Prison sex-offender treatment programme

Historical account. This tab describes a programme that has been withdrawn. It is not a treatment guide. It sets out what the programme involved, how it may have caused harm, how its components can still show up in practice, and what replaced it.

1. What it was

Version followed: the prison-based Core Sex Offender Treatment Programme (Core SOTP) of HM Prison Service, later NOMS and then HMPPS, in England and Wales. It was accredited in 1992 and withdrawn in 2017. The account mainly describes the programme as delivered to the men in the Ministry of Justice (MoJ) impact evaluation, who started between 2000 and 2012 (Mews, Di Bella & Purver, 2017). It also refers to the earlier 1990s version, the related Extended, Rolling and Booster programmes, and the interim revisions of 2015. [source: Mews et al. 2017; Grubin & Thornton 1994; The Detail 2017]

Who developed and ran it. HM Prison Service designed it as a national, multi-site, cognitive-behavioural group programme. The Prison and Probation Services Joint Accreditation Panel, later CSAAP, accredited it for prison use in 1992. Its early content was closely modelled on programmes in other jurisdictions. Trained prison staff delivered it as facilitators: psychologists, psychology assistants, officers and probation staff. [source: Mews et al. 2017; Stanbury 2017] The MoJ report does not describe how facilitators were qualified or trained. [uncertain]

Where and when. It ran in about one-sixth of male prisons in England and Wales. The Northern Ireland Prison Service delivered it from 1998, at Maghaberry and Magilligan, with 44 enrolments from 2001. [source: Mews et al. 2017; The Detail 2017]

Stated rationale. The programme assumed that sexual offending cannot be cured, but that men can learn to recognise and manage the thinking and behaviour that led to their offending. [source: Big Issue North 2022, account of Dr Rebecca Myers] The 1994 national strategy described the Core Programme as structured group work designed to increase a man's motivation to avoid reoffending. It had three main targets: challenging cognitive distortions, building empathy for victims, and mapping each man's offence cycle as the basis for relapse prevention. Relapse prevention ran through the whole model. [source: Grubin & Thornton 1994] By the 2000s the main targets were:

cognitive restructuring of offence-related thoughts and beliefs;
an offence account required to be "consistent with the victim's statement";
exercises to increase victim empathy. [source: Mews et al. 2017]

How it was sold to prisoners, referrers and decision-makers. Core SOTP went into men's sentence plans as the accredited route for addressing sexual offending risk. For men serving indeterminate sentences, completing the plan could decide whether they were released. For men on determinate sentences, it was presented as a way to ease licence conditions. Ievins calls this "hard incentivisation": men experienced the course as a hurdle to clear, not a chance to change. [source: Ievins 2017] Men were told they needed SOTP for release decisions. Men who denied their offence could not take it, and men who would not work in a group were told they could not progress. [source: Stanbury 2017] A 2003 Home Office study found a positive effect for the pre-1996 programme, and this helped keep the programme in use (Friendship, Mann & Beech, 2003). [source: Stanbury 2017; Birmingham repository record]

Related programmes (named variants).

Core SOTP: the main programme. From 2006 it was for men assessed as medium risk or above on RM2000/S. [source: Mews et al. 2017]
Rolling SOTP: usually used for lower-risk men. [source: Mews et al. 2017]
Extended SOTP: for high and very high-risk men who had already completed Core. Combined Core and other programmes could total up to about 360 hours. [source: Ievins 2017; Mews et al. 2017]
Booster programme: delivered close to release. [source: Grubin & Thornton 1994]
Adapted SOTP, for men with intellectual disabilities, and Better Lives Booster are widely named in the UK literature, but no source describes them. [uncertain]

2. What was actually done

This is a typical course as described in the sources. Session-by-session manual content was not available and is not reproduced. Where the order of components is inferred, it is marked.

1. Selection and assessment. The intended criteria were:

a sentence of 12 months or more;
a current or previous sexual offence;
willingness to engage;
not being in categorical denial of the offence.

