Evaluation of the Arrest Referral Pilot Schemes

Listen

EXECUTIVE SUMMARY

CHAPTER ONE - BACKGROUND TO ARREST REFERRAL, THE PILOT SCHEMES AND THE EVALUATION

1.Arrest Referral ( AR) is one of a range of recent policy initiatives intended to disrupt the link between substance misuse and offending by improving uptake of services among arrestees whose offending is linked to drug or alcohol use. The development of AR was given new impetus in Scotland by the announcement, in 2003, of Scottish Executive funding for a series of pilot projects. The 6 schemes funded were: Edinburgh & Midlothian ( EMARS), Tayside, Renfrewshire, East Renfrewshire & Inverclyde, all extended from existing projects and three new schemes in Lanarkshire; Dumfries & Galloway and Glasgow. All were operated by voluntary sector agencies under contract to the Social Work Department, and were multi-site, except Glasgow. The schemes included a mix of police-based and court-based locations and of police-mediated and direct access to arrestees. The Scottish Centre for Social Research (ScotCen) and the Social Work Research Centre at Stirling University ( SWRC) were commissioned to conduct a mixed method, process and outcome evaluation with three largest projects providing case-studies ( EMARS, Glasgow and Tayside).

CHAPTER TWO - PROCESS ISSUES RELATING TO PLANNING, IMPLEMENTATION AND OPERATION

2. The pilots were at widely different stages of development and therefore the relevance of key issues varied across schemes. Initially the preparation of suitable premises and AR staff recruitment, plus the associated Disclosure Scotland process, proved to be the most problematic areas. Other tasks included assessments of risk within police offices; training of all operational staff involved; final agreement of protocols and procedures including information sharing agreements; and building awareness of the scheme.

3. The location of the AR scheme, within a police office or court setting, clearly affected the way it operated. There were more challenges to overcome in introducing the AR process within the context of a busy police office, open 24 hours a day, with large numbers and high turnover of police staff working changing shifts. AR had to fit alongside the primary task of custody management in an environment with its own strong sense of professional identity. Good working relationships between AR and police staff were developed but took time to build and maintain. Furthermore individual officers varied in their understanding of and commitment to the scheme, and their ability and willingness to promote its benefits to appropriate arrestees. Many issues were common across police settings but there were important variations - whilst the police risk assessment for the new Glasgow scheme precluded direct access by Arrest Referral workers to arrestee in cells, more established schemes, notably at Dundee and earlier at Dalkeith, had negotiated direct access.

4. Direct-access, to prisoners at the cells, appeared more possible in the court setting where office-type hours operate and there is greater continuity of staffing. The custody staff (from Reliance) have a more limited role and found it relatively easy to accommodate the needs of AR. Custody staff, in both settings favoured interview spaces with dividing screens, for safety reasons, but these were often perceived as a barrier by Arrest Referral workers.

5. Stakeholders thought that: the time in custody offered arrestees an important opportunity for reflection and was an excellent opportunity for supportive intervention. They also thought arrestees understood that compliance with AR would not affect subsequent prosecution or disposal, although it was also felt that Defence Agents might use it in mitigation and that Sheriffs might be influenced at some level.

CHAPTER THREE - ARRESTEES, THROUGHPUT AND AR ASSESSMENT

6. Pilot throughput ranged from 100-900 a year with the 3 largest, urban schemes achieving a higher level of AR interviews than expected. Overall, those arrestees accepting referral to AR were predominantly male, white and aged under 40. The vast majority were not in employment and were repeat offenders. Pilots successfully reached arrestees with substance misuse problems - of arrestees interviewed in the 3 case study areas, only 8% had not used either drugs or alcohol in the 24 hours prior to their arrest and around 1 in 6 had used both. Arrestees interviewed also confirmed that drug or alcohol use contributed to the arrest in at least 52% cases, and 39% of survey arrestees had committed crime in the past 4 weeks to fund their habit. Many arrestees interviewed also indicated that substance misuse was disrupting their lives through dislocation of relationships with family and friends.

7. Information on those refusing the offer of AR was only available from the Glasgow police-mediated scheme. Those refusing AR were more likely to be young (aged 16-20), male and to have been arrested in relation to drunken and violent behaviour. Conversely, older arrestees, females and those arrested in relation to court-related offences were relatively more likely to accept the offer.

8. In all areas, ARWs provided basic harm reduction information, assessment, new referrals on to other agencies or liaison with services which the arrestee was already in contact with. The completion of the baseline assessment (Glasgow) and the Drug Problem Service Assessment (Tayside) was recognised good practice.

CHAPTER FOUR - REFERRING ON, OUTCOMES AND COSTS

9. The ease of information exchange between AR teams and service agencies affected stakeholders' perceptions of the efficacy of referrals. The Glasgow scheme was unique, in that it was embedded in the local Community Addiction Team ( CAT) structure and this - along with its co-location with the CAT and other statutory services - facilitated easy referral on. Other pilot schemes referred to a multiplicity of service agencies using multiple pathways. Here the establishment of effective referral mechanisms took time and shared commitment, for example the Tayside Drug Prescribing service trained all the ARWs in completing the TDPS assessment and the EMARS team negotiated to hold surgeries at several service agency premises.

10. It appears that around two-thirds of initial AR interviews in Glasgow and 30-40% in Tayside and EMARS were with arrestees not linked to support and treatment services. Most arrestees interviewed were referred on: in Glasgow virtually all to the local CATs who would facilitate further assessment, support and treatment; whilst in Tayside and EMARS, around three-quarters were referred to a variety of specialist drug and alcohol support agencies. Of those referred on to services, a smaller proportion were 'new' services users in EMARS and Tayside (around 40-50%); whilst in Glasgow the Community Addiction Teams found as many as 69% not previously known to them. Not surprisingly, arrestees offered help in relation to alcohol were less likely to have been offered help or support previously.

