RCT
The Ministry of Justice has just – 11 December 2025 – published (the rather grandly titled) A Randomised Controlled Trial in Four Prisons: Impact of Incentivised Substance Free Living Wings on Prison Stability which is part of the Tackling Drug Misuse in Prisons Evaluation Programme. Written by Darren Churchward, Peter Smolej, Gurmukh Panesar, Maika Terashima, Jo Voisey, Hala Elsayed and Lucy Cuppleditch, this small-scale evaluation focuses not on drug treatment measures but on safety and stability outcomes by examining the rates of assault and self-harm on ISFL and non-ISFL wings. I say small-scale because although it measures outcomes across four prisons, there were just 60 people in the total cohort (28 people living on ISFLs and 32 living on other wings).
The study concludes that ISFLs have a high probability of providing a more stable environment for prisoners, compared to non-ISFL wings. Those living on ISFL wings are 31% less likely to involved in an assault, 31% less likely to self-harm and 32% less likely to be involved in a disorder incident (all three figures are medians within a large range).
ISFLs
Incentivised Substance Free Living wings (ISFLs) are defined as dedicated spaces within prisons for prisoners who want to live drug-free. On ISFLs, prisoners are given incentives (for example additional access to gym equipment, entertainment equipment, additional time out of cell) to remain drug free, alongside undergoing regular drug testing.
In an ISFL, new residents sign a compact which outlines expectations for acceptable behaviour and what leads to removal from the ISFL. The creation of a positive and supportive environment by both prisoners and staff is a key factor in the functioning of an ISFL. Therefore, the guidance for ISFLs emphasises the importance of having residents and staff who are both self-motivated and committed to the ethos of ISFLs.
ISFLs are the successors to Voluntary Testing Units which were introduced back in 1998. In April this year there were currently 85 prisons with ISFLs in operation, out of 123 prisons in England and Wales.
The study
This innovative study utilised a waitlist randomised controlled trial which randomly assigned the order in which prisoners would move from a non-ISFL prison wing to an ISFL prison wing, in four prisons. Those randomised to the top half of the list formed the intervention group, and outcomes were compared to those in the control group (bottom half of the list). The primary outcome measure was time to involvement in an assault incident. This is used as a measure of prison wing stability and was measured over a 3-month period. For the intervention group this follow-up period began from the point of moving onto the ISFL (which was dependent upon spaces becoming available), for the control group follow-up began from the point of randomisation. Results were analysed using Bayesian survival regression.
Findings
As noted, a total of 60 prisoners were involved in the final intention to treat analysis for the study’s primary outcome measures (28 ISFL and 32 non-ISFL). Bayesian survival analysis estimated there to be a 93.1% probability that ISFLs have a beneficial effect of any magnitude, in terms of reducing assault incidents. Our analysis estimates, with 80% uncertainty, the size of reduction in assault incidents to be between 5% and 50%, with a median estimate of 31% – ISFL’s compared to non-ISFL’s.
This means those on ISFL wings were 31% less likely to be involved in an assault incident compared to those on a non-ISFL wing. Similar effects were seen for self-harm (80% probability between 4% and 50% less likely to self-harm, with median at 31%) and disorder (80% probability between 6% and 51% less likely to be involved in a disorder incident, with median at 32%).
Conclusions
The study concludes that ISFLs have a high probability of providing a more stable environment for prisoners, compared to non-ISFL wings. Of course, this sort of RCT design is unable to suggest any reasons for this – but it could be argued that any wing which provides a better regime than those which most under-staffed and over-crowded prisons are currently able to deliver would result in a more stable living environment for residents and staff.
Thanks to Andy Aitchison for kind permission to use the header image in this post. You can see Andy’s work here
