Evaluating P-ASRO outcomes | findings

High quality evidence is required to show that P-ASRO is working and is a cost effective drug treatment programme to run within prisons. A randomised controlled trial would be the most robust evaluation method to provide this evidence, though there would be some  challenges to conducting such a trial.

Random allocation should therefore be carried out by an external team and it would be necessary to have measures in place to manage any negative impacts on offenders. Ethical and practical barriers were also identified in relation to the control group condition, and the receipt of 'treatment as normal' was the preferred approach. However, what this treatment would be was not entirely clear, though suggestions included work with CARATs teams and possibly a mixture of other drug treatment programmes and interventions in general.  It was felt P-ASRO should be shown to have 'better outcomes' than from the offenders involvement with CARATs due to the additional resources it requires to run. However, the likelihood of P-ASRO having better outcomes than the other Offending Behaviour Programmes (OBPs) was questioned.  An important requirement of the control condition was that there was a reasonable chance of demonstrating that P-ASRO has a measurable impact.

A sufficient sample size could be obtained by changing the level of P-ASRO provision within sites, such as running the programme less frequently or with smaller groups so that some currently eligible prisoners can be included in the control group. A second option would be to change the current eligibility requirements. A possible upcoming policy change regarding eligibility and the most widely discussed option for an RCT, would be to include offenders on a methadone maintenance prescription. Including those on a methadone maintenance prescription had both advantages and disadvantages.

This was based on both views of what the aims of the evaluation should be and because of practical issues about the way the programme is currently being implemented. Including only sites at an agreed 'best practice' standard of delivery would allow the evaluation to show whether P-ASRO works when delivered with integrity. If the findings from an evaluation would be required to be generalisable across the prison estate it would be necessary to include a diverse range of sites and offenders. However, if it was felt an evaluation should focus on whether P-ASRO 'worked', irrespective of whether this was for all or a majority group, having such sample diversity may not be necessary.

This support would include the MoJ, Directors of Offender Management (DOMs), governors, operational staff and offenders. From the outset it would be also be necessary to have systems in place that informed an evaluation team of any upcoming policy changes which could impact on the running of an RCT.

Broader areas discussed included: substance misuse (both type of drug use and how it is used); offending behaviour; engagement with services; and, social, health and practical outcomes. However there was also difficulty in articulating what specific outcomes could be attributable to P-ASRO, especially as the intervention was seen as playing a part in the whole system of rehabilitation, and its success was also dependent on the larger system it is working within.

An evaluation could use a range of existing P-ASRO datasets as well as external datasets to monitor outcomes, though it would be important to bear in mind their reliability, completeness, the resources involved in accessing them and their validity. The idea of running a follow up survey to generate new data on outcomes was supported. However, it would be necessary to have systems and strategies in place to manage attrition. Probation were identified as a good resource to use for facilitating access to ex-offenders. However, data collection by completely neutral interviewers was preferable for obtaining accurate responses. A smaller qualitative component is recommended and would contextualise the quantitative outcomes as well as providing explanatory insight.  It would also be important to manage offenders' concerns about participating in this stage of the evaluation, especially post-release.