From 2006, Core was limited to men at medium or higher risk on RM2000/S. Treatment managers prioritised men with less time left to serve. [source: Mews et al. 2017] The 1994 strategy describes a standard assessment package at a prison assessment centre to identify treatment needs and set baseline measures. [source: Grubin & Thornton 1994] Phallometric testing was used in UK assessment for part of this period, but no source confirms it was part of Core SOTP selection. [uncertain]

2. How consent was obtained. Formally, taking part was voluntary and needed the man's consent. [source: The Detail 2017] In practice, consent was shaped by four pressures:

SOTP sat on the sentence plan;
the Parole Board and prison decisions on moves to open conditions looked for it;
men who denied their offence were excluded and so could not complete their plan;
men waited past their tariff for a place.

Ievins met dozens of men held beyond tariff while waiting for courses. One man refused a move to open conditions said the only thing missing from his plan was "Just the SOTP". [source: Ievins 2017] Stanbury describes a client who maintained his innocence and was refused release because he had not done SOTP. He was freed after eight years, only when his conviction was quashed. [source: Stanbury 2017]

3. Setting, staff and format. The programme was delivered in closed groups of about eight men in prison treatment rooms, by facilitator teams. The course took about 180 hours. [source: Mews et al. 2017] The 1994 Core programme was 70 to 80 hours, so the dose more than doubled over time. [source: Grubin & Thornton 1994] Exact session frequency and length are not stated in the sources. Contemporaneous accounts commonly describe several sessions a week over several months. [uncertain] Groups were the only mode of delivery. There were no individual modules, which the MoJ authors later identified as a possible problem. [source: Mews et al. 2017]

4. Component: establishing the group and challenge norms. Men were expected to talk about their offending in front of peers. Peers and facilitators questioned and challenged that talk as a core group task. [source: Grubin & Thornton 1994; Mews et al. 2017] The exact rules and opening sessions were not available. [uncertain]

5. Component: the offence account (disclosure). Each man gave a detailed account of his offence to the group. The account was judged against the victim's statement, and the group and facilitators questioned it until it was "consistent" with that statement. Minimising, justifying or leaving out details were treated as targets. [source: Mews et al. 2017] Facilitators engaged closely with offence detail. One facilitator later described being unable to think about the child victims while in the room, as a way to cope. [source: Big Issue North 2022] In practice the whole group heard every member's offence account in detail. The MoJ authors later suggested this could normalise the behaviour and allow contacts and sources to be shared. [source: Mews et al. 2017]

6. Component: cognitive distortions. Cognitive restructuring was used to identify and challenge offence-supportive thoughts and beliefs. These included excuses, minimisations, and beliefs about victims, entitlement and harm. Challenge came from facilitators and peers. [source: Mews et al. 2017; Grubin & Thornton 1994] The specific exercise formats were not available. [uncertain]

7. Component: victim empathy. Structured exercises aimed to build cognitive and emotional understanding of the victim's experience. [source: Mews et al. 2017; Mann & Barnett abstract] In the 1990s, one documented method was the offence re-enactment role play:

the man played his own victim;
the facilitator played the man, in the offender role;
props were sometimes used;
the scene was the offence itself.

A facilitator recalls a man reduced to tears while playing his abused niece. Looking back, she said: "I can't believe we did that." [source: Big Issue North 2022] Victim empathy work also commonly included writing letters to the victim that were not sent, and hearing or reading victim accounts. The sources do not confirm these for Core SOTP specifically. [uncertain]

8. Component: offence cycle and relapse prevention. Each man mapped his offence cycle or chain of events, from background factors to the offence. This map was the basis for a relapse prevention plan identifying high-risk situations and avoidance strategies. [source: Grubin & Thornton 1994] Later versions are generally described as adding future-oriented and Good Lives content, but this is not confirmed in the sources. [uncertain]

9. Homework. Written out-of-session work, such as offence chains, accounts and letters, is typically described. The sources do not specify it. [uncertain]

10. How progress was judged. At the end, facilitators wrote a post-programme review or report. It recorded whether the man had a "positive" outcome and fed into parole, sentence planning and category decisions. [source: Ievins 2017] Facilitators interviewed later worried that victim empathy work mainly produced superficial compliance rather than real change. Displays of remorse or empathy are easy to rehearse, but they were a visible marker of progress. [source: Norton 2013] The scoring criteria for reviews were not available. [uncertain]