11. Pilots had difficulty in establishing robust systems for collecting data on arrestee contact with and retention in services. Small scale tracking exercises suggest that a clear majority of arrestees referred will attend at least one appointment and that of these many will still be in contact with services weeks or months later. Stakeholders identified service waiting times as an issue in Tayside, and Dumfries and Galloway, however incomplete monitoring data prevented examination of the impact of this upon the AR services

12. The vast majority of arrestees interviewed (84%) said they would recommend AR to other people - 45% would do so 'strongly'. Over a quarter (28%) had heard of AR prior to arrest, but awareness was highest at longer established schemes where 45% survey arrestees had previously seen an ARW. The majority of survey arrestees understood the voluntary nature of AR and around half understood that compliance with AR would make no difference to their trial and sentence - however 15% thought it would make a difference and the remaining 32% were not sure. Only 24% (37) survey arrestees were traced for follow-up, they continued to positively endorse AR but there was little reported on its impact on them.

13. The cost of AR varied from around £75 per offer of AR to £340 per achieved initial AR interview (in a police setting). Average costs per AR interview (£150) appeared lowest where a direct-access model in a court setting was followed. The national pilots considered the Scottish Executive ( SE) funding essential to underpin the initial multi-agency planning and start-up process and particularly in Tayside, Edinburgh and Glasgow to enable schemes to meet the current high level of demand. Non SE funded schemes were of similar format and experienced similar operational difficulties, were under review or were applying for SE funding.

CHAPTER FIVE - CONCLUSIONS

14. The anticipated level of demand for AR was reached or exceeded by the EMARS, Glasgow and Tayside pilots. The smaller pilots in police settings faced more difficulties in fully establishing schemes and the Lanarkshire and Dumfries and Galloway pilots were moving towards accessing arrestees via courts. Court settings achieved a much higher than expected level of initial AR interviews in the year (Dundee, 545; EMARS, 907 - almost twice as many as projected, largely as a result of gaining direct access at the Sheriff Court).

15. All pilots appeared to be successful in contacting the main target groups - individuals with significant drug and alcohol problems and offending behaviour linked to substance misuse. Schemes targeted at both drug and alcohol users were seeing both target groups but the relative proportions of individuals with drug use, alcohol use and poly-drug use was not clear. Glasgow saw the largest proportion of alcohol users.

16. AR is not a standalone service but part of a range of pathways into treatment and support. For AR to be effective the onward referral routes and care pathways need to be clearly specified and waiting times for treatment and support services need to be as brief as possible. All stakeholders valued the Glasgow model, where the AR service is fully embedded within the fully integrated community addiction service. In other pilots, specific mechanisms developed, such as use of service agency assessment forms and shared drop-in arrangements, smoothed the referral pathway from AR to services.

17. Professional stakeholders highlighted the potential significance of even relatively limited interventions at the point of arrest. The arrestees interviewed confirmed that the provision of harm reduction information brought immediate benefits in some cases and service information was a catalyst towards contacting services either immediately or at a later date.

18. The evaluation found reasonable grounds for supposing that AR is successfully linking arrestees into services. Most arrestees seen by ARWs were being referred on, almost all in Glasgow's integrated service and around three-quarters from Tayside and Edinburgh. Stakeholders considered the AR advocacy and safeguarding role with 'existing' service users (30-60%) to be as important as the signposting role for 'new' service users (40-69%).

19. Small scale tracking exercises provided some grounds for optimism about longer-term engagement or retention in support and treatment agencies: around three quarters of Edinburgh and Glasgow arrestees referred to agencies had attended at least one appointment and furthermore a substantial proportion were still in contact with the agencies concerned weeks or months later.

20. AR in a police setting faces more structural and organisational challenges than AR in a court setting, where there is less potential for conflict of purpose. The changeover to Reliance for custody transportation was a major factor influencing AR operating hours at police offices and in particular in EMARS' (and Lanarkshire's) decision to operate from court. It is more difficult to maintain a comprehensive and consistent offer of AR via police custody staff whose primary focus is one of custody management. Direct access by ARWs is more possible within the court setting but has also been very successful in the police setting at the longer established Dundee site. ARWs prefer direct-access as they feel better placed to identify need and to 'sell' AR to arrestees.

21. Decisions about the location and operational hours of AR schemes should reflect the chosen target arrestee groups. Police settings are more likely to offer opportunity to capture arrestees at an early stage of their substance-related offending career but the ARW needs to be on site at the right time. AR teams in police settings felt that they were missing a proportion of the following vulnerable groups: young people aged 16-25; vulnerable women; those arrested for minor offences and released early; those too intoxicated to be interviewed at that time.

22. Court-based AR ensures high throughput and is cost effective, while reaching a higher proportion of repeat offenders. Court based ARWs could potentially see all police custodies transferred to court but would miss those released from police custody with an undertaking to appear at court from the community.

23. A key aim of the evaluation had been to obtain data on offending patterns pre- and post-referral to AR, but for a variety of theoretical and logistical reasons, this proved not possible within the short time-scale of this evaluation.

24. The national monitoring framework was not consistently implemented and poor data quality and non comparable information proved a significant constraint on the evaluation. Although monitoring arrangements to record basic throughput were adequate if not consistently maintained, data on referring on, contact with and retention in services was poor.