to 45. Paired sample t-tests showed that both these increases were statistically significant 10. 10 Pre- and post-cisp Physical Component Scores: t=3.56, df=66, p<0.001 Pre- and post-cisp Mental Component Scores: t=4.3, df=66, p<0.001 Melbourne Criminological Research and Evaluation 57

SF-12 Component Score pre and post-cisp 60.00 50.00 40.00 30.00 Pre-test Post-test 20.00 10.00.00 Physical Component Score Mental Component Score Figure 3.7 Pre- and post-cisp PCS and MCS scores As only about one-third of those who completed the first round SF-12 went on to complete the survey a second time at or near exit, it is important to check that there were no systematic differences between the two groups in their entry scores. An analysis of variance test comparing the first round scores for those who, and didn t go on to the second round showed no significant difference in their first round scores. 3.7 Comments on client characteristics and service needs The CISP client group exhibits the high prevalence of drug, alcohol and mental health problems known to be a common feature of offender populations. Interestingly, their self-reported physical well-being shows scores comparable to those of the non-offending population, indicating that any detrimental physical impacts of their drug and alcohol use are not apparent. This may be related to the relative young age of most clients. There were two unexpected findings about client characteristics. The first was the high prevalence of acquired brain injury in the group. The Australian Institute of Health and Welfare (1999) provides five Australian population estimates of the prevalence of ABI, of Melbourne Criminological Research and Evaluation 58

which four are in the range 0.2% to 0.4% 11. Bearing in mind that the CISP only provided a screening assessment designed to identify persons with indicators of possible ABI, around 9% of clients were in this category. Chapter 6 provides further information on ABI status at the end of clients engagement with CISP. The second unexpected finding was the very low levels of self-reported well-being for mental health. While the prevalence of mental health issues is relatively high compared with the general population, their self-reported well-being is significantly below community levels. 11 The report cites rates of 161, 240-290, 290, 400 and 1,696 per 100,000. The higher figure is based on a population modelling approach and is for brain injury in general. Melbourne Criminological Research and Evaluation 59

Chapter 4 CISP service model The CISP service model is complex. The program provides a wide range of services to clients and the courts (assessment, referral, reporting, supervision, and direct support) across a variety of service sectors (mental health, ABI, alcohol and drug, accommodation, employment, social support) each of which has its own distinctive referral and service delivery processes. Prior to the establishment of CISP court services were provided through a number of specialised programs that were available in certain court venues. These included CREDIT/Bail Support program, an Aboriginal Liaison Officer, Disability Co-ordinator, and Mental Health Liaison Officers. The establishment of CISP involved the integration of these service functions into the CISP teams, although mental health liaison officers remained staff members of the Victorian Institute of Mental Health. In addition, service agreements were established with housing agencies as part of the Justice Housing Support Program, with arbias for the provision of Acquired Brain Injury assessment and support services, and with the Community Offenders Advice and Treatment Service (COATS) for drug and alcohol assessment and referral services. The relationships between CISP and service agencies are set out in protocols that specify the staffing, financial and other responsibilities and obligations between the parties, and the referral, communication ands reporting systems. For some service arrangements, service agency staff members are located at CISP venues and deal directly with clients. This chapter provides an overview of the CISP service model and examines: The clinical and procedural basis of the model, and how this aligns with the goals of the program; Trends and patterns in the level of service activity (referrals and interventions); Client feedback on the support provided to them; Support systems for the service model (assessment processes, policy and procedures manual, Platypus Systems Case Management System); and Staffing issues associated with the service model Chapters 5 to 7 provide more detailed analysis of service issues associated with drug and alcohol services, mental health services and accommodation services respectively. The Melbourne Criminological Research and Evaluation 60

material presented in these four chapters is based on interviews with staff members and team leaders at all three program venues, interviews with clients exiting the program, review of program documentation and assessment and referral data from the PSCMS. 4.1 Clinical and procedural basis of CISP The primary features of the CISP service model are: A multidisciplinary team-based approach; Case managers carry out a range of assessment, compliance, reporting, support and referral functions; Clients are allocated to the three program levels based on their assessed risk 12 ; Once engaged, a client is the responsibility of a CISP case manager, but other specialists may be consulted; CISP provides a range of direct services, with drug, alcohol, ABI and accommodation service interventions delivered by contracted service providers; Additional services may be delivered by referral to external agencies, with brokerage funds available to pay for a range of treatment and support services including emergency accommodation, pharmacotherapy assessment or treatment, and education or other programs. Multidisciplinary team approach The multi-disciplinary team approach is generally regarded by program staff as successful in addressing the complex needs associated with the program s clients. Workers valued the capacity to consult with other staff members who have more specialized knowledge in particular areas, and to be able to refer clients with particular service requirements to mental health, arbias and housing workers who were on-site. Magistrates commented favourable on the capacity to have defendants problems addressed directly and quickly, and to have independent advice to the court on a range of complex problems. Much of the interaction between team members is informal (one spoke of the approach as an all in together model), and it is apparent that the level of interaction between case managers is higher than that between case managers and workers from the contracted 12 The allocation of clients across program levels is detailed in chapter 2. Melbourne Criminological Research and Evaluation 61

service provider agencies, who often felt that they were in a twilight zone between being accepted as part of the CISP team and feeling like you were not part of it. It was acknowledged that the CISP model had developed substantially over the life of the program, from its origins in the CREDIT/Bail Support Program. Several who had worked on the CREDIT /Bail Support program prior to coming to CISP commented on the advantages of the CISP model. However, the multi-disciplinary approach also poses problems of adjustment for new team members, and in the administration of the program. Some workers who had come from a service area where they had been in specialized roles found it a challenge to adjust to the more varied demands of CISP. One commented that if you re a disability specialist and you expect that your caseload will consist solely of clients with disability issues, then that s a shock. Ideally, clients and workers should be matched on the basis of the clients priority needs, but high workloads meant that this was not always possible. The high prevalence of drug and alcohol problems in the client group means that there is a correspondingly high demand for workers with drug and alcohol skills. Those from other areas of specialisation noted that they had to pick up drug and alcohol skills in order to be able to work effectively. Workloads also impeded workers capacity to consult with their colleagues about clients problems, and a commonly expressed source of frustration was the inability to engage in systematic case reviews for clients with more complex problems. Weekly allocation meetings involve CISP team leaders, case managers and contracted services were implemented in August 2009. The purpose of the meeting was to assist with the equitable allocation of case load for staff and enhance a collaborative approach to the management of clients. These meetings commenced in the very final stages of this evaluation and we are therefore unable to offer any comment on their impact on these issues. Workers roles CISP case managers work involves a wide variety of roles, including carrying out assessments, reporting to the court, providing direct support to clients, supervision and compliance management, and making referrals to contracted and other services. The role of case manager therefore involves elements of clinical forensic practice in combination with the compliance and reporting responsibilities of a court officer. From the beginning Melbourne Criminological Research and Evaluation 62

of the program there has been uncertainty on the part of CISP staff about how these roles should come together. Staff members feel a responsibility to engage with clients directly (a clinical or case manager role) but are also aware that this role is time-limited and will ultimately involve the client moving on to a longer-term relationship with one or more treatment or support agencies (a brokerage/advocacy role). A related issue is that the client-worker relationship that is central to conventional clinical practice must also take into account the responsibilities to the court that are inherent in the CISP case management role. Workers who had come to CISP from other service areas often found the legal aspects of the program very challenging. The support of Magistrates was seen as fundamental to the success of the program, but it was recognized that different Magistrates had quite different expectations of the program, and that it took some time learn how to manage and respond to these differences in approach. A related problem was that workers felt that there was a lack of recognition of their expertise by other court professionals. In particular, lawyers were inclined to challenge in court recommendations they didn t like. Typically this involved defendants who were assessed as unsuitable for the program, with the result was that workers were required to justify their recommendations in the adversarial environment of the court. Staff turnover has been a significant problem in the operation of the CISP service model. All those consulted in the course of this evaluation (staff members, Magistrates and external stakeholders) commented on the high rate of staff turnover and its consequences. These consequences included difficulties in establishing stable and cohesive program teams, ongoing requirements for training of new staff, and increased workloads on existing team members. External stakeholders referred to the problems arising from dealing with staff members who were unfamiliar with court procedures or the operations of external services. Team leaders The role of team leader is central to the effective operation of the CISP model, and both workers and Magistrates commented on the difference that having an experienced team leader made to the quality of the work. Given the high rate of staff turnover in CISP, Melbourne Criminological Research and Evaluation 63

team leaders have to devote a substantial amount of time to the training and development of new staff, as well as carrying a 50% caseload. Team leaders are also the primary point of communication between CISP and the other court professionals. At the Latrobe Valley and Sunshine sites the line management for the CISP team is through the Senior Registrar, but this role is mainly concerned with administrative issues concerning the team and responsibility for clinical and case management issues rests with the team leaders. While some CISP team leaders came into the role with substantial experience, new team leaders find the role very challenging. While training has been provided to team leaders in case management, assessment, leadership and staff management, team leaders also report that direct observation and hands-on experience are critical elements in the development of these skills. Program venues Some important variations are apparent in the way the three program venues operate. Melbourne is distinguished from Sunshine and Latrobe Valley in that it is a much larger and more complex court environment, with greater work pressure demand on workers and more variation across Magistrates in their expectations about the program. In contrast workers at Sunshine and Latrobe Valley have more direct knowledge of Magistrates and other court staff. One significant variation between the sites is in the way teams are managed. At Melbourne line management is through the Program Manager, which at Sunshine and Latrobe Valley line management is through the Senior Registrar. Both arrangements have their strengths but the existence of both simultaneously gives rise to uncertainty about who is responsible for the various CISP functions. Geographical barriers to access by clients are an important issue for the Latrobe Valley site. Many defendants at the Morwell Court come from communities where only limited public transport access is available. The court also sits at other sites in Gippsland like Korumburra, Wonthaggi and Bairnsdale where the program is not available. Some of the referral services available to the program are also located in other communities in the valley, and again this can mean that clients find it difficult to access these on a regular basis. Melbourne Criminological Research and Evaluation 64

4.2 Referrals and interventions A total of 3,661 referrals to treatment and support services were made in 2007 and this increased by three-quarters to 6,371 referrals in 2008. This is equivalent to 3.3 referrals per engaged client in 2007, and 5.1 referrals per engaged client in 2008. A large part of this increase was in material aid, which accounted for just over a quarter of all referrals in 2007 but around one-third in 2008 (see Tables 5.1 and 5.2). Referrals to selected services : 2007 and 2008 2500 2000 Number of referrals 1500 1000 2007 2008 500 0 Drug and alcohol - COATS Pharmacotherapy Mental health ABI - CISP Housing CISP Material aid Medical Figure 4.1 Referrals to selected services: 2007 and 2008 The volume of referrals increased across all categories. Referrals to contracted ABI services doubled between 2007 and 2008, referrals to the contracted housing services went up by 270%, and there was an increase of 464% in referrals to medical services. Some of this increase reflects the increasing caseload across the program as well as delays in getting contracted services established early in 2007. The COATS alcohol and drug services program was available at courts prior to 2007, and there was no increase in the average number of referrals per engaged client between 2007 and 2008. In contrast there were only four referrals to arbias in the first three months of 2007, compared with 70 in the next three months. Similarly, there was only one referral to the contracted accommodation service at Sunshine in the first three months of 2007, but 18 in the next three months when a SASHS worker was present. Melbourne Criminological Research and Evaluation 65

There are also some significant differences in referral patterns between the three program venues. Sunshine makes relatively more referrals than Melbourne (4.4 referrals per client in 2007 and 6.8 per client in 2008 compared with 2.5 and 4.2 at Melbourne in 207 and 2008 respectively) with Latrobe Valley falling between the two in its referral rate (4.2 referrals per client in 2007 and 4.3 in 2008). Again, much of the difference is accounted for by variations in the use of material aid, which accounted for half of all referrals at Sunshine in 2008. Latrobe Valley makes very few referrals for pharmacotherapy and contracted housing services, but makes ABI referrals at a higher rate than the other venues. These service specific differences are discussed in more detail in the following chapters. Melbourne Criminological Research and Evaluation 66

Service Referrals in 2007 Latrobe Valley Melbourne Sunshine Total Drug and Alcohol - COATS N 211 960 580 1751 % 50.4% 58.4% 36.3% 47.8% Drug and Alcohol - Non- N 31 8 3 42 COATS % 7.4%.5%.2% 1.1% Pharmacotherapy N 0 101 200 301 %.0% 6.1% 12.5% 8.2% Mental Health N 40 59 16 115 % 9.5% 3.6% 1.0% 3.1% ABI Services - CISP N 40 69 42 151 % 9.5% 4.2% 2.6% 4.1% ABI Services - Non-CISP N 2 0 1 3 %.5%.0%.1%.1% Housing - CISP HIR N 0 87 31 118 %.0% 5.3% 1.9% 3.2% Housing - Other N 4 23 0 27 % 1.0% 1.4%.0%.7% Disability Services N 6 8 1 15 % 1.4%.5%.1%.4% Material Aid N 53 256 701 1010 % 12.6% 15.6% 43.9% 27.6% Medical N 10 18 6 34 % 2.4% 1.1%.4%.9% Men's Behavioural Change N 18 30 9 57 % 4.3% 1.8%.6% 1.6% Other (inc. problem N 4 26 7 37 gambling, vocational, family % 1.0% 1.6%.4% 1.0% services) Total N 419 1645 1597 3661 % 100.0% 100.0% 100.0% 100.0% Table 4.1 CISP service referrals in 2007 Melbourne Criminological Research and Evaluation 67

Service Referrals in 2008 Latrobe Valley Melbourne Sunshine Total Drug and Alcohol COATS N 413 930 736 2079 % 50.7% 34.9% 25.5% 32.6% Drug and Alcohol Non- N 2 32 20 54 COATS %.2% 1.2%.7%.8% Pharmacotherapy N 7 179 266 452 %.9% 6.7% 9.2% 7.1% Mental Health N 85 120 70 275 % 10.4% 4.5% 2.4% 4.3% ABI Services - CISP N 64 144 106 314 % 7.9% 5.4% 3.7% 4.9% ABI Services Non-CISP N 20 7 4 31 % 2.5%.3%.1%.5% Housing CISP HIR N 7 219 104 330 %.9% 8.2% 3.6% 5.2% Housing Other N 23 38 20 81 % 2.8% 1.4%.7% 1.3% Disability Services N 1 19 4 24 %.1%.7%.1%.4% Material Aid N 116 603 1484 2203 % 14.2% 22.6% 51.3% 34.6% Medical N 11 113 34 158 % 1.3% 4.2% 1.2% 2.5% Men s Behavioural Change N 30 50 7 87 % 3.7% 1.9%.2% 1.4% Other (inc. problem N 36 212 35 283 gambling, vocational, family % 4.4% 8.0% 1.2% 4.4% services) Total N 815 2666 2890 6371 % 100.0% 100.0% 100.0% 100.0% Table 4.2 CISP service referrals in 2008 Melbourne Criminological Research and Evaluation 68

4.3 Client feedback A small group of CISP clients were asked to participate in an interview about their first hand experience of the program. Originally it was intended that a larger sample of CISP clients who were about to finish the program would be asked to do the interview, however logistically it was very difficult to arrange times to meet with clients before their sentence hearing. Instead members of the evaluation team spoke with clients as they came to the CISP office in MMC. In the interview clients were asked a set of questions that were designed to help them reflect and evaluate their time on the program. Question 1: Asked clients to indicate the main reason for referral to CISP: Most clients interviewed said that they were referred to CISP to help them with issues such as drug and alcohol addiction, homelessness, and mental health problems. A couple of respondents identified that participating in the CISP program was likely to improve their court outcome, and that this was their main reason for attending the program. Questions 2 &3: Asked clients whether they thought their access to treatment and support services improved with being on CISP, and if so how much. Eight of the nine respondents agreed that their access to services was improved by being on CISP. Comments about accessing services were generally very positive. Of those eight individuals who agreed that their access to services was enhanced by being on CISP, six thought that their access was a lot better, and two thought that their access was slightly better. Question 4: Asked clients to report how CISP improved their access to services. At least two respondents had difficulty answering this question, and didn t provide a response, for the individual who said their access wasn t enhanced this question wasn t asked. The six individuals that did respond to this question made comments such as the program: put you in contact with the right people or I wouldn t have known about Melbourne Criminological Research and Evaluation 69

[particular service] or Case managers explained the different services, type of counselling, and which was suitable. Question 5 & 6: Asked clients to rate how satisfied they were with their individual case manager, and what in particular they were satisfied or unsatisfied with. All nine respondents said that they were Very Satisfied with their individual case manager. Generally comments centred on the case managers high level of availability to the individual, their empathy, support and knowledge of the court system. Question 7: Asked clients to identify the most important aspect of support from CISP Answers to this question depended to a large extent on the reasons that the individual had been referred to CISP in the first place (e.g. drug and alcohol problems), however a number of comments noted that having a sense of support, loyalty, someone to talk to and getting people to listen were the most important aspects of their experience. Question 8: Asked clients to identify any improvements that could be made to the service Approximately half of the respondents could not think of any improvements. Some improvements that were suggested include: Better communication between CISP and legal aid; that the program was too short; that psychologists reports need to be able to be integrated better into CISP reports; and being able to access and get referrals to primary health care facilities. Melbourne Criminological Research and Evaluation 70

4.4 Support systems The CISP service model is supported by three elements of program infrastructure: a policy and procedures manual, a referral and screening assessment process, and a client records system (the Platypus Systems Case Management Systems PSCMS). Each of these systems was developed to support the specific features of the CISP program. Policy and procedures manual Work commenced on the CISP Policy and Procedures Manual in early 2007, but the completed manual was not issued to staff in its final form until late 2008 13. The manual set out the organisational arrangements for the program, eligibility requirements, case management procedures, professional development arrangements and policies on a range of matters including disclosure of information, records management, duty of care and OH&S. The manual is regarded highly by staff. Experienced staff members view the manual as a source of detailed information that is relevant to the day to day demands of the work. New staff members report that the information in the manual allows them to understand the duties and procedures involved in their work, and get up to speed quickly. This situation is contrasted with that before the manual was available when we were flying by the seat of our pants and new staff had to be extensively briefed by existing staff on the procedures involved in the program s operations. A number of staff members suggested additional material for inclusion in the manual, including advice on court procedures and information about the management of blood-borne viruses. Assessment process Two assessment instruments were developed for CISP: a referral assessment with basic details of defendants applying for entry to CISP, and a screening assessment to support decisions about program engagement. The Referral Assessment is typically completed by 13 Draft versions of the manual were in circulation in 2007 and early 2008. Melbourne Criminological Research and Evaluation 71

the defendant or defendant s solicitor, and the Screening Assessment is completed by the case manager assigned to do the initial assessment. The Screening Assessment was the subject of a detailed review in the first part of 2008 to determine whether it provides an effective basis on which to assess and classify the risks associated with CISP clients. The review was concerned with the following questions: How well is the Screening Assessment being completed? How frequently do the specific risks covered in the Screening Assessment occur? Is the distribution of clients across risk categories appropriate? Do the risk categories derived from the Screening Assessment align with the CISP program levels of clients? The review was conducted by selecting 40 cases from each of the three CISP program venues and examining the Screening Assessment forms for each case. These cases were selected on the following criteria: The client was assigned to Level 2 or Level 3 of CISP, and The assessment was completed after 30 June 2007. The review found that: The rate of completed data entry on Screening Assessment forms is high. For most data items, information is recorded on around 95% of forms. There were a few data items where recording rates were significantly lower (including Indigenous status, whether receiving pharmacotherapy and indicators of ABI and intellectual disability). The definitions and recording instructions associated with these items were subsequently reviewed in consultation with CISP staff. The specific risks in the Risk Assessment section all target risks that occur frequently in the CISP client population and the distribution of the numbers of risks across clients was appropriate. The sub-score sections of this part of the assessment were not always completed, and the risk classification section is completed in less than half of assessments. Where the risk assessment sections of the form were not completed this was frequently in cases where the case manager recommended or the court determined that the client was unsuitable for CISP. In cases where the risk classification section was competed, higher risk was strongly related to more intensive program involvement, both as a Melbourne Criminological Research and Evaluation 72

recommendation and court outcome. Two risk items were identified as having only a weak relationship with case outcome (problems with family relationships and inability to cope with custody). It was recommended that these items be removed from the form. Following the 2008 review the Screening Assessment form was revised to improve clarity and align the form more closely with the content and screen structure of the PSCMS. In interview with staff during 2009 there was a generally positive view of the Screening Assessment. I quite like the assessment as it is it s quick and works well in the cells, with people who are distressed or upset. It gives you the basics. It gets the ball rolling, and then you can see where you should go after that, in a more relaxed environment, after they re on the program. However, program staff also reported that assessment remains a difficult aspect of their work. Around 60% of assessments are conducted while the defendant is in custody and frequently when the defendant is distressed, anxious, suspicious or otherwise unable to provide accurate information. They noted that assessment is frequently done under considerable time pressure associated with the court process, and that this often takes priority over and interferes with other aspects of their work. Platypus Systems Case Management System The Platypus Systems Case Management System (PSCMS) is used to manage client records. The PCSMS records clients personal details, contacts with CISP, maintains an assessment and referral history, keeps track of court dates, generates reminders for a variety of events and produces a range of statistical reports on program activity. It is supported by a detailed manual, and new staff members are trained in the use of the system. Staff members views about the system are mixed. While most agree that the system provides ready access to client information and greatly assists in case management, there are concerns about the amount of time required to input data into the system. Melbourne Criminological Research and Evaluation 73

4.5 Recommendations: Recommendation: There should be a general review of clinical service approaches in the courts, taking into account the CISP, CREDIT/Bail Support, Neighbourhood Justice Centre, Drug Court, Family Violence Court and Koori Court services. The CISP service delivery model is one of a range of clinical and support program service approaches currently operating in Victorian courts. The other examples include CREDIT/Bail Support, the Neighbourhood Justice Centre client services function, Drug Court, Koori Court, and the applicant and defendant support functions for the Family Violence Courts. Each is based on different assumptions and service models, and each involves some specific skill sets. However, underpinning all these services is a general body of clinical knowledge and technique and common case management, support and other court-based functions. It is proposed that there should be a general review of these approaches with a view to creating a court support services function that would provide the basis for the delivery of a range of clinical, support, referral, supervision and case management services to court clients. Recommendation: The CISP Policy and Procedures Manual should be supplemented with training and procedural advice on assessment procedures and working in a court environment. The material on assessment procedures should include more detailed information about the clinical aspects of forensic practice, assessment criteria for the elements of the Screening Assessment, and the interpretation and integration of assessment information across multiple problem domains (especially substance abuse and mental health). The development of more explicit eligibility criteria may also assist staff by making such decisions more externally accountable. The material on working in a court environment should include advice on court processes, reporting in court, and the roles of other professional groups in the court (especially lawyers and police) and the development and management of effective relationships with them. Melbourne Criminological Research and Evaluation 74

Recommendation: The CISP Screening Assessment should be subject to regular review and staff should receive feedback on the aggregated results of assessments. It is important to maintain staff commitment to the structured assessment model for CISP. Two strategies for this are regular reviews of the assessment process (every second year) to ensure that the assessment content and procedures remain up to date and consistent with work practices, and demonstrating the value of the collection of assessment data through communication with staff about the results of assessments. Melbourne Criminological Research and Evaluation 75

Chapter 5 Drug and alcohol services 5.1 Provision of drug and alcohol services The provision of drug and alcohol assessment, detoxification and treatment services is a major component of the CISP service delivery model. Drug and alcohol service arrangements are set out in a protocol between the Department of Human Services, the Department of Justice and the Community Offenders Assessment and Treatment Service (COATS). COATS provide drug and alcohol assessment services and a treatment brokerage service. Under the brokerage model, COATS develops a treatment plan and purchases treatment services from a range of community alcohol and drug treatment and residential program providers for clients referred by Justice agencies including CISP. Some CISP case managers are also accredited drug and alcohol clinicians and can make assessments directly, and the COATS assessment service therefore supplements CISP in this area. Some drug and alcohol services can also be provided outside the COATS process, although this is usually only done as an interim measure while awaiting a COATS assessment. CISP case managers can also arrange assessments for pharmacotherapy through Directline, and where appropriate purchase pharmacotherapy from general practitioners and pharmacies. 5.2 Drug and alcohol service needs of CISP clients Drug and alcohol problems are highly prevalent in the CISP client group, with cannabis, heroin and amphetamines the most commonly reported drugs. The prevalence of drug problems varied across program venues, with the highest rates at Sunshine, followed by Melbourne, and the lowest rates at Latrobe Valley. The service needs of clients are also varied. A proportion of clients already have some involvement with a drug or alcohol treatment program. For those assessed as having current drug use, around 40% had some involvement in a drug program at the time of their entry to CISP, and 28% were involved with some form of pharmacotherapy program (mainly methadone). Around half of those with current pharmacotherapy recorded were also on a drug program. While the proportion of clients with some drug program involvement was slightly higher Melbourne Criminological Research and Evaluation 76

at Latrobe Valley and lower at Sunshine, the proportion on pharmacotherapy shows a substantially higher level of existing service involvement at Sunshine and a low level at Latrobe Valley. This appears to be consistent with the greater severity of drug problems at Sunshine (see Chapter 3). Case location Latrobe Valley Melbourne Sunshine Total Drug program N 85 372 230 687 involvement % 41.3% 38.6% 34.3% 37.3% Current N 16 254 245 515 pharmacotherapy % 7.8% 26.3% 36.5% 28.0% Total with current drug N 206 964 671 1841 use % 100.0% 100.0% 100.0% 100.0% Table 5.1 Drug program involvement for CISP clients Current alcohol Case location use level Latrobe Valley Melbourne Sunshine Total Abuse Alcohol N 7 25 5 37 program % 14.6% 11.0% 6.9% 10.6% involvement Total with alcohol abuse N 48 228 72 348 % 100.0% 100.0% 100.0% 100.0% Dependence Alcohol program involvement Total with alcohol dependence N 18 15 22 55 % 21.4% 13.6% 13.2% 15.2% N 84 110 167 361 % 100.0% 100.0% 100.0% 100.0% Table 5.2 Alcohol program involvement for CISP clients For those assessed with alcohol dependence or abuse, between ten and fifteen percent have some involvement with an alcohol program. For alcohol problems, there was a higher likelihood that clients at the Latrobe Valley site had an existing involvement with a treatment program than those at Melbourne or Sunshine. Melbourne Criminological Research and Evaluation 77

5.3 Service output measures Referrals to COATS make up the largest single category of CISP program referrals. There were 1,751 referrals to COATS in 2007 and 2,079 in 2008. Just over half the referrals (53%) were for assessment, although where the client is assessed as requiring some form of program intervention this is likely to result in referral to a detoxification, treatment or support program. A typical CISP case where there were four referrals for drug-related issues included: assessment by COATS, pharmacotherapy, and referral to the Counselling, Consultancy and Continuing Care treatment program while awaiting a place in a residential rehabilitation program. Taking into account all referrals to COATS in the period January 2007 to March 2009, 30% of cases were referred for one service (usually assessment), around half of cases were referred for two services, and 15% were referred for three or more services (Table 5.3). Number of Number of cases % of cases referrals More than 5 12 0.4% 5 28 0.9% 4 93 2.9% 3 341 10.6% 2 1803 55.9% 1 947 29.4% Total cases 3224 100.0% Table 5.3 Number of COATS referrals per case (2007 2009) COATS Assessments comprised a little over half of all referrals, and referrals to the Counselling, Consultancy and Continuing Care (Four-Cs) treatment program made up a further one-third of COATS services. About one in twenty clients were referred for drug withdrawal, and small numbers were directed to residential and youth drug programs. These referral patterns were stable over the two years 2007 and 2008, and there were few differences between the three program sites other than a higher rate of drug withdrawal services at Latrobe Valley (Table 5.4). Melbourne Criminological Research and Evaluation 78

Case location Latrobe Valley Melbourne Sunshine Total 2007 Drug and alcohol 94 (44.5%) 520 330 944 (53.9%) assessment (54.2%) (56.9%) Counselling, Consultancy 62 (29.4%) 363 216 641 (36.6%) and Continuing Care (37.8%) (37.2%) Drug Withdrawal 40 (19.0%) 39 (4.0%) 25 (4.3%) 104 (5.9%) Residential Rehabilitation 7 (3.3%) 24 (2.5%) 3 (0.5%) 34 (1.9%) Youth drug and alcohol 7 (3.3%) 10 (1.0%) 4 (0.6%) 21 (1.2%) programs Koori alcohol and drug 0 (0.0%) 1 (0.1%) 0 (0.0%) 1 (0.0%) services Supported Accommodation 0 (0.0%) 2 (0.2%) 2 (0.3%) 4 (0.2%) Total 211 (100.0%) 960 (100.0%) 580 (100.0%) 1751 (100.0%) 2008 Drug and alcohol assessment 188 (45.5%) 589 (63.3%) 408 (55.4%) 1185 (57.0%) Counselling, Consultancy 156 (37.8%) 304 288 725 (34.9%) and Continuing Care (32.7%) (39.1%) Drug Withdrawal 47 (11.4%) 23 (2.5%) 22 (3.0%) 92 (4.4%) Residential Rehabilitation 14 (3.4%) 11 (1.2%) 6 (0.8%) 31 (1.5%) Youth drug and alcohol 3 (0.7%) 2 (0.2%) 6 (0.8%) 11 (0.5%) programs Supported Accommodation 1 (0.2%) 0 (0.0%) 5 (0.7%) 6 (0.3%) Koori alcohol and drug services 2 (0.4%) 1 (0.1%) 1 (0.1%) 4 (0.2%) Total 413 930 736 2079 (100.0%) (100.0%) (100.0%) Table 5.4 Referrals to COATS drug and alcohol services: 2007 and 2008 (100.0%) CISP case managers can also make referrals to non-coats drug and alcohol programs. There were 42 such referrals in 2007 (of which 31 were made at the Latrobe Valley venue) and 54 in 2008. These non-coats referrals include a relatively higher proportion (around 15%) of referrals to residential programs. Melbourne Criminological Research and Evaluation 79

Pharmacotherapy is also an important form of service response for clients with drug problems. Frequently CISP referrals support a pre-existing treatment arrangement. There were 301 pharmacotherapy referrals in 2007 and 452 in 2008, the majority for methadone therapy (Figure 6.1). As might be expected from the higher rate of pre-cisp involvement in pharmacotherapy at Sunshine (see above), the rate of referrals for pharmacotherapy were higher at this venue. Two-thirds of all pharmacotherapy referrals in 2007, and 60% of those made in 2008, were from Sunshine. Referrals to pharmacotherapy: 2007 and 2008 350 300 250 200 150 100 50 0 Buprenorphine Methadone Naltrexone Suboxone Other Figure 5.1 Pharmacotherapy referrals by program type: 2007 and 2008. 5.4 Change in drug and alcohol problems during CISP engagement The status of CISP clients drug and alcohol problems is recorded at program entry, and again at program exit. In the case of alcohol problems, this assessment is made on the same basis: an overall assessment is made as to whether the client s alcohol use constitutes use, abuse, or dependence. Table 5.5 shows the distribution of alcohol status at program entry and exit for clients who successfully completed CISP. These data Melbourne Criminological Research and Evaluation 80

appear to show a dramatic reduction in the proportion with problems of alcohol abuse or dependence, with almost all those assessed with problems of this severity no longer having these problems at exit. In the case of clients drug use the pre- and post-program comparison is more complex because clients are assessed at entry in relation to each drug that they report using, while at exit only an overall assessment of their drug use is provided. However, these data represent case managers assessments of clients progress and need to be viewed with caution. The majority of drug and alcohol treatment interventions provided to clients are through the COATS brokerage program, and there is no provision for post-program assessment results to be reported back to CISP. Many clients will not have completed their treatment programs by the time they exit CISP, and for some, treatment will continue under court order. Another problem is that post- CISP outcomes are only known for those clients who complete the program. It seems likely that those who fail to attend or continue to offend are also those who do not make progress in addressing their drug and alcohol problems. Entry Exit Assessed alcohol Number of % Number of % problem status cases cases Use 179 33.0 183 33.7 Abuse 158 29.1 33 6.1 Dependence 196 36.1 43 7.9 None 10 1.8 244 44.9 Unknown 40 7.4 Total 543 100.0 543 100.0 Table 5.5 Assessed alcohol problem status 5.5 Relationship between CISP and drug and alcohol agencies The evaluation team met with the COATS and four drug and alcohol agencies that provide services to CISP (Odyssey House, Anglicare, DASWest and Latrobe Community Melbourne Criminological Research and Evaluation 81

Health 14 ). CISP case managers were also asked about their relationships with agencies. Drug and alcohol services to justice clients were well-established prior to CISP, and all the agencies consulted receive justice clients from a variety of sources in the justice system. CISP clients represent a small component in the overall stream of justice clients who go through COATS and the service agencies, and their relationship with CISP is seen in the context of this wider relationship. A comment that was made to the evaluation team on a number of occasions was that most drug and alcohol clients have histories that include involvement with the justice system. The various justice referral processes thus constitute pathways into treatment rather than attributes of a client that are important in terms of the services and supports that are offered to them. Clinical issues (treatment readiness, severity of drug or alcohol problems) and support issues (accommodation, family support and financial status) are much more important considerations, and much of the feedback from drug and alcohol agencies concerned how justice processes generally bore on these factors. The main issues identified through consultation with drug and alcohol agencies were: Clients are frequently referred from the justice system because their drug and alcohol issues constitute a justice problem, rather than because they are ready and motivated for treatment. While there is acknowledgement that the threat of justice outcomes can be important in generating motivation, a genuine personal commitment to change is more important. The punitive approach of justice programs is considered to be inconsistent with the harm-minimization approach of most treatment programs; The main perceived advantage of the CISP model is that clients also receive the social and material support that assists them to engage with drug and alcohol treatment; Availability of treatment places is an important constraint. This is particularly acute for residential programs where there can be lengthy waiting periods for a program place; Clients who have both substance abuse and mental health problems are particularly difficult to deal with. Agencies expect that CISP case mangers will make appropriate clinical decisions about whether a client is able to engage with a 14 Latrobe Community Health is also a provider of mental health and ABI services. Melbourne Criminological Research and Evaluation 82

drug or alcohol program. However it is acknowledged that this is a complex issue to determine that requires clinical experience with drug, alcohol and mental health issues; Generally, day to day relationships between agency workers and CISP case managers are good, although all but one agency reported examples of poor communication or inadequate knowledge of policies and procedures specified in the protocols on the part of CISP staff. Frequently this was ascribed to CISP staff being inexperienced; There is relatively little information about clients exchanged with COATS. The brokerage model means that direct contact should be between CISP case managers and treatment agencies. 5.6 Recommendations Drug and alcohol services are the most well-developed services component of the CISP model. The COATS brokerage model provides a broad-based assessment and treatment referral process, and draws on a range of community based service providers with substantial experience in the delivery of treatment and related programs. Recommendation: CISP program goals for drugs and alcohol should be concerned with effectiveness of the referral process and maintaining clients engagement with treatment programs. Improved drug and alcohol outcomes are part of the CISP goal set, however it seems inappropriate to hold the program responsible for treatment goals that are beyond its direct control. The key service delivery issues for CISP are how effectively it operates as a referral pathway and case management service. The indicators of success in this area should be whether drug and alcohol program referrals are based on a comprehensive assessment of clients risks and needs, and whether clients are provided with the support and supervision that ensures they satisfy the attendance requirements for drug and alcohol services. Melbourne Criminological Research and Evaluation 83

Chapter 6 Mental Health and Acquired Brain Injury Services 6.1 Provision of mental health and ABI services Court-based mental health assessment and advice services are provided by the Victorian Institute of Mental Health (Forensicare). A full-time mental health clinician is present at Melbourne Magistrates Court and Sunshine. Forensic mental health assessments at the Latrobe Valley venue are provided by the Latrobe Valley Community Mental Health Service which has a staff member at the court for around 2.5 days per week. Mental health clinicians are not part of the CISP multi-disciplinary team, but are co-located with the team members at Melbourne and Sunshine, and in practice there is a great deal of day-to-day contact between them. Direct case management and secondary consultation services for defendants with Acquired Brain Injury are provided by staff from the Alcohol Related Brain Injury Australian Service (arbias) who are attached to the CISP teams 15. This service comprises 1.5 EFT positions at Melbourne, 1.0 EFT at Sunshine and 0.5 EFT at Latrobe Valley. In addition, the protocol with arbias provides for two pro-bono full neuropsychological assessments each month for CISP clients. CISP case managers can also use brokerage funds for additional neuropsychological assessments by arbias or appropriately qualified clinicians. 6.2 Mental health service outputs for CISP clients Around 35% of CISP clients are identified as having some form of mental health problem, and of these around 40% are currently receiving treatment. Around one in eight clients (12.7%) with an identified mental health problem had never received treatment, had been assessed but not treated, or were currently being assessed. Case 15 Some CISP case managers are also undertaking training in ABI screening. Melbourne Criminological Research and Evaluation 84

managers made 115 referrals for mental health services in 2007, and 275 referrals in 2008. Unlike drug and alcohol referrals, most clients (87%) were referred for only one mental health service. The rate of referrals therefore represents one referral for every ten engaged clients in 2007, and one referral for every five engaged clients in 2008. There were large variations in the patterns of mental health service referrals between 2007 and 2008. In 2007, the largest number of referrals was for counselling services (41.7%), followed by referrals to the Area Mental Health Service (29.6%). However, in 2008 the largest group of referrals was to psychologists (including psychological assessment), accounting for 36.4% of referrals, followed by general practitioner referrals (18.9%). The number of Area Mental Health referrals was about the same across the two years, but fell as a proportion with the large rise in total mental health referrals in 2008. There were also significant variations in the rate of mental health referrals across the three program venues. Latrobe Valley made mental health referrals at a much high rate than the other two program sites. In 2007, there were 40 referrals across 100 engaged clients at this site, compared with 59 across 654 engaged clients at Melbourne in the same year. The pattern in 2008 was similar. This may relate to the problems in accessing the mental health court service at Latrobe Valley (see below) and the consequent use of external referrals to address this need. Assessing the change in clients status during their engagement with CISP is problematic. For those who were identified as having mental health issues at program entry, around half were recorded as also having issues at exit. In contrast, for those who weren t identified as having mental health issues at entry, 80% were also identified as not having them at exit, while nine percent not identified at entry were recorded as having issues at exit. However, it needs to be borne in mind that the CISP assessment is not a clinical assessment, and the change in status from entry to exit may represent better knowledge about clients mental health status rather than some change in condition for better or worse. Given the relatively short period of engagement with CISP, it seems unreasonable to expect any significant change in anything other than relatively minor mental health problems. Melbourne Criminological Research and Evaluation 85

2007 Counselling Area Mental Health Service General Practitioner Psychiatrist Psychologist Psychological Assessment Total Case location Latrobe Valley Melbourne Sunshine Total N 18 26 4 48 % 45.0% 44.1% 25.0% 41.7% N 18 8 8 34 % 45.0% 13.6% 50.0% 29.6% N 0 8 1 9 %.0% 13.6% 6.2% 7.8% N 0 4 0 4 %.0% 6.8%.0% 3.5% N 2 10 2 14 % 5.0% 16.9% 12.5% 12.2% N 2 3 1 6 5.0% 5.1% 6.2% 5.2% N 40 59 16 115 % 100.0% 100.0% 100.0% 100.0% 2008 Counselling Area Mental Health Service General Practitioner Psychiatrist Psychologist Forensicare Psychiatric Assessment Psychological Assessment Total N 8 16 5 29 % 9.4% 13.3% 7.1% 10.5% N 15 10 7 32 % 17.6% 8.3% 10.0% 11.6% N 4 24 24 52 % 4.7% 20.0% 34.3% 18.9% N 4 9 1 14 % 4.7% 7.5% 1.4% 5.1% N 6 38 11 55 % 7.1% 31.7% 15.7% 20.0% N 11 8 15 34 % 12.9% 6.7% 21.4% 12.4% N 2 7 5 14 % 2.4% 5.8% 7.1% 5.1% N 35 8 2 45 % 41.2% 6.7% 2.9% 16.4% N 85 120 70 275 % 100.0% 100.0% 100.0% 100.0% Table 6.1 Referrals for mental health services Melbourne Criminological Research and Evaluation 86

Mental Health issues indicated at CISP entry Mental health issues at CISP exit Yes No Unknown Total Yes No Unknown Total N 219 142 61 422 % 51.9% 33.6% 14.5% 100.0% N 55 512 62 629 % 8.7% 81.4% 9.9% 100.0% N 23 49 23 95 % 24.2% 51.6% 24.2% 100.0% N 297 703 146 1146 % 25.9% 61.3% 12.7% 100.0% Table 6.2 Mental health issues at CISP entry and exit 6.3 Service relationships for mental health The service relationships around mental health are complex in that Forensicare court staff can take referrals from CISP but can also take referrals direct from defendants lawyers. The Forensicare criteria for taking on a case are essentially the presence or suspected presence of a major mental illness, while CISP may also record other forms of mental disorder (like personality disorders, anxiety or depression) as a mental health issue. Thus, the mental health clinicians are part of the CISP team but also work independently (the Melbourne clinician estimated that the bulk of his work came directly rather than through CISP), and not all mental health issues can be referred through them. This process appears to work well in two locations (Melbourne and Sunshine). The mental health clinician provides a secondary consulting service for cases that fall outside the Forensicare eligibility criteria and there appear to be close and productive working relationships with the CISP team members. This is not the case at Latrobe Valley, where a clinician is only present on a part-time basis and there is little contact on a day-to-day basis with the CISP team. It was reported that there was little referral to this service from the court. A major problem with court mental health services is that difficulty in organizing referral to area mental health services. It was reported that these services were reluctant to take referrals of any clients who were not in crisis, and that CISP team members generally Melbourne Criminological Research and Evaluation 87

were required to organize a psychiatrists assessment before area mental health services would agree to take a referral. The lack of a specialist response for borderline and antisocial personality disorders was also acknowledged as a gap in this system, as these clients are likely to re-offend but have little access to community-based treatment services. 6.4 ABI service outputs for CISP clients One of the unexpected results of the CISP was the high rate of apparent ABI in the client group. Around one in ten engaged clients were identified at screening as having indicators of ABI, with a much higher identification rate at Latrobe Valley than the other two venues. The result was that there was a high level of demand for arbias assessment services. In 2007 there were 122 screening assessment conducted and 29 full neuropsychological assessments, and in 2008 the level of ABI assessment activity doubled to 247 screening assessments and 59 neuropsychological assessments. The increase in demand for ABI assessment services is also reflected in the use of non-arbias services. In 2007 there were three referrals for non-arbias assessments, but in 2008 this had increased 30. The relatively high level of identified cases at Latrobe Valley is not reflected in the rate of referrals for assessments which are more or less proportionate to caseloads, especially in 2008. For those who were identified as potential ABI cases at CISP entry, around one-third were recorded as having this status at exit. However, there were also significant numbers of cases recorded as unknown or not-abi at entry that had ABI status recorded at exit. At least some of the unknown cases at exit may also ultimately be determined to have ABI. In the middle stages of this evaluation there were long waiting times for full neuropsychological assessments, however by late 2008 additional assessment resources had been organized and delays had been reduced to around four weeks. Demand for ABI assessments is likely to remain an issue in the management of CISP. Melbourne Criminological Research and Evaluation 88

ABI assessments Case location Latrobe Valley Melbourne Sunshine Total 2007 Neuropsychological N 16 12 1 29 Assessment % 40.0% 17.4% 2.4% 19.2% Screening Assessment N 24 57 41 122 % 60.0% 82.6% 97.6% 80.8% Total N 40 69 42 151 % 100.0% 100.0% 100.0% 100.0% 2008 Neuropsychological N 6 29 27 59 Assessment % 9.4% 20.1% 25.5% 18.8% Screening Assessment N 57 112 78 247 % 89.1% 77.8% 73.6% 78.7% Case Management (Internal) N 1 3 1 5 % 1.6% 2.1%.9% 1.6% Total N 64 144 106 314 % 100.0% 100.0% 100.0% 100.0% Table 6.3 Referrals for arbias assessments Acquired Brain Injury indicated at CISP entry Yes Acquired brain injury at CISP exit Yes No Unclear Total N 36 30 36 102 % 35.3% 29.4% 35.3% 100.0% No N 12 730 54 796 % 1.5% 91.7% 6.8% 100.0% Unknown N 32 112 103 247 % 13.0% 45.3% 41.7% 100.0% Total N 80 872 193 1145 % 7.0% 76.2% 16.9% 100.0% Table 6.4 ABI at CISP entry and exit Melbourne Criminological Research and Evaluation 89

6.5 Service relationships for ABI The service relationship for ABI is simpler in that arbias workers are placed with the CISP teams and virtually all of their work is the result of referrals from CISP case managers. The arbias team has also experienced a fairly high rate of staff turnover, and at some venues no specialist ABI worker has been available at times. The most problematic issue for ABI services is the volume of referral and assessment work generated by CISP. As noted earlier, the number of neuropsychological assessments referred greatly exceeds the number that was anticipated in the planning for CISP, and one consequence was that waiting times for full assessments increased. It was reported that during late 2007 and early 2008 there were waiting times of six to eight months for full assessments, which greatly exceeds the period of time clients are likely to be on CISP. This problem has now been resolved. The problem of demand for ABI assessments is exacerbated by the uneven distribution of referrals in relation to the allocation of arbias staff. The CISP team at Melbourne refers relatively fewer cases to arbias staff than those at Sunshine or Latrobe Valley, but Melbourne has the largest arbias staff allocation. In this kind of situation it is important that initial screening of clients of clients is as effective as possible, that communication between workers is clear, and that protocols are followed. All of these issues have been sources of dissatisfaction for arbias, although there have been steps taken in 2008 to address these problems through staff training. 6.6 Recommendations While there are specific issues with the delivery of ABI services to CISP clients, it seems unlikely that they can be resolved without some more general strategy to address the needs of ABI clients in the justice system. This should include training of CISP staff in the management of ABI cases, but should also take into account the need to provide long-term outreach services for those clients. Melbourne Criminological Research and Evaluation 90

Recommendation 8: Review the provision of services to justice system clients with suspected Acquired Brain Injury The rate of suspected Acquired Brain Injury in program clients is much higher than allowed for in the demand modelling for CISP. This points to a high rate of ABI in justice client populations generally, and indicates that a comprehensive strategy to address this issue is required. While arbias and other ABI service providers offer a highquality assessment and service response for these defendants, the timescales involved are longer than can be accommodated in a pre-trial program. CISP case managers should receive additional advice on the management of such cases while awaiting full neuropsychological assessment process. There should also be consideration of the continuing management of these clients as they progress through other stages of the justice system. Melbourne Criminological Research and Evaluation 91

Chapter 7 Accommodation support services 7.1 Provision of housing services The Justice Housing Support Program (JHSP) is a collaborative project between the Homelessness Assistance Unit of the Office of Housing, Department of Human Services under the Victorian Homelessness Strategy, the Department of Justice, Magistrates Court of Victoria and the Neighbourhood Justice Centre. The program was established to provide clients in the justice system with transitional housing, referrals and links to support services in the short-term, and secure permanent housing in the longer term. Housing services are provided to the CISP program through HomeGround Services at Melbourne, the Salvation Army Social Housing Services (SASHS) at Sunshine 16, and Quantum Support Services (QSS) at the Latrobe Valley venue. The arrangements for the provision of housing advice and referral services to CISP are set out in the JHSP Inter-agency Protocols, issued in June 2008. This protocol covers the full range of JHSP services to the CREDIT/BSP, CISP and Neighbourhood Justice Centre programs, as well as other venues of the Magistrates Court of Victoria. The protocol specifies the roles and responsibilities of co-ordinators and workers from the participating agencies, collaborative work protocols, data collection and reporting arrangement and grievance procedures. The primary relationship between CISP and the JHSP agencies is through the Housing Information & Referral workers. A HomeGround worker is present at Melbourne four days a week, and a SASHS worker is on-site at Sunshine Court, also for four days each week. Quantum Social Services is mainly used on a referral basis at the Latrobe Valley venue. The JHSP funding and service agreement sets a target of a total of 55 Transitional Housing Properties (THM) available to justice clients, including 18 properties available through the Melbourne and Sunshine venues and two properties available to clients at the Latrobe Valley venue. Not all properties covered by the JHSP are available at any 16 Note that SASHS is the referring agency at Sunshine but support services are provided by HomeGround. Melbourne Criminological Research and Evaluation 92

time, as a result of delays in acquisition, turnover in properties, or repairs or refurbishment of properties. In June 2008, a total of 38 properties were available to the JHSP, including ten to CISP at Melbourne and Sunshine and two to CISP Latrobe Valley and by June 2009 this had increased to 45 properties. In addition to placing clients in properties where CISP has nomination rights, the JHSP agencies can place clients in temporary accommodation, assist in preparing applications for public housing, and provide case management and support on housing-related issues. The preparation of housing applications involves compiling a detailed account of a client s case history, and this in turn may require access to information from the client s file to support the application. Where clients are referred through CISP but continue to require housing support after they exit this program, case management responsibility may be transferred to the housing support agency. 7.2 Housing service needs of CISP clients Improving access to stable housing is an important element in the CISP service delivery model. While poor or unstable housing does not directly cause crime, it is a critical factor in achieving other program outcomes. Defendants who are homeless or in emergency accommodation have a higher risk of non-compliance with bail, are more likely to be remanded in custody, less likely to attend treatment programs and have poorer outcomes from treatment and support interventions (Baldry, McDonnell, Maplestone, & Peeters, 2003). The housing services component of CISP is specifically targeted at improving outcomes for clients who are homeless or in emergency accommodation, and forms part of the Justice Housing Support Program. Six percent of CISP engaged clients were assessed as homeless in 2007 and 2008, a further nine percent were in emergency or transitory accommodation, and 12% were in boarding house or other short-term rental accommodation (Table 7.1). In total, over a quarter of all engaged clients were in unstable accommodation when they were assessed for entry to CISP. Overall, housing issues are most prevalent in clients at Melbourne, with little difference between Sunshine and Latrobe Valley. Melbourne Criminological Research and Evaluation 93

Accommodation stability at entry Owned or buying Long-term public or private Case location Latrobe Valley Melbourne Sunshine Total N 25 107 122 254 % 8.6% 8.4% 15.4% 10.8% N 171 540 394 1105 rental % 58.8% 42.4% 49.7% 46.9% Living with family N 29 147 90 266 % 10.0% 11.5% 11.3% 11.3% N 31 195 66 292 Boarding house or shortterm rental % 10.7% 15.3% 8.3% 12.4% Emergency or transitory Homeless Supported accommodation Unknown Total N 18 147 54 219 % 6.2% 11.5% 6.8% 9.3% N 15 76 45 136 % 5.2% 6.0% 5.7% 5.8% N 0 14 1 15 %.0% 1.1%.1%.6% N 2 47 21 70 %.7% 3.7% 2.6% 3.0% N 291 1273 793 2357 % 100.0% 100.0% 100.0% 100.0% Table 7.1 Accommodation stability of CISP clients at program entry 7.3 Service output measures The JHSP reports on a quarterly and annual basis against a number of performance criteria, and these reports provide summary data for each of the CISP venues. The following service measures are drawn from the 2007/08 report which covers the period February 2007 to February 2008, although some activity measures up to March or May 2008 are reported. The HIR service at the Melbourne Magistrates Court saw 293 clients between February 2007 to March 2008 for a total of 822 contacts (an average of 2.8 contacts per client) and has submitted a total of 120 housing applications The HIR service at Sunshine Court recorded a total of 511 contacts between August 2007 and May 2008. These resulted in 26 completed Segment 3 applications, with 12 submitted for approval and the remainder on hold awaiting further documentation or contact from the client. Twelve Segment 1 applications were commenced with 2 submitted for endorsement. Melbourne Criminological Research and Evaluation 94

At the Latrobe Valley venue, QSS reported five clients successfully housed out of a total of six referrals. In total, three Clients have exited from transitional properties and both properties were occupied as of February 2008. Referrals made by CISP staff to the JHSP services are also recorded on the CMS, as are other housing-related referrals (mainly referrals to emergency or temporary accommodation). In the 2008 year there were 324 cases where a CISP HIR referral was recorded, of which 215 were made at Melbourne, 102 at Sunshine and only seven at Latrobe Valley. In addition, there were 79 referrals to other housing services, of which 37 were from Melbourne, 20 from Sunshine and 22 from Latrobe valley. Direct comparison between these two sources is difficult as they are based on different time periods and may use different ways of counting service activity. If the referrals counts from the CMS are expressed as a proportion of the number of engaged clients at each venue in 2008, it can be seen that the likelihood that clients will be referred to the CISP HIR is greatest at Melbourne but that relatively few Latrobe Valley clients receive this form of referral, and this appears to be consistent with the general patterns of activity reported by the JHSP. An important limiting factor on access to THM properties by CISP clients is that the average period of tenancy for clients in the JHSP is around 35 to 40 weeks. As a result, vacancy rates are low and the ratio of referrals to property vacancies in the JHSP is around four times. JHSP reporting shows that 28 CISP clients were housed in the period to December 2008, of whom 19 were allocated to properties in 2008. The average length of tenancy so far is around 60 weeks, and since only three CISP clients have exited JHSP properties to date it is inevitable that the actual length of stay will be much greater. This problem is particularly serious where the number of properties is small, as in the Latrobe Valley. Two CISP clients who were referred in mid-2008 and exited CISP in December 2008 were still in the properties 50 weeks after entry. As a result, there is effectively no access to JHSP properties for clients at this venue. It also seems likely that relatively few new CISP clients will be able to enter the JHSP in 2009. Melbourne Criminological Research and Evaluation 95

Proportion of engaged clients referred to housing services 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% Clients referred to CISP HIR Clients referred to other housing 10.0% 5.0% 0.0% Melbourne Sunshine Latrobe Valley Figure 7.1 Proportion of engaged clients referred to housing services CISP clients accommodation status is recorded at their initial assessment, and again when they exit the program. Table 7.2 shows accommodation status at entry and exit for those clients who exited CISP after having successfully completed the program. The aggregate number of clients who were homeless fell from 42 to 12 over this period, although this is slightly misleading as only four of the 42 who were recorded as homeless at program entry were still homeless at exit, while another eight who were not homeless when they entered the program had become homeless by the time they exited. Of those who had been homeless, around a third were in emergency or transitory accommodation at exit, and 20% were each in long-term rental, supported accommodation and boarding house or short-term rental. However, there was no change between entry and exit in the aggregate number recorded as in emergency or transitory accommodation, and only a small reduction in those in boarding house or short-term accommodation. The number who were in the owned or buying category fell by a third - most of these clients moved to long-term rental or residing with family status. Again, it should be noted that these are aggregate changes and there was substantial movement within these groups. Melbourne Criminological Research and Evaluation 96

Status at entry Status at exit Accommodation status Frequency Percent Frequency Percent Owned or buying 149 12.4 99 8.2 Long-term rental 607 50.5 667 55.5 Residing with family 119 9.9 153 12.7 Supported accommodation 8.7 23 1.9 Boarding house or shortterm 144 12.0 126 10.5 rental Emergency or transitory 100 8.3 100 8.3 Homeless 46 3.8 12 1.0 Unknown 27 2.2 21 1.7 Total 1201 100.0 1201 100.0 Table 7.2 Accommodation status at CISP entry and exit 7.4 Relationship between CISP and housing agencies While both the JHSP and CISP recognize the importance of dealing with housing issues, each program has different priorities and each works under different constraints. The priority for the JHSP is to tackle the problem of homelessness in the population of justice clients. The screening process for JHSP services is primarily concerned with risks and needs relating to homelessness 17 and the goal of their interventions is to provide long-term stable accommodation. It is accepted that this is likely to require long-term engagement with clients, partly because the application and placement process for public housing takes a long time, and that this engagement should continue until a satisfactory outcome is achieved. In contrast, CISP is concerned with accommodation as a means to an end achieving better bail outcomes and allowing clients to engage in other forms of treatment. The process of CISP engagement is driven by the timelines imposed by the court, and the process of engagement is time-limited. When a client exits CISP the case manager s involvement ceases whether or not the person s housing issues have been resolved. 17 See Appendix A in JHSP protocol Melbourne Criminological Research and Evaluation 97

These distinctly different views are reflected in the feedback about the relationships between case workers on each program. A common theme in our interviews was that workers reported that the other party didn t adequately understand the capacities and limitations associated with their role. For example, one housing worker reported that a problem with working with CISP was that referrals are made based on the clients need to receive bail rather than their actual need for housing. Conversely, there were concerns expressed by CISP workers that the JHSP properties were permanently occupied, that housing workers were mainly concerned with clients long-term needs at the expense of short-term issues that were more directly relevant to case management in CISP, and that as a result the program was of relatively minor significance for CISP. The time pressure of work was also a source of dissatisfaction, with some housing workers arguing that up to two weeks were required to make a full assessment of CISP clients housing needs, while CISP workers wanted a response within days. Another source of dissatisfaction for housing workers was lack of access to client assessment and treatment information. Each of these issues reflects the differences in program goals and processes, and the frustration experienced by workers attempting to solve intractable problems in an environment where resources are strictly limited. A related theme in the interviews with workers was that the two services don t always work together as effectively as possible. It was claimed that housing workers aren t invited to take part in CISP team meetings. One proposal was that there should be monthly case reviews involving all support workers and the client, to provide an opportunity for the team involved with the client (including the housing worker) to discuss the issues affecting the client. 7.5 Access to housing services for CISP clients The fundamental problem about access to housing services for CISP clients is the lack of available accommodation. While the JHSP provides a pool of medium- and long-term housing, the number of properties is limited and vacancy rates are low. Access to these properties represents an undeniable improvement for those who are placed in them, but the number who benefit is small in comparison with the total demand. Nevertheless, there is evidence that the JHSP is able to reduce the rate of homelessness in CISP clients, at least where clients are able to successfully complete the program. Access to Melbourne Criminological Research and Evaluation 98

temporary and private rental properties is also limited. Interviewees noted that with the increasing pressure on the rental market, even previously accessible options such as boarding houses have become difficult to access, especially for justice clients, and that waiting times of six weeks are common for emergency housing. The long-term solution for many CISP clients is public housing, but the application process is detailed and there are long waiting times. The result is that for most CISP clients any change in their housing situation is likely to take longer than the period of their involvement with the program. Melbourne Criminological Research and Evaluation 99

Chapter 8 Magistrate perspectives on CISP and Therapeutic Jurisprudence 8.1 Magistrates use of CISP The assessment and case management services provided by CISP can only be effective if the program is used appropriately by Magistrates. This chapter examines Magistrate perspectives on the program, and asks how Magistrates use the services offered by CISP, how this bears on their daily practice, whether they are satisfied with the quality and timelines of CISP services, and how they see CISP developing in the future. This analysis is based on interviews and focus groups conducted with Magistrates at each of the CISP venues, together with feedback from senior Magistrates in the Magistrates Court of Victoria. In the last year, CISP services have been extended to the County Court and to persons appearing in relation to family violence matters, and this chapter also examines the implications of these developments in the CISP model. There is considerable variation in the extent to which individual Magistrates use CISP. Some Magistrates use the program on a regular basis, others use it only for cases where there is a pressing need for support, and others use it rarely or never. Variation in frequency of use is present at each of the three CISP venues, although it is less evident at Sunshine where the majority of Magistrates are frequent users of CISP. The Magistrates who agreed to be interviewed for this evaluation were mainly the more frequent users of the program. Magistrates reported using the services provided by CISP in four ways: i Providing a comprehensive and independent assessment of the problems and issues associated with a defendant. Magistrates noted that they frequently deal with cases where claims are made about defendants problems. These may be made by the defendant s legal representative, the police or by other parties, and Magistrates must take these claims into account in their decision-making. Magistrates noted that while they deal with these problems on a daily basis, there are often complex clinical, social or personal issues involved that they do not Melbourne Criminological Research and Evaluation 100

have the expertise or time to deal with directly. CISP case managers are recognized by Magistrates as officers of the court who can be relied on to provide independent advice. In this way, CISP provides a means whereby defendants can be formally assessed and the status of their issues or problems can be quickly and independently validated. Before CISP the concept of dealing with people with special needs was hit and miss ii Selecting and organizing an appropriate therapeutic response Where a Magistrate identifies that a therapeutic response is appropriate, there remains the problem of determining what this response should be, and identifying an appropriate service agency. CISP allows Magistrates to know what treatment resources are available, assists in choosing the most appropriate form of intervention, and generally improves the court s timeliness and efficiency in organizing a therapeutic response. I have no idea about the scores of institutions, programs and resources that are available. The fact that the CISP workers know about that is just fantastic. We wouldn t have a clue about those things, where to send someone or how to get assistance. iii Supervising defendant s through the course of their bail period Magistrates saw the supervision and case management process as helping to keep defendants on track and providing them with regular feedback on the progress of defendants. A key element in this was the idea that the case manager was part of the chain of accountability between the defendant and the court. The CISP workers are responsible to me, and if they (defendants) are in breach, they know that I will be told and told quickly. iv Preparing defendants for a CCS order Involvement in CISP helps to prepare clients for a Community Corrections Order by reducing the chaos in their lives and giving them the experience of being a client in a highly structured and accountable program. Successful completion of CISP demonstrates that a defendant is likely to be able to handle the requirements of a CCS order. Melbourne Criminological Research and Evaluation 101

The way that Magistrates view CISP is largely a function of the extent to which their practice incorporates elements of therapeutic jurisprudence. In general, Magistrates saw CISP as a vital element in making therapeutic jurisprudence work effectively. A comment that was frequently made was that CISP makes it possible for me to do my job properly. Several Magistrates referred to cases that would previously have been regarded as hopeless or a waste of the court s time but who were now able to be successfully placed on bail in the expectation that this would result in real improvement in their problems. I ve seen great results from it, what seemed like hopeless cases but seeing them over the four months, there are huge transformations that take place. 8.2 Relationship to practice When CISP commenced it was evident that some magistrates were uncertain about whether it was a distinct program or simply a rebadging of the CREDIT/Bail Support program model, and whether CISP staff were only concerned with the assessment and referral of defendants, or whether they were clinicians in their own right. These issues appear to have been resolved the Magistrates interviewed were all clear about the features of the CISP service delivery model and the role of staff, and some noted that they found it significantly better than the CBS model. However note that CISP staff members continue to express concerns about the recognition of their role and functions (see additional comments in Chapter 5). Magistrates were asked how CISP related to their daily practice in court. There were two distinct practice approaches described, with their use depending on the circumstances of the case. The first can be described as an intensive supervision approach. Where a defendant was judged to be high risk, the Magistrate would be directly and regularly involved in their supervision while on bail. Frequently the Magistrate would mark the case as part-heard, and the defendant would be required to report back at frequent intervals. An important part of this approach is communicating to the defendant that the Magistrate is taking a direct interest in their case, and that he or she is personally accountable to the Magistrate for their performance on CISP. Melbourne Criminological Research and Evaluation 102

It s important to know who is really making an effort, and I find having the CISP worker presenting the report in court and being able to ask them questions is quite useful. And the defendant s in court and they hear that they hear that we re taking an interest in the case, and it makes them aware that it really is important, that the Magistrate really does take this seriously. The feedback from clients is that the feedback is important, the penny drops for them. In the second approach the emphasis is on CISP as a referral and support service. This approach is appropriate where defendants don t pose a serious risk but require support (eg. where the defendant is homeless)and possibly treatment (eg. where the defendant has a mental disorder) in order to be eligible for bail. In such cases there is less need for direct personal involvement by the Magistrate, or for frequent reporting by the defendant. For the others, I think they should be brought back for supervision but I won t mark myself as part-heard. Anyone can review the file and deal with that case. In both of these practice approaches there is an acknowledgment that not all defendants will comply with the requirements of the court. Most Magistrates see that involvement in CISP can place significant demands on defendants and that not all defendants will be willing or able to comply with them. For some defendants (especially those with a history of failure to comply with court orders) CISP is a waste of time and they should be screened out at the beginning. Realistic assessment and judicial monitoring are critical to ensuring that appropriate defendants are placed on CISP, and that they then act in accordance with the program requirements. Sometimes it is difficult for people, in the first month, to appreciate what s required. I invariably give them one go, I ll read the riot act to them but if that doesn t wake them up then I have no hesitation in taking them off. Magistrates were asked to comment specifically on the decline in rates of judicial monitoring (see 2.6 above). Those who responded to this enquiry expressed strong support for the principle of judicial monitoring. It was noted that judicial monitoring was central to therapeutic jurisprudence and should generally be encouraged as a constructive adjunct to all CISP referrals. A general issue that was acknowledged by Melbourne Criminological Research and Evaluation 103

Magistrates was the impact of work pressure on their capacity to deal with complex cases effectively. The process of assessment and repeated case review means that defendants on CISP require more court time than those where the court is only concerned with whether bail should be granted or refused. Effective screening of cases to ensure that only those who genuinely warranted CISP interventions were placed on the program was seen as a key requirement in limiting the work-flow impact of the program. Two other possible reasons for the fall in rates of judicial monitoring were nominated: Movement and certainty of Magistrates scheduling. Judicial monitoring is more difficult if a Magistrate is not certain that she/he will be around to deal with a matter marked part-heard. It was noted that this is particularly an issue at Melbourne where there are relieving Magistrates coming in here for a day or so, or regulars who themselves do a bit of relieving and therefore they cannot guarantee they will be here on the next occasion ; The availability of police prosecutors. It was suggested that the former arrangement of part-heard Fridays at Melbourne allowed Magistrates to schedule part-heard matters in the knowledge we would have a dedicated Prosecutor in that Court for as long as we needed them. Magistrates were also asked about the impact of CISP on their sentencing decisions. Generally, involvement in CISP was viewed as relevant to sentencing but not a determinant of the sentencing outcome. It was noted that with CISP there was no deal being, but that in general good performance on CISP would be taken into account when sentencing. However, where cases involved serious offences, the court would be unable to make any significant reduction in sentence severity. Conversely, Magistrates noted that failure to comply with CISP didn t mean any increase in sentence severity, although it might mean that some rehabilitative sentencing options might not be considered as suitable. Several Magistrates made the point that successful involvement in CISP demonstrated that the defendant was committed to his or rehabilitation, provided evidence of an established and effective treatment relationship, and allowed decisions about rehabilitative sentencing to be made with greater confidence. If you re dealing with really serious offences (drug trafficking, sex offences against kids) you can t give people any major benefit for being on the program. But overall I tend to treat success stories Melbourne Criminological Research and Evaluation 104

from the CISP with a reduction in the level, the type of penalty, changing a prison term to a suspended term or a community based order. There is no deal being offered - that if they do CISP they ll get a certain outcome, that they won t go to jail where they would otherwise. But if they do CISP, a good performance on CISP is a relevant matter to consider in sentencing. Successful participation in CISP may also be a consideration in suspending a sentence of imprisonment. Section 27 of the Sentencing Act (1991) specifies that a court must not suspend a term of imprisonment unless it is satisfied that such an order is appropriate because of the existence of exceptional circumstances. Similarly Section 31 provides that where a person breaches a suspended sentence of imprisonment, the court should order that the original term of imprisonment be restored unless exceptional circumstances mean that it would be unjust to do so. In each case successful participation in CISP may be held to qualify as exceptional circumstances. An important constraint on the use of CISP in cases involving serious indictable offences was identified by Magistrates. Where a defendant is charged with this kind of offence and then bailed, there is limited capacity for a Magistrate to exercise judicial supervision with regular reviews of the defendant. In conventional summary or indictable matters dealt with in the Magistrates Court, bail can be managed such that the defendant can be required to report for review at intervals. In the case of serious indictable matters the defendant can be placed on CISP but the next appearance date may be months in the future. Arguable, such cases warrant greater levels of judicial supervision but no mechanism for this exists in the current bail arrangements. 8.3 Satisfaction with CISP services In general, Magistrates expressed a high level of satisfaction with the quality and timeliness of CISP services. Assessment reports were considered to be of a high quality and were generally provided within a reasonable time. Similarly, case management and monitoring were reported to be of a high standard. CISP staff members were viewed as being strongly commitment to helping their clients, knowledgeable about the clinical and support issues involved in case management, and generally as an important source of Melbourne Criminological Research and Evaluation 105

information and advice. Magistrates reported that they trusted the judgment of most CISP staff (especially experienced team leaders) and welcomed their input into the judicial monitoring process. If I get some material put to me (by the lawyers) about the effect of certain therapies, I use them as a resource, I can bounce ideas off them, pick their brains. The two most commonly cited sources of dissatisfaction were limitations on case loads, and naïve or unrealistic assessment of clients. These two problems were present at all three venues but were more apparent at some times than others. Inappropriate gate keeping by CISP staff resulted in unsuitable cases filling up the available program places with the result that Magistrates were advised that caseloads for the month were fully occupied and no new cases could be accepted. This in turn was seen as being the consequence of inexperienced staff providing inadequate reports, accepting clients onto the program who were unsuitable, or failing to terminate those who were not showing an acceptable level of commitment. A number of Magistrates commented on the high turnover in CISP staff and the impact that this had on the quality of service, both as a result of vacancies in case manager positions, and the loss of experienced case managers. Before we had (team leader), every defence lawyer would refer everybody to CISP, and there was no gatekeeper, and CISP would take them all. But (team leader) was able to identify the ones who were not suitable for the program. We were getting emails to say that caseloads for the month were full, sometimes by the middle of the month While CISP was viewed by Magistrates as greatly enhancing their capacity to identify and organize treatment and support services, there was also recognition that sometimes the necessary program or service resources were not available. The two areas most commonly nominated were places in residential drug detoxification or treatment programs, and assessments for Acquired Brain Injury. Magistrates reported that there might be delays of weeks before a place in a drug program became available, and that this was a significant impediment to dealing effectively with such cases. It was noted that the bail process was a window of opportunity where defendants with drug problems were highly motivated to deal with their problems, but if no effective response was made Melbourne Criminological Research and Evaluation 106

available their continuing drug use made them unsuitable for CISP, and if remanded they were unlikely to receive any effective treatment. In the case of ABI assessments, it was acknowledged that the rate of referral to arbias had been much greater than anticipated, with the result that arbias were drowning in demand. At the beginning we were told we would get 5 ABI assessments a month, but within 6 months there were 40 people waiting for assessment, so if you work that out, half of them weren t going to get to the head of the queue within the 4 months. These comments were made by magistrates after the delays in ABI assessments had been addressed. This indicates that regular communication with magistrates about the availability of specialist assessment and treatment services is an important part of maintaining confidence in CISP. Melbourne Criminological Research and Evaluation 107

8.4 Future directions for CISP Magistrates were asked how they would like to see CISP develop in the future. The most commonly nominated enhancement to the program was its extension to other court venues, including regional courts and other specialist courts. Several referred to the variations in practice that were apparent when they were working at court venues where CISP was not available, and the inherent unfairness of this to defendants ( postcode justice ). Magistrates whose work took them to regional courts commented on the limitations this placed on their capacity to respond effectively to defendants. Similarly, it was noted that the availability of a CISP-style program would greatly enhance the effectiveness of the Koori Court, especially in regional venues where it was perceived that the local Koori-based service organizations were under a great deal of strain and close to being burnt out. It was noted that extension to regional venues would involve some modification to the existing program model such as telephone or video-link access, or a visiting case management team. As CISP has demonstrated that it is capable of managing defendants charged with more serious offences, it raises the question of the extension of the program to the County Court. The primary services offered by CISP will always remain in the domain of the Magistrates Court, but as more serious offenders make their way into the County Court either on bail or for sentencing, the case management information from CISP becomes relevant. It is already the case that some County Court judges are aware of this and take these matters into account in their decisions. However, for this to be fully effective, County Court judges need to have access to information about the program and the services that it offers. The future of CISP is intimately bound up in the future of therapeutic jurisprudence in the courts. While there is a general acknowledgement that therapeutic interventions are neither suitable nor appropriate for all defendants, the feedback from Magistrates also makes it clear that they see the availability of the services offered by CISP as a necessary investment to make therapeutic jurisprudence work. Magistrates view CISP as more than just a service that allows defendants to be safely managed while on bail. CISP is viewed Melbourne Criminological Research and Evaluation 108

as a key element in the process of therapeutic jurisprudence that generates tangible outcomes for defendants in the form of better understanding of the issues that lie behind their offending, and effective responses to those problems. 8.5 Recommendations It is clear from Magistrates comments that CISP is integral to the process of therapeutic jurisprudence. However, this also means that for CISP to work effectively, therapeutic jurisprudence processes must also be supported. That is, CISP should be seen not just as an intervention intended to produce better outcomes at the bail and pre-conviction stage, but as an integral element in therapeutic jurisprudence processes that commence at the bail stage and continue throughout the offender s involvement in the court process and through into supervision under court order. This analysis also suggests that for CISP to work effectively, two areas of program policy and procedure should be examined with a view to reform. These are: Provision of more explicit referral criteria. Both Magistrates and case managers (see Chapter 5) are aware that some referrals are made that are not appropriate for the program. This can be addressed by providing more explicit referral criteria that take into account factors such as the defendants previous bail history, any breaches of suspended sentences or Community Corrections orders, and other factors relevant to the court s bail decision. One possibility would be to establish three referral outcomes: unsuitable and rejected referrals (as at present), referrals accepted for assessment, and referrals where a Magistrate s approval should be sought before accepting the defendant for assessment Legislative or procedural amendments to allow defendants charged with serious indictable offences to be subject to judicial monitoring. This might require establishing reporting dates intermediate between the Filing Hearing and Committal Mention appearance that would apply to defendants placed on CISP. Melbourne Criminological Research and Evaluation 109

Chapter 9 outcomes Reoffending and compliance 9.1 Outcome targets for CISP The primary goals of CISP are concerned with improving treatment, court, sentence and re-offending outcomes for program participants. These goals are all linked to one another: effective treatment of the causes of offending is intended to bring about better health and well-being outcomes and reduce bail breach rates. These outcomes lead in turn to an increased likelihood of rehabilitative sentencing and improved order compliance. The ultimate outcome for CISP is reduced rates of post-release re-offending. Treatment, health and well-being outcomes were reported in earlier chapters of this report and this chapter examines data on the three justice process outcomes of bail compliance, order compliance and re-offending. 9.2 Outcome assessment methodology CISP clients are selected in two ways. Firstly, they must apply or be referred to the program and this referral must be based on some identified problem or disadvantage. Secondly, the decision to accept the referral and engage the person as a CISP client is based on an assessment of the person and a judgment about the suitability of the person and the likely outcome of the case. Defendants with no or only mild problems either don t apply, are screened out during the assessment process, or are referred to Level 1 of the program. Defendants who are judged likely to be remanded in prison because of the seriousness of the offence or their criminal history are also likely to be excluded. Thus, in order to make meaningful comparisons of outcomes between CISP clients and other defendants, it is necessary to look at outcomes for defendants who present at court venues where CISP is not available and who share the same general demographic and criminal profile as CISP clients. Melbourne Criminological Research and Evaluation 110

The general methodology for this analysis involved comparing the post-court records of 200 persons who had completed CISP in 2007 with a sample of 200 persons sentenced in other Magistrates Court venues in the same period. There were 100 CISP clients selected from Melbourne, 51 from Sunshine and 49 from Latrobe Valley. CISP clients who had exited without completing the program as a result of re-offending, noncompliance or who were remanded were excluded from the CISP sample. Each CISP client was matched with a person from a non-cisp venue. Unfortunately, court records provide little information about defendants other than their age, sex and court history. The primary matching criteria used were age, gender, major charge and number of charges. Records of any subsequent court records and outcomes occurring between the primary episode date in 2007 (CISP exit for the program group or sentence date for the comparison group) and 30 November 2008 were collected. Re-offending events included proven offences and unproven charges yet to be determined, as well as any records of ICO and CBO breach matters. Minor traffic and public transport offences were not counted as a re-offending event. 9.3 Bail compliance For a pre-trial program like CISP, improving bail compliance rates is obviously a critical outcome. However, bail data is not currently able to be extracted in a manner that allows reliable and consistent analysis of bail outcomes for CISP clients and Victorian offenders generally. There are several factors that complicate these calculations. Bail may be granted by police and not recorded on any court data systems. Calculating compliance rates with bail orders is difficult, as any individual may have multiple bail orders in the course of a single case, and any bail order may give rise to multiple charges of fail to appear. The absence of bail data constitutes a significant barrier to the evaluation of court programs in Victoria. 9.4 Order compliance Advice from Magistrates indicates that successful participation in CISP is viewed as an important indicator of suitability for a CCS order. In addition, the treatment and support referrals provided through CISP have the potential to prepare offenders for longer-term Melbourne Criminological Research and Evaluation 111

case management under a CCS order. Community corrections orders are a common outcome for CISP clients: for those clients whose court outcomes are known at the time they complete CISP, around one-third receive a community corrections order (ICO or CBO) (see Chapter 2). This compares with less than ten per cent of defendants in all Magistrates Court cases. Thus, the impact of CISP on successful completion of CCS orders is a key outcome for this program. Ideally, the order breach rate for all CISP clients placed on a CCS order would be compared with the breach rate for other CCS orders. Unfortunately, data on CCS order completions for CISP clients was not available for this evaluation. As an alternative, breach rates were compared for the sample of CISP orders and comparison sample as described above. However, this must be seen as a less desirable basis of comparison than a comparison based on all CISP clients. In the sample of 200 CISP cases there were 70 who had received a CCS order (35%), and in the control group 84 defendants (42%) received a CCS order. This suggests that the members of both groups were on average more serious offenders than lower court defendants overall, which is to be expected given the selection process for the two groups. It is unclear whether the variation between the two groups in CCS order rates is indicative of a difference in overall case severity. The overall successful completion rates for orders up to June 2009 were 49% for the CISP sample, and 45% for the control sample. The successful completion rate for CCS orders state-wide in 2007/08 was 58%. While the proven breach rate for CISP clients was 3% lower than for the control group, a chi-square test showed that there was no statistically significant difference in CCS order outcomes between the two groups (Likelihood ratio value 0.17, df=2, p=0.9). Melbourne Criminological Research and Evaluation 112

Group Control CISP Total Breach category No breach Proven Breach Total breach charge unfinalized N 38 41 5 84 % 45.2% 48.8% 6.0% 100.0% N 34 32 4 64 % 48.6% 45.7% 5.7% 100.0% N 72 73 9 154 % 46.8% 47.4% 5.8% 100.0% Table 9.1 CCS order outcomes for CISP and control groups 9.5 Re-offending rates There are three measures of re-offending that are bear on the assessment of program outcomes: whether a person re-offends at any time, the period that elapses until they do re-offend, and the number of offences they commit in a given period. Proportion of recidivists For the CISP group, around 50% cent were classed as recidivists, of whom 40% had proven charges against them, and a further 10% had charges that had not been finalized. In the comparison group, 64% were classed as recidivists, with 50% having proven charges recorded and a further 13% having unfinalized charges. A chi-square test showed that there was a statistically significant difference in the proportions of recidivists between the two groups (Likelihood ratio = 6.87, df=1, p<0.05). Melbourne Criminological Research and Evaluation 113

Outcome Proven charges CISP Comparison group N 79 99 % 39.5% 49.5% Unproven N 20 26 charges % 10.0% 13.0% No further N 101 75 charges % 50.5% 37.5% Total N 200 200 % 100.0% 100.0% Table 9.2 Proportion of recidivists in CISP and comparison groups Time to re-offence Re-offending was measured from the time that the members of the CISP exited the program and the members of the control group were sentenced until 30 June 2009. For each CISP re-offender, the number of days between their program exit and their first proven subsequent offence was calculated. For the control group, the number of days between their matching case sentence date and their first proven subsequent offence was calculated. Any time spent in custody was deducted from the time at risk. Where a prison term incorporating a maximum and non-parole period was given, only the nonparole period was deducted. The follow-up period for non-re-offenders varied from just under 400 days to over 900 days, with a mean of 670 days for the CISP group and 720 days for the control group. In order to examine whether there is a difference between the time distribution of reoffending between the two groups, the Kaplan-Meier Survival Analysis procedure was used. This procedure plots the survival function for event history data (in this case, the time to first re-offence) and calculates whether there is a significant difference in the survival functions for the two groups. For this analysis only proven offences were classified as recidivist events. Melbourne Criminological Research and Evaluation 114

Figure 9.1 Kaplan-Meier Cumulative Survival Function for CISP and Control groups The plot of the survival function 18 (Figure 9.1) shows that for both groups the rate of recidivism increases fastest in the period immediately after the commencement of time at risk (i.e. exit from CISP or sentence date). Initially the two survival curves overlap but after about 180 days there is a divergence between the curves that is maintained thereafter. At 100 days, around 20% of both groups have recidivated, and at 200 days around 30% of the CISP group and 32.5% of the control group have recidivated. By 400 days the degree of divergence is six percent (37% of the CISP group and 43% of the control group) and by 600 days it is eight percent (40% of the CISP group and 48% of the control group. The extent of divergence reaches its maximum of ten percent by around 700 days. 18 The 1-survival function is plotted as this is equivalent to a recidivism function and hence more readily interpreted. Melbourne Criminological Research and Evaluation 115

The Mantel-Cox log-rank test compares whether these two survival curves have equal distributions. The test has a chi-square distribution and yields a value of 2.36 with one degree of freedom, and a probability of 0.125. This is not a significant difference but it should be noted that the group sizes of 200 are the minimum recommended by the Home office to detect variations in recidivism between a program and control group. Melbourne Criminological Research and Evaluation 116

REFERENCES Adams, K., Sandy, L., Smith, L., & Triglone, B. (2008). Drug Use Monitoring in Australia: 2007 annual report on drug use amongst police detainees. Research and Public Policy Series No. 93. Canberra: Australian Institute of Criminology. Australian Bureau of Statistics. (2008). Prisoners in Australia. Cat. No. 4517.0. Canberra: Australian Bureau of Statistics. Australian Institute of Health and Welfare. (1999). The definition, incidence and prevalence of Acquired Brain Injury in Australia. AIHW Cat. No. DIS 15 Canberra: Australian Institute of Health and Welfare. Melbourne Criminological Research and Evaluation 117

APPENDIX 1 REGRESSION MODELLING OF PROGRAM COMPLETION Around four in ten CISP clients fail to complete the program. If the likelihood that client was going to fail to complete the program could be accurately predicted, program targeting could be improved. There are a number of factors that show a statistically significant relationship to failure to complete, and this analysis examines a range of bivariate and multivariate models of program completion. Defining program completion Program completion can be defined in simple terms of completed/not completed or by taking into account why a client fails to complete. The main reasons for failing to complete are that the client fails to attend the program, is adjourned to another court level or venue or is remanded in custody. Non-completion as a result of being adjourned or remanded can arise from a breach of bail conditions, or further offences committed while on CISP, or additional offences which may predate the CISP episode. These outcomes cannot be accurately distinguished from one another from CISP records. The PSCMS system distinguishes between clients who fail to attend and those who no longer wish to attend, however these can be regarded as equivalent fail to attend outcomes. Thus, in examining program completions it is appropriate to distinguish between completion as a general category, and completion and non-completion as a result of failing to attend (excluding non-completion as a result of adjournment or remand). In the following analyses these are used as the two dependent variables and have the following value distributions and group sizes based on all completing clients to March 2009: Completion status Completion Non-completion Non-attendance Number of cases 1201 847 431 Bivariate relationships The first stage in the analysis was to examine variables that were likely to show a bivariate relationship with the outcomes of interest. These bivariate relationships help to Melbourne Criminological Research and Evaluation 118

select variables for inclusion in the multi-variate analysis and show whether any re-coding is required to eliminate categories (eg of unknown value) or re-order or combine categories in relation to their relationship with the dependent variables. The set of variables selected included the primary demographic characteristics of clients (age, sex, indigenous status), key program participation attributes (court venue, referral source, custody status at assessment, judicial monitoring and whether CISP was a bail condition) and client characteristics likely to be related to future risk (drug and alcohol status, ABI and mental health status, accommodation and financial status). All variables except age are categorical, and the bivariate analyses were done using Chisquare tests. To facilitate comparison with the other analyses, age was converted to a categorical variable (young, middle, older). For all variables, unknown values were set to missing. The tables below summarize the results of these bivariate analyses. The independent variables referral source, judicial supervision, current alcohol use, ABI status, and mental health issues did not show a statistically significant relationship with either of the dependent variables and were eliminated from further analysis. This left nine variables with statistically significant relationships with completion that were included in the multivariate analysis. Dependent Independent variable Chi-square P value variable (L.R.) value Completed or not completed Court location 33.6 0.000 Referral source 10.94 N.S. Custody status at assessment 111.16 0.000 CISP as a bail condition? 16.77 0.000 Gender 1.36 N.S. Indigenous status 10.93 0.000 Judicial monitoring 2.36 N.S. Age group (young, middle, older) 12.92 0.002 Current drug use 25.36 0.000 Current alcohol use level (use, abuse, 3.18 N.S. dependence) ABI indicated? 0.43 N.S. Mental health issues indicated? 0.81 N.S. Accommodation stability at entry 45.09 0.000 Level of financial support (none, some, 7.41 0.025 significant) Bivariate relationships for completion / not complete outcomes Melbourne Criminological Research and Evaluation 119

Dependent Independent variable Chi-square P value variable (L.R.) value Completed or not attend Court location 19.73 0.000 Referral source 6.58 N.S. Custody status at assessment 30.64 0.000 CISP as a bail condition? 9.71 0.002 Gender 6.32 0.01 Indigenous status 16.09 0.000 Judicial monitoring 0.31 N.S. Age group (young, middle, older) 18.32 0.000 Current drug use 19.05 0.000 Current alcohol use level (use, abuse, 3.66 N.S. dependence) ABI indicated? 0.37 N.S. Mental health issues indicated? 0.39 N.S. Accommodation stability at entry 62.6 0.000 Level of financial support (none, some, 16.90 0.000 significant) Bivariate relationships for completion / not attend outcomes Multivariate analysis The two dependent variables are categorical (complete/not complete and complete/not attend) and a logistic regression method was used for the multivariate analysis. This was done in two steps. Firstly, all variables were entered into the logistic regression (ENTER method), and then the same variables were entered using a FORWARD STEPWISE method, with the Likelihood ratio as the entry criterion. Court location was entered as a categorical variable. Only one of the nine variables (sex of the offender) was excluded from the forward stepwise model for completion/non-completion, and two were excluded from the model for completed/not attended (sex and court location). It seems likely that the exclusion of cases with a completion outcome of remanded or adjourned also removed much of the variance associated with the court location variable. With this exception, the final models produced by the two methods are more or less identical. These models are shown in the two summary tables below. Melbourne Criminological Research and Evaluation 120

Variable B S.E. Wald df Sig. Exp(B) Court location 13.058 2.001 Court location (Latrobe V vs other) -.683.189 13.057 1.000.505 Court location (Melbourne vs other) -.146.114 1.655 1.198.864 CISP as bail condition.768.151 26.038 1.000 2.156 Age group -.194.071 7.432 1.006.823 Current drug use -.435.126 11.948 1.001.648 Accommodation stability.232.050 21.539 1.000 1.261 Financial support -.074.034 4.806 1.028.929 In custody at assessment.939.107 77.187 1.000 2.556 Indigenous status -.465.211 4.852 1.028.628 Constant -1.279.587 4.753 1.029.278 Final logistic regression model for completed / not completed Variable B S.E. Wald df Sig. Exp(B) CISP as bail condition.740.182 16.571 1.000 2.096 Age group -.278.090 9.512 1.002.758 Current drug use -.543.160 11.565 1.001.581 Accommodation stability.100.031 10.325 1.001 1.105 Financial support -.153.044 12.368 1.000.858 In custody at assessment.644.131 24.270 1.000 1.905 Indigenous status -.473.241 3.865 1.049.623 Constant.139.699.039 1.843 1.149 Final logistic regression model for completed / not attended The Exp(B) values and Wald statistic provide measures of the contribution of each variable to the regression model. It can be seen that the two models show a similar ordering of variables, with CISP as a bail condition and In custody at assessment the two most important components of the models, followed by Accommodation stability. This ordering of variables is also reflected in the order of inclusion in the forward stepwise process, with In custody at assessment and CISP as a bail condition the first and second selected variables in both models. Melbourne Criminological Research and Evaluation 121

Both models provide a good fit for the data, as measured by the Hosmer & Lemeshow goodness of fit statistic. However, the classification tables show an important difference between the two models. For the regression model for completed/ not completed the model correctly predicts the outcomes of two-thirds of the cases, but is about twice as good at predicting completion outcomes as it is at predicting non-completion outcomes. In contrast, the regression model for completed/ not attended correctly predicts a higher proportion of cases overall, but fails to predict not attended cases. This suggests that the partitioning of completion outcomes is not a useful strategy as the excluded cases contribute important predictive information to the model. Actual value Predicted values Completed Not completed Percent correct Completed 862 194 81.6 Not completed 418 283 40.4 Overall Percentage 65.2 Classification table for final logistic regression model for completed / not completed Actual value Predicted values Completed Not completed Percent correct Completed 1027 29 97.3 Not completed 336 14 4.0 Overall Percentage 74.0 Classification table for final logistic regression model for completed / not attended Melbourne Criminological Research and Evaluation 122

APPENDIX 2 EVALUATION METHODS Analysis of PSCMS data The analysis of client case flows, service referrals and program outcomes used data extracts from the Platypus Systems Case Management System (PSCMS) used by CISP staff to record assessment and case management information. Extracts were provided in the form of de-identified unformatted text documents. These were transferred to Excel format and from there into SPSS data files. Data files incorporating information from multiple PSCMS files were compiled by matching on case and client numbers. The client number is retained by an individual across multiple CISP episodes, while a new case number is assigned at each new contact episode. Details of the file structures can be found in the PSCMS Export Manual. Extracts were made at five points during the period of the evaluation. An initial extract is March 2007 allowed data quality issues to be checked. The first interim report on program case flows was done using an extract taken in June 2007 (six months after program commencement). A third extract at the end of 2007 allowed further monitoring of case flows and served as the basis for the second interim report in February 2008. A third monitoring report was based on data extracted in June 2008, and a final extract in March 2009 provided the basis for the final evaluation report. Note that some internal CISP reporting is based on end of month or quarter PSCMS reports while the evaluation extracts were done some weeks after the end of the period of interest. As a result there are likely to be minor discrepancies in reported case numbers and other measures. SF-12 Survey of Health and Well-being The Sf-12 survey was administered to CISP clients at the beginning and near the end of their time on the program. Survey forms were distributed to CISP staff at each venue together with notes on administration, and evaluation team members visited each venue and briefed staff on the administration of the survey. Participation in the survey was Melbourne Criminological Research and Evaluation 123

voluntary, and the survey questions were administered verbally by the participant s case manager who also recorded the clients responses. Only participants in the intermediate and intensive levels of the program were eligible. Completed forms were held at the program venue and either collected directly by evaluation team members or mailed to them. The SF-12 was administered over a twelve month period from February 2008. A total of 197 clients completed the survey at the time they commenced the program, and 67 were re-tested at or near their exit from the program. CISP staff and stakeholder interviews Stakeholder interviews were conducted over the course of the evaluation. The initial round of fifteen interviews (January to June 2007) focused on feedback from staff about program implementation issues. In a second round of staff interviews in the second half of 2008 covered a further eight staff members. Evaluation team members also attended conferences of CISP staff at the outset of the program and after the submission of the interim report, and team meetings at each venue throughout the project. Stakeholder interviews covered a range of service providers including Homeground, Quantum, Latrobe Community Health, DASWest, arbias, Odyssey House, Anglicare, Forensicare. Some service provider agencies were interviewed on more than one occasion over the course of the evaluation. In addition, interviews were conducted with Registry staff and lawyers at two program venues. Magistrates were interviewed in mid- 2007 in relation to program implementation issues, and a second round of interviews was conducted in late 2008 and early 2009. Copies of the staff and stakeholder interview schedules used in the implementation and outcome stages of the evaluation follow: Melbourne Criminological Research and Evaluation 124

IMPLEMENTATION INTERVIEWS WITH STAFF AND STAKEHOLDERS 2007 1. Training and experience of program staff What professional training do you have? What is your experience in working with offenders? How are program staff recruited and selected? To what extent are program managers involved in delivery of program services and the supervision of staff? 2. Development of program policy and content What work was undertaken to develop the program model? (literature review, pilot program, study of comparable programs) What assessments were undertaken regarding the need for the program? How do the values and goals of the CISP program align with the existing values of the courts and related programs (CREDIT-BSP)? Is program funding adequate? 3. Assessment of participants Is there a clear definition of the program s target group(s)? How are presenting problems (e.g. substance abuse, sexual offending, violence etc.) assessed? Is this assessment process effective and reliable? Appropriateness: Is the type of client presently received appropriate to the target group? Exclusions: Is there a rational basis for the exclusion of certain clients? Are styles or modes of service matched to characteristics of offenders? 4. Intervention model What do stakeholders understand to be the theoretical basis for the program? What documentation exists for program policy and practice? How adequate is this? What training do program staff members receive in program policy & practice? How effective was this? What criminogenic behaviours and attitudes are targeted? What other behaviours or deficits are targeted? How are interventions structured in relation to assessed risks and needs? 5. Feedback and quality assurance processes What quality assurance is provided through program checks, clinical supervision, client feedback Melbourne Criminological Research and Evaluation 125

Is consumer satisfaction measured? How is this information used? What ongoing assessment is conducted of clients in relation to target behaviours? What other follow-up data is gathered? How is feedback about program implementation and outcomes disseminated to stakeholders? Melbourne Criminological Research and Evaluation 126

CISP SERVICE PROVIDER INTERVIEW SCHEDULE 2007 What does your agency do? What services does your agency provide that are required by CISP referees? What is your role in relation to CISP service provision? What is the basis for your service relationship with CISP? (Written protocols etc) How useful, comprehensive, effective is the policy framework for your relationship with CISP? Is funding adequate? Can you assess the quality of your service links with CISP staff in relation to: - Communication about individual clients? - Information about legal status issues (bail, sentence, supervision requirements)? - Review and development of services and programs/ 1. What are the main problems or issues that characterize clients referred to you by CISP? (Circle more than one if appropriate, rank in order of significance) a) _ Drug and alcohol abuse/ addition b) _ Homelessness or accommodation problems c) _ Poor educational achievement and/ or long term unemployment d) _ Mental illness e) _ Physical health problems f) _ Limited social and familial support links g) _ Other, please specify 2 a. To what extent are these problems related to their offending behaviour? 1. Very related 2. A little related 3. Neither related nor unrelated 4. A little unrelated 5. Very unrelated 2 b. How are the problems related to their offending behaviours (e.g. need to support additions, etc)? Melbourne Criminological Research and Evaluation 127

3. In general, how receptive were the CISP referrals to the services provided? 1. Very receptive 2. A little receptive 3. Neither receptive nor unreceptive 4. A little unreceptive 5. Very unreceptive 4 a. In general, how useful was it to provide services to CISP referee s while they were on bail? 1. Very useful 2. A little bit useful 3. Neither useful nor useless 4. A little useless 5. Very useless 4 b. (IF RESPONSE NOT 3) Why was it useful or not useful to provide CISP referees with services while they were on bail (e.g. were they more motivated because of what was at stake, or were they in too great a state of crisis)? 5 a. In general, how appropriate do you view CISP referrals in relation to the services provided by your agency (e.g. does your agency provide services that are often required by CISP referees)? 1. Very appropriate 2. Somewhat appropriate 3. Neither appropriate nor inappropriate 4. Somewhat inappropriate 5. Very Inappropriate 6. In general, what level of priority for treatment do CISP referrals receive from your agency? 1. High priority 2. Moderately high priority 3. Moderate priority 4. Low priority 5. Very low priority What needs to be done to improve things? Melbourne Criminological Research and Evaluation 128

SERVICE PROVIDER INTERVIEW SCHEDULE 2009 Description of the services provided to CISP clients Relationships and communication 1. Are there formal protocols for engagement between CISP and the referral agency? How well are these protocols understood and adhered to, both on the part of your own agency and CISP staff? 2. How much investment of time and other resources goes into developing links between CISP and your own service agency? 3. Do you ever encounter grey areas or contradictions in treatment models that formal protocols do not adequately address? Example? Does double up occur is the services provided by CISP and your service? 4. Are there formal protocols for identifying and meeting specific cultural/ linguistic/ religious needs of clients from different cultural backgrounds? 5. Does the assessment and screening done by CISP case mangers prior to referral provide you with adequate information about the client to make a decision about accepting or not accepting a referral? What further assessments would you need to conduct with a client post referral? 6. Are there consistent referral strategies between case managers? Do some case managers refer to you more than others? (e.g. both under and over-referral 7. In general, how appropriate are the referrals you receive from CISP case managers? If not, in what way are referrals inappropriate? What resources are available to case managers to inform them of the specific services that your agencies do and do not provide? 8. How satisfied are you with communication about client and case information from CISP staff? If you are not satisfied, what specific things are not being communicated well? What are the consequences of this (e.g. risks, double up in assessment of clients etc)? 9. In general, do clients seem to understand what it is that you can or can t do for them as an advocacy worker in the context of the justice system? Melbourne Criminological Research and Evaluation 129

Appraisal of CISP model 1. How effective is the assessment and referral model of intervention at meeting the needs of CISP clients? (As opposed to a model that incorporates direct service provision to the client). 2. Are assessment outcomes and treatment recommendations made by your service made available to CISP staff for the purposes of follow up and case management? i. If so, what processes or systems support this to happen? ii. In what ways could these processes or systems be improved? 3. Is appropriate intermediate support provided to CISP clients referred to services with long wait lists, given CISP s time limited period of involvement with clients? In particular, in cases of suspected ABI or mental illness, what safety nets are provided to ensure appropriate treatment is provided in a timely fashion. 4. How effective is CISP at meeting the needs of clients (e.g. addressing drug and alcohol or mental health issues) given the time limited nature of the program. 5. Should CISP continue as a pre-sentencing program? Or is there scope for continued involvement with clients who are not given custodial sentences? 6. General views about their experience of working with CISP staff and clients: i. Does CISP improve upon other existing service models in the justice system? If so, in what ways does it do this? Are there particular kinds of clients that CISP services well? Are there some clients the program is not effective for? ii. Are there ways in which CISP is not as effective at meeting the needs of clients on bail, as other existing programs Melbourne Criminological Research and Evaluation 130

APPENDIX 3: CISP SCREENING ASSESSMENT FORM Melbourne Criminological Research and Evaluation 131