11. How non-response, denial and refusal were handled.

Men in categorical denial were excluded. [source: Mews et al. 2017]
Refusing or failing to complete meant the sentence plan was incomplete, with knock-on effects on release and progression. [source: Ievins 2017; Stanbury 2017]
Men at high risk could be sent on to Extended SOTP. [source: Ievins 2017]

12. Interim revisions (2015). From 2015 the programme moved away from developing empathy for past victims and towards preventive strategies. Men were encouraged to identify and take responsibility for the risk factors behind their offending, instead of giving detailed offence accounts in the group. [source: The Detail 2017] The MoJ authors note that offence accounts and victim empathy were "phased out" as better evidence on risk factors emerged, citing Mann, Hanson & Thornton (2010). [source: Mews et al. 2017]

3. How it caused harm

3.1 Increased reoffending in the national evaluation. The study compared 2,562 men who started Core SOTP between 2000 and 2012 with 13,219 matched men who did not take it. Propensity score matching used 87 variables. Average follow-up was 8.2 years, with a maximum of 13.9. [source: Mews et al. 2017]

Outcome (MoJ 2017)	Treated	Comparison
Proven sexual reoffending	10.0%	8.0%
Sexual offences per offender	0.59	0.45
Child sexual image reoffending	4.4%	2.9%
Extended SOTP starters, sexual reoffending (sensitivity analysis)	17.5%	10.0%

The MoJ reported 51 more reoffenders in the treated group. [source: The Detail 2017] Matching could not account for deviant sexual interest or other unobserved factors, such as self-selection, so the result is an association and not proof of causation. [source: Mews et al. 2017] For clinicians, the practical point is that a large, intensive, mandated-in-effect programme showed no benefit and a probable small harm.

3.2 Mechanisms of harm. The MoJ authors proposed the first three of these mechanisms; the rest are from other sources, as tagged.

Normalisation through group exposure. Hearing detailed accounts of other men's offending may have made the behaviour seem more normal. [source: Mews et al. 2017]
Sharing contacts and sources. In the worst case, group members may have exchanged contacts and sources linked to offending. This fits the larger rise in child image reoffending. [source: Mews et al. 2017]
Group-only delivery. There was no individual work to address each man's needs. [source: Mews et al. 2017]
Targeting the wrong things. Effort went into offence-account accuracy, victim empathy and general distortions. These were later judged weak or non-criminogenic targets. A review of three meta-analyses suggests victim empathy work is unnecessary or even harmful. [source: Mews et al. 2017; Mann & Barnett abstract]
Shame. Shame-inducing work, such as re-enactments, detailed disclosure and confrontation, can deepen shame instead of building responsibility. The process study of the replacement programme recommended more work on shame and guilt. [source: Big Issue North 2022; Wilkinson & Powis 2019] The idea that shame (as opposed to guilt) works against change is a mechanism from the wider literature, not tested here. [synthesis]
Performance instead of change. Because progress reports affected release, men learned to show the expected remorse. Facilitators suspected this compliance was superficial. [source: Norton 2013; Ievins 2017]

3.3 Documented harms to participants.

Detention beyond tariff while waiting for, or being refused, the course. [source: Ievins 2017]
Continued detention of men maintaining innocence, including at least one later found wrongly convicted. [source: Stanbury 2017]
Distress after the 2017 findings. One man said: "I spent seven years on that. Seven years away from my family." Others feared that having completed the programme would now count against them at parole, and their trust in prison decision-making fell. [source: Ievins 2017]
Emotional distress during re-enactment role plays. [source: Big Issue North 2022]

No physical injuries or deaths attributable to Core SOTP are reported in the sources. Self-harm or suicidal crises linked to disclosure work are plausible but undocumented here. [uncertain]

3.4 Harms to victims and the public. The probable rise in sexual and image reoffending is itself the main harm to the public. [source: Mews et al. 2017]

