Probation and Addiction Treatment Data Has Been Missing the Full Picture

A landmark government study shows official records have significantly undercounted how many people on probation engage with drug and alcohol treatment. People who stay in treatment are far less likely to reoffend. That finding alone should reshape how policymakers think about community sentence treatment requirements.

The Ministry of Justice (MoJ) and the Office for Health Improvement and Disparities (OHID) published the report jointly in March 2026. Researchers tracked nearly 46,000 community orders and suspended sentence orders. Each carried either an Alcohol Treatment Requirement (ATR) or a Drug Rehabilitation Requirement (DRR). The observation period ran from August 2018 to March 2023.

What they uncovered raises urgent questions about how the system measures success.

The Data Gap Nobody Was Talking About

Researchers at the MoJ and OHID first tried to match probation records with structured treatment data from the National Drug Treatment Monitoring System (NDTMS). They could only link 49% of the 45,943 requirements issued. On the surface, that suggested roughly half the people given community sentence treatment requirements simply did not show up.

That picture turned out to be seriously incomplete.

Researchers used Natural Language Processing (NLP) to read thousands of probation contact notes. They found that an additional 41% of those unlinked records showed clear evidence of treatment engagement. The standard data matching process had simply missed it.

When researchers combined both sources, up to 90% of people with ATRs and DRRs had engaged with some form of treatment. For those specifically on Alcohol Treatment Requirements, that figure reached 93%.

This is not a minor data quirk. Policymakers have made decisions about probation and addiction treatment on figures that missed nearly half the picture.

Staying in Treatment Cuts Reoffending Sharply

The study produced some of the clearest evidence yet that sustained treatment engagement works.

Within 12 months of sentencing, 36% of people linked to structured treatment were reconvicted. Among those with no identified treatment link, that figure rose to 44%. The gap is significant. It grows wider when you break outcomes down further.

For people on ATRs who completed or stayed in structured treatment, the reconviction rate sat at just 13%. Among those who dropped out, it jumped to 41%. For DRRs, the contrast was even more striking. People who remained in treatment saw a 26% reconviction rate. Those who dropped out saw 60%.

Engagement with community sentence treatment requirements is not a bureaucratic formality. It is one of the strongest predictors of whether someone goes on to reoffend.

Who Engages With Probation and Addiction Treatment Most Successfully

The study shed light on who is most likely to link to structured treatment after receiving a sentence with an ATR or DRR.

People aged over 50, women, and those in settled accommodation all showed higher engagement rates. Quick access to treatment also made a measurable difference. For ATRs, starting within three weeks of sentencing produced the highest completion rates, reaching 57%. For DRRs, the best window fell between three and six weeks.

Rough sleeping stood out as a serious barrier. Only 36% of ATRs and 37% of DRRs involving people sleeping rough linked to structured treatment. For those in settled accommodation, those rates rose to 58% and 48%.

Housing instability does not just make daily life harder. It actively blocks access to the support a court order is meant to provide.

The Regional Picture Is Uneven

Where someone lives clearly affects their chances of accessing treatment.

For ATRs, the North West, East of England and London each recorded a 58% linkage rate to structured treatment. In South Central, only 43% of ATRs linked to treatment at all.

For DRRs, London stood alone as the only region where roughly half of all requirements linked to structured treatment. The South West recorded the lowest rate, at 40%.

The independent sentencing review previously described this variation as a postcode lottery. The data gives that phrase real statistical weight. Courts across England issue community sentence treatment requirements every day. What actually happens next depends heavily on location.

AI Tools Are Filling Gaps in Justice Research

One of the more striking aspects of this report is what it shows about the potential of artificial intelligence in justice data work.

The NLP model read probation contact notes and distinguished between meaningful details. It could tell the difference between a note recording that someone expressed interest in a support group and one confirming that they attended. That kind of contextual reading goes well beyond what keyword searches or data matching alone can do.

Researchers describe the methodology as experimental. Contact note quality does vary across practitioners. But the results made enough of a case that the report recommends developing the approach further. A separate NLP model to analyse reconviction data could follow.

What Probation and Addiction Treatment Policy Needs to Do Next

The report does not make sweeping policy demands. But its findings point in clear directions.

Speed matters. For ATRs, each additional week before treatment starts links to a higher dropout rate. People who waited more than 12 weeks before beginning treatment showed a 44% dropout rate. Those who started within three weeks showed just 32%.

Stable housing is not a separate issue from community sentence treatment requirement outcomes. It sits at the centre of them. Services working with people on ATRs and DRRs need to treat accommodation as a core part of the plan, not a side concern.

The data infrastructure also needs urgent improvement. The report calls for a common unique identifier, such as an NHS number, to link health and justice records accurately in future. Without it, researchers will keep patching gaps with probabilistic matching and AI workarounds.

The Broader Case for Treatment Over Custody

A companion report cited alongside this study found that people sentenced with an ATR took an average of 12 more days to reoffend compared with those leaving short custodial sentences. DRR recipients were less likely to receive a further custodial sentence: 47%, against 53% for those released from short custodial terms.

The case for probation and addiction treatment is not only about individual recovery. It is also about what happens in communities when people either get or do not get support at a turning point.

This study builds that case on firmer evidence than ever before. Whether the system acts on it remains the open question.

Source: dbrecoveryresources