3.5 Harms to staff. Facilitators were exposed repeatedly to graphic offence material. One described being "devastated" by the findings and blamed rapid expansion that "diluted" quality. [source: Big Issue North 2022] Facilitators also reported ambivalence about victim empathy work. [source: Norton 2013]

3.6 Who was most at risk.

Higher-risk men. Extended SOTP showed the largest gap: 17.5% vs 10.0%. [source: Mews et al. 2017]
Men with image offences, given the image-reoffending signal. [source: Mews et al. 2017]
Men on indeterminate sentences, whose release depended on completion. [source: Ievins 2017]
Men maintaining innocence, who were excluded and so could not progress. [source: Stanbury 2017]
Men with intellectual disabilities, histories of their own abuse or high shame. This is an inference: no source analysed these groups. [synthesis]

4. Recognising it today

Core SOTP itself is gone. Its components and assumptions can still appear in in-house programmes, private and community groups, overseas services that copied the UK model, risk reports and parole recommendations. [synthesis]

4.1 Features that signal the old model. All of these are [synthesis] built on §2 and §3.

Full disclosure as a condition of entry or progress. Examples: "he must give a full account consistent with the victim's statement", or admission treated as a condition of treatment.
Denial treated as disqualifying in itself, or as a central risk factor that must be broken down before other work can start. The replacement programmes deliberately accept men who maintain innocence. [source: Stanbury 2017; Wilkinson & Powis 2019]
Group-only delivery in which every member hears every other member's offence in detail.
Victim empathy as a core target using victim role play, playing the victim, re-enacting the offence, letters to victims that are not sent, or graded exposure to victim accounts presented as the route to lower risk. [source: Big Issue North 2022; Mann & Barnett abstract]
Confrontational "challenge" as the main therapeutic stance, with the group used to break down excuses.
Progress judged by shown remorse or tears, or by how detailed and "honest" the account is, instead of by change on assessed risk factors.
One fixed, long dose for everyone regardless of assessed risk and need. Core grew to about 180 hours. [source: Mews et al. 2017]
Completion presented as risk reduction in parole, MAPPA or court reports.

4.2 Phrases and rationales to notice. These are illustrative of the old framing, not quotations from manuals: "He needs to take full responsibility before we can work with him"; "He's still in denial, so he's untreatable"; "Until he feels what the victim felt he's a risk"; "He completed SOTP, so his risk is reduced." [synthesis]

4.3 Rebrandings. Offence-focused group programmes with victim-empathy and full-disclosure modules may carry other names in community, private, faith-based or overseas settings. Ask what is actually done in sessions; the label tells you little. Stanbury reports a psychologist's view that Kaizen reuses some discredited techniques. That is one critic's opinion and has not been verified. [source: Stanbury 2017] [uncertain]

4.4 Red flags in a colleague's or service's practice. [synthesis]

Group members routinely hear each other's detailed offence accounts, with no rationale tied to individual risk.
Re-enactment or role play in which the client plays his victim.
Programme reports gate parole or progression with no outcome data for the programme.
People who maintain innocence are excluded from all risk work.
Shaming language: "monster", "predator", or "a sex offender" used as an identity.
Staff are not supervised for exposure to offence material.
The service has no individual sessions.

4.5 How legitimate current work differs.

Targets. Current programmes target assessed risk and need domains: problem-solving, self-regulation, relationships, and sexual attitudes and behaviours. They do not target offence-account accuracy or victim empathy for its own sake. [source: Wilkinson & Powis 2019]
Orientation. Horizon is described as strengths-based and future-focused, built on RNR, the Good Lives Model and desistance research. [source: Wilkinson & Powis 2019]
Offence discussion. Discussing the offence is limited and serves a purpose. The aim is to understand one's own risk factors, not to give a statement-matched account in front of a group. [source: The Detail 2017 on the 2015 changes] Kaizen participants valued being able to discuss offence details without judgement, and sharing was hardest at the start. [source: Valois, PSJ 275]
Individual work. Kaizen mixes individual and group sessions, partly in response to the MoJ findings. [source: Valois, PSJ 275]
Shame. Shame is treated as a clinical target, not used as a tool. [source: Wilkinson & Powis 2019] [synthesis]

5. What to do instead and how to help those harmed

5.1 Current recommended alternatives (HMPPS, England and Wales).

Horizon. Accredited in 2015 and introduced in March 2017. It is for men with a sexual conviction assessed as medium risk or above. It can be delivered in custody or the community and runs for about 60 hours, against 120 for the previous programme. It is strengths-based, draws on RNR, Good Lives and desistance work, and is suitable for men who maintain innocence or minimise. [source: The Detail 2017; Wilkinson & Powis 2019]
Kaizen. For high and very high-risk men; it replaced Extended SOTP. It is delivered individually or in rolling groups of up to eight, with a sexual-offending strand alongside intimate-partner and general violence strands. It targets four risk domains plus a desistance domain, sense of purpose. It uses Good Lives, RNR and desistance principles and accepts men who maintain innocence. [source: Valois, PSJ 275; The Detail 2017; Stanbury 2017]
Healthy Sex Programme. One-to-one work for men with a sexual conviction or an offence with a sexual element. [source: gov.uk HMPPS accredited programmes page]
Other programmes. iHorizon (internet offending), Becoming New Me+ and New Me Strengths (adapted for learning disability and challenges) are also named in HMPPS listings. The gov.uk accredited programmes page recorded on 12 May 2026 that Horizon, Kaizen, Becoming New Me+, New Me Strengths and Living as New Me had been removed from that listing, for reasons not stated. Check the current HMPPS programme suite before relying on any of these names. [source: gov.uk page] [uncertain]

5.2 Assessing a client who went through Core or Extended SOTP. [synthesis unless tagged]

Ask what the course involved for them: disclosure, role plays, group challenge, and what the post-programme report said. Ask how it affected their sense of self, hope, shame and trust in professionals.
Screen for shame-related depression, suicidality and self-harm using standard risk-assessment practice. Screen for trauma symptoms if they took part in victim role play or have their own abuse history.
Ask about their reaction to the 2017 findings. Some men felt seven years were wasted, or feared completion now counted against them. [source: Ievins 2017]
Ask whether they made contacts in the group linked to offending, or met material or sources through the group, framed as a risk-management question and not an accusation. [source: Mews et al. 2017 mechanism]
Use a current, validated risk assessment by appropriately qualified assessors. Do not treat SOTP completion as protective.

5.3 Treatment considerations. [synthesis]

Do not repeat statement-matched offence accounts or victim role play.
Work collaboratively on dynamic risk factors and approach goals, using a Good Lives or desistance framing.
Address shame directly, for example with compassion-focused or other shame-focused approaches, while keeping responsibility for behaviour clear.
Rebuild the alliance with someone who may associate psychology with coercion: be explicit about the limits of confidentiality and about what reports will say.
For a client who maintains innocence, offer risk-relevant work that does not depend on admission.
Treat co-occurring problems: substance use, mental illness, intellectual disability.
Refer to current accredited programmes or forensic specialist services where indicated.

5.4 Ethical, reporting and legal duties. [synthesis] Check your own regulator's and jurisdiction's rules.

Safeguarding comes first. Disclosures of current or planned harm to children or adults at risk, or of contact with offending networks, may require breaching confidentiality and informing police, MAPPA or children's or adult social care. Tell clients these limits at the outset.
Report writing. Parole, court and MAPPA reports should not present Core SOTP completion as evidence of reduced risk, given the MoJ findings. They should not count non-completion against a man whose refusal or exclusion reflected the programme's own rules, such as maintaining innocence. [source: Mews et al. 2017; Stanbury 2017]
Professional practice. Follow HCPC and BPS standards (or your regulator's) on evidence-based practice, informed consent and avoiding harm. Raise concerns through supervision and governance if a service still uses re-enactment, shaming or compelled-disclosure methods.
Legal routes. Men who believe their progression was blocked by SOTP requirements may want independent legal advice. Stanbury doubts that compensation claims would succeed. [source: Stanbury 2017]
Staff welfare. Provide supervision and support for clinicians exposed to offence material. [source: Big Issue North 2022]

Sources

Sources: