Source · Prevention of Future Deaths

Sundeep Ghuman

Ref: 2025-0625 Date: 15 Dec 2025 Coroner: David Manknell Area: London Inner South Responses identified: 1 / 2 View PDF

Systemic misunderstanding of the CSRA policy for prisoners posing racist risks led to incorrect 'standard risk' categorization instead of 'high risk,' reflecting a significant training and operational failure.

Date 15 Dec 2025

56-day deadline 9 Feb 2026

Responses identified 1 of 2

State Custody related deaths

Coroner's concerns

AI summary

View full coroner's concerns

The following matters are addressed both to the Secretary of State for Justice, and the Governor of HMP Belmarsh. 1) The operation of the CSRA process, and application of the CSRA policy at HMP Belmarsh and more widely in the prison service HMPPS provided a witness who holds a senior position in the safety group for HMPPS and is responsible for various policies and processes including the CSRA policy (“the policy witness”). His evidence was that the CSRA policy is to the effect that if a prisoner poses a risk because of racism, that they should be categorised as “high risk”, but if appropriate that they could be permitted to share with a prisoner of their own ethnicity. He confirmed that this had been the policy position in 2019 and remains the position now. The policy witness was not aware of the “S1” system operated at HMP Belmarsh (where such prisoners would be categorised as “standard risk” with a note to say that they can share with their own ethnicity) until he was preparing for his evidence in this case. He also emphasised that the policy is intended to enshrine a cautious approach to risk. By contrast, none of the witnesses from HMP Belmarsh (or HMP High Down) were aware that a prisoner who posed a risk because of racism should be high-risk, and all were under the impression that such a prisoner should be standard risk, with an “S1” label indicating that they should only share with their own ethnicity. This was not only the view of the prison staff responsible for the CSRA process, but also the evidence of the former Head of Residence and Services (now Head of Business Assurance), and the Head of Security and Intelligence, whose evidence was that he was unaware of the correct position until shortly before giving his evidence. The evidence of those staff was that HMP Belmarsh had operated in this way for many years. Nor were staff at HMP Belmarsh aware of the policy position that the guiding principle is one of caution. Rather, they emphasised the need to maximise occupancy and the risk of prisoners causing occupancy to be lost by falsely claiming to be racist. The misunderstanding of the need for such prisoners to be categorised as high risk creates itself a significant risk of fatal events such as in this case, because a standard risk prisoner (whether noted as S1 or not) is not noted in the same way on the electronic system, and it is not obvious to staff who subsequently make decisions in respect of that prisoner that the individual indeed presents a high risk. In addition, the lack of understanding of the policy by all staff (including senior management) at HMP Belmarsh appears to reflect a systemic failing of training and operational understanding, and a disconnect between those responsible for creating and maintaining the policy and those who take operational decisions within prisons. The fact that the comprehensive misunderstanding of the policy by HMP Belmarsh over a period of many years was entirely unknown to those responsible for the policy, even in circumstances where HMP Belmarsh operated its own system (complete with bespoke “S1” stickers) and not picked up by any audit or monitoring procedure also indicates a failure in policy implementation. The overall widespread and longstanding failure in policy implementation by HMP Belmarsh creates a risk of future deaths as a result of racist violence. If policies designed to prevent these events are fundamentally misunderstood and are incorrectly implemented to this extent, there is a real risk of departure from the minimum standards set in policy and specifically designed to reduce this very risk.
2) A lack of either guidance or operational understanding of how NOMIS alerts for racism should be treated by staff. The evidence of staff at HMP Belmarsh (and HMP High Down) was highly inconsistent as to how an active alert for racism would be treated. Evidence of some officers was that if an alert was currently marked as “active” then they should assume that there was a real risk of danger posed by racism, or require further investigation and have the active status removed if no longer appropriate. This appeared to be the approach endorsed by the policy witness, on that basis that an officer seeing an alert will not have the benefit of the underlying intelligence, or knowing what led to the alert being made and maintained as active. The majority of officers from HMP Belmarsh who gave evidence on this issue however, expressed the view that it was a matter for them as to what weight they should give an active alert for racism. Their evidence was that a racist alert might be old, that the alert was only part of the picture, and that they would give at least equal weight to what was said by the prisoner upon arrival at the prison. In the case of [“M”], the evidence of the relevant officer (who carried out the secondary CSRA in reception) was that he was aware of the alert, but because it was old, and because there was no concern apparent when he spoke to [“M”], that the alert could be disregarded. No officer recalled any specific training as to how an active alert for racism should be approached when conducting a CSRA. The lack of understanding of how an active alert for racism should be approached when assessing suitability cell sharing creates a risk of future fatal events. An unstructured approach and lack of training creates a risk that, as in this case, staff may inappropriately disregard an active alert for racism, leading to potentially fatal racist violence.
3) The lack of consideration of risks, beyond the minimum requirements of the CSRA, when allocating prisoners to share a cell a) As to the risk of violence: The evidence of HMPPS staff and officials was that the threshold for making someone High Risk on a CSRA as a result of the risk of them being violent is a very high threshold. The Head of Security and Intelligence’s evidence was that, in general, a prisoner would only be made high risk for violence on the CSRA if they had either previously killed a cellmate or they had previously inflicted life-threatening injuries on a cellmate. Lesser violence, even against other prisoners and cellmates would not lead to them being High Risk. The finding of weapons in cells was not regarded as relevant unless there was evidence that they had been used. Similarly, the threshold for someone being regarded as “vulnerable” for the purposes of the CSRA was said to be very high. The evidence of the Head of Security and Intelligence was that this was in effect limited to prisoners with significant disabilities including serious learning difficulties. The policy witness’s evidence was that the CSRA policy only provides a “baseline” and he would expect measures to be in place to allow judgements to be made beyond this that minimise any problems caused by cell sharing. His evidence was that he would expect staff to be exercising judgement beyond what is stated on the formal CSRA, although the extent to which they are able to do that will be dependent on a range of things, including what information they have got about the individuals, what the circumstances are for the move, what the capacity pressures may be. The evidence of staff at HMP Belmarsh (including from the Head of Security and Intelligence) was by contrast to the effect that moves into a cell are simply a daily occurrence and that as long as a CSRA says the individuals concerned can share then they can be moved in, and that no further consideration is necessary or appropriate. Consistently with this, the evidence of officers was that when arranging a cell move they would not look at the NOMIS notes of a prisoner, or other records, but solely at the CSRA to check that they were not High Risk (or “S1”). In this case, the officers confirmed that when deciding to move [“M”] and his cousin into Mr Ghuman’s cell, no consideration was given to the appropriateness of the move save for checking that the CSRA permitted sharing. This approach creates a substantial risk that a prisoner who is prone to violence against his cell mate will be placed with a prisoner who is at risk of such violence. The fact that the violent prisoner may not have previously killed a cellmate or inflicted life-threatening injuries on a cellmate in the past is not a sufficient safeguard against a future cellmate being subject to a level of violence that may result in death. This risk is heightened if the other prisoner has a level of vulnerability, or some other feature that makes violence against them more likely, including any relevant prejudice. A more nuanced approach appears to be required in respect of individual cell sharing decisions, and this appears to be the intention of the policy, but not reflected in implementation at HMP Belmarsh. b) Drug use The evidence from all witnesses was that use of illicit drugs in prison (including Spice) had no bearing on cell sharing suitability either in the CSRA or otherwise. Although the effect of the drug Spice was not a factor in Mr Ghuman’s death, the evidence that emerged during the inquest on this issue as it relates to cell sharing is an issue that gives rise to a concern of future deaths. The evidence of multiple witnesses from staff was that Spice use is widespread in the Prison, and that officers are well aware of prisoners that use it in their cells. It was also apparent that the secondhand or passive effects of Spice can be dramatic and seriously harmful. One of the witnesses, an Officer at HMP Belmarsh, described how having been exposed to secondhand inhalation from outside a cell for a very short period, he was unable to stand, and remained unwell and disoriented, requiring rest and supervision overnight as a result. The evidence was that Spice use can make prisoners very unwell, and can also make them act irrationally and in a way that has no regard for their own safety. There was evidence in this case that prisoners who were Spice dealers would persuade young and easily exploited prisoners to try Spice, including by falsely claiming that it was a regular roll-up cigarette, in order to “test” a batch of the drug or simply to take pleasure in seeing them suffer the effects. There appears to be no consideration in HMPPS of the risks to prisoners from inhalation of Spice when a cellmate is a user or supplier, or of the risks of more easily exploited prisoners being led into Spice use by cellmates who use the drug, and no assessment of whether it is safe to place a prisoner in a cell with a user or supplier of Spice. I am aware that deaths have been directly linked to Spice use. In the absence of any assessment of the risks in individual cases, there appears to be a risk of death of a prisoner through being placed into a cell with a known user or supplier of Spice.
4) In respect of HMP Belmarsh more generally, a concern for the safety of prisoners given the evidence of such widespread and seemingly ubiquitous violence and drug use in the prison. There was evidence of concern in respect of the widespread levels of prisoner-on-prisoner violence at HMP Belmarsh, and the widespread use of drugs, especially Spice, and the evidence of staff that this was unavoidable. The evidence was clear and consistent about both of these issues, with a number of witnesses effectively giving evidence that “lots of prisoners here” or “most prisoners at Belmarsh” are violent or have a violent history, and that it was therefore accepted that there will be regular incidents of violence by prisoners against other prisoners. Similarly, the evidence about Spice use was that it is known to be widespread in the Prison, and that a very significant proportion of prisoners are Spice users. However, that evidence on these topics did not correspond to recorded findings in Adjudications or criminal proceedings. In [“M”]’s case there were multiple instances of violence (and indeed racism) being identified in intelligence reports and in some cases on NOMIS, but without matters proceeding to an adjudication. Similarly, although it was apparently well-known that [“M”] and his cousin were regular Spice users, no formal proceedings for drug possession were taken against them. The evidence overall was that HMP Belmarsh has widespread levels of violence by prisoners against other prisoners, and seemingly ubiquitous levels of Spice use in the prison. This seems to be at a level where staff do not consider that they can always take robust action and formal measures when this occurs. In those circumstances, I have a concern that the Prison may not currently be capable of providing a safe and secure environment for prisoners accommodated there, and that there is a risk of future deaths from drug use or violence.
6. ACTION SHOULD BE TAKEN In my opinion action should be taken to prevent future deaths and I believe you and your organisation have the power to take such action.

Responses

1 respondent

HM Prison Probation Service Central Government

9 Mar 2026 PDF

Action Taken

HMP Belmarsh has withdrawn the S1 system for cell sharing risk assessment and reviewed all prisoners under the previous system, updated their risk level to be in line with national policy. HMPPS is updating the CSRA policy and naloxone is now available across all residential units. (AI summary)

View full response

Dear Mr Manknell KC

REGULATION 28 REPORT TO PREVENT FUTURE DEATHS: MR SUNDEEP GHUMAN

Thank you for your Regulation 28 report of 15 December 2025 following the inquest into the death of Sundeep Ghuman at HMP Belmarsh on 19 February 2020.

I know that you will share a copy of this response with Mr Ghuman’s family, and I would first like to express my condolences for their loss. Every death in custody is a tragedy and the safety of those in our care is my absolute priority.

You have raised several concerns relating to the operation of the cell sharing risk assessment (CSRA) process at HMP Belmarsh and the application of the policy at both a local and national level.

You have identified that practice at HMP High Down and HMP Belmarsh was not in line with the CSRA policy that is set out in PSI 20/2015 The Cell Sharing Risk Assessment. HMP Belmarsh has confirmed that it has now withdrawn the S1 system and is operating fully in line with national policy. All prisoners previously categorised under S1 have been reviewed and allocated to the appropriate level in accordance with the relevant guidance. In addition, individuals identified with a racist marker are now designated as ‘high risk share’ to ensure full compliance with national standards. HMP High Down has also confirmed that it now complies with the policy and uses the correct descriptor for the risk level of such prisoners.

HMPPS is currently updating the CSRA policy, which will be issued as a policy framework that will supersede PSI 20/2015. There are two improvements that we intend to make which address the concerns you have raised. First, the forms in the current PSI contain only two options, ‘high risk’ and ‘standard risk’. Where a prisoner is high risk but only to certain people (as “M” was), the details of those at risk will be set out separately and then entered into the

electronic record in separate places. We will do this by including three options: ‘high risk general’ (cannot share with anyone); ‘high risk specific’ (cannot share with certain people or groups); and ‘standard risk’ (can share with anyone). Where a prisoner is high risk specific the details of those at risk, with whom he or she cannot share, will have to be set out clearly next to that result, where they cannot be missed. The electronic record will be amended to match this. Second, the policy framework will instruct prisons not to use local terms or abbreviations to describe CSRA results. The framework will also contain a statement that any elevated risk, even to one person, makes a prisoner high risk, and that standard risk is not an option in that situation.

Another improvement in the new framework will be to require all prisons to put in place a quality assurance process. This will involve checking samples of initial assessments to see that all available evidence has been taken into account, and that the results are based on that evidence. It will also cover how CSRA is used in practice, to check whether prisoners’ cell allocations match their risk level. In addition, independent audits of the CSRA process by our Performance, Assurance and Risk Group will continue.

You suggested that training could be improved to strengthen operational understanding. Alongside the new policy framework we are redesigning the CSRA training materials and these will address how NOMIS alerts for racism should be treated by staff. When implementing the framework, we plan to give prisons enough time to ensure a core group of staff will receive the new training and be able to reinforce its content with colleagues before they are required to comply with it.

In the meantime, all managers at HMP Belmarsh are currently completing the mandated CSRA training as part of scheduled training days. In addition, all staff are required to undertake the CSRA e-learning package to support improvements in assessment quality. Completion of these training requirements is being monitored by the Head of Business Assurance, who holds overall responsibility for the establishment’s training programme.

The Long-Term and High Security Estate (LTHSE) Safety Team will also deliver a CSRA upskilling session at a future Heads of Safety Meeting to ensure consistent, policy-compliant application of CSRAs across all LTHSE prisons. It will emphasise that a cautious approach should be taken so that decisions default to the safest option when risk is uncertain or information is incomplete.

As well as improving training, quality assurance measures have been formally embedded within HMP Belmarsh’s internal audit processes. These arrangements ensure that decisions and practice remain compliant with national policy and are subject to regular review. Additionally, the LTHSE Safety Team will incorporate CSRA checks during safety assurance visits to provide the Prison Group Directors (PGDs) with assurance that prisons are policy compliant in CSRA application. This will include reviewing a sample of 10% of CSRAs completed in the last 60 days (on reception, transfer, and following incidents) for process compliance and quality. CSRA forms will be checked to ensure there is a clear rationale for

decisions about risk levels and that all data sources used are recorded. Any CSRA found to be non-compliant or poor quality will be escalated to the Head of Safety, Deputy Governor and Governor, with a recommendation of improvement actions. If improvement actions are required, re-checks will be completed within 30 days. If assurance gaps still exist, this will be escalated and reported to the PGD within five working days for Senior Management Team oversight and targeted support.

You have raised concerns about an inconsistent response to alerts about racism. The expectation of staff conducting CSRAs is that they will exercise caution and such alerts will be taken seriously. Where time is available the basis for them will need to be explored further and a decision not to act in accordance with them may be appropriate. But such a decision must not be reached solely on the basis of what the prisoner says. And where there is doubt and/or where there is no time to explore further, the alert will need to be taken at face value and the individual made high risk (cannot share with a particular group) until further investigation can take place. Clarification of this point will be included in the training that will be provided for the roll out of the new policy framework, and compliance will be checked through the new quality assurance process described above.

Some of the problems that arise in this situation are the result of the presence of an excessive number of alerts on NOMIS, caused by the absence of any consistent process for reviewing, and where appropriate deactivating, old alerts. There are broader plans to address this which are currently on hold pending other work, including the transitioning of information management from the legacy NOMIS system to the modern Digital Prison Services platform (which will improve accuracy, accessibility and operational efficiency generally, as well as bring specific benefits in terms of the way in which CSRAs are recorded). These plans will be taken forward at the earliest opportunity with a view to reducing the number of inappropriate alerts and thereby making the task of staff making decisions on the basis of such alerts much more straightforward.

You have observed that during the allocation of prisoners to shared cells there is scope to strengthen the consideration of risks beyond the minimum CSRA requirements. You suggested this may be particularly relevant in relation to risks associated with violence and drug use.

The CSRA is the process for managing the risk of serious in-cell violence, and decisions about cell sharing should always be made in accordance with the prisoner’s CSRA. However, this is not the only consideration in such decisions. Whilst not something that is set out in a formal policy (because it would be impossible to describe all the various factors that may be relevant or all the different contexts in which such decisions may need to take place) there is an expectation that staff will exercise their judgement when allocating cells with a view to anticipating and minimising any conflict and maximising prisoner welfare. The extent to which this is possible will depend on the information available about the individuals and the circumstances in which the decision is being taken. But it is a potentially important way in which staff can exercise control over the residential environment. As well as prisoner

preferences (which it may not always be appropriate to follow for various reasons), these decisions will be influenced by what is known about the individuals in a more general sense than the narrow range of factors that is considered in the CSRA.

This may include information about the potential for, or vulnerability to, intimidation or manipulation, and any suspected substance misuse. Depending on the circumstances such information may affect decisions about cell sharing in a number of different ways. As you point out, the risk of a suspected drug user coercing another (previously non-using) prisoner into drug use must be considered and may lead to a decision that sharing between two individuals is not appropriate. However, there may be instances in which the risk to a suspected drug user could be mitigated by sharing with a non-user who may be able to raise the alarm in the event of an overdose. These are therefore difficult and finely balanced decisions that must be made on the basis of the details of the individual case.

Drug misuse is not, of itself, a reason to change a prisoner’s CSRA. Not all individuals who use drugs become more violent, and staff are instructed to remain alert to any behavioural changes that may indicate an increased risk of violence or harm to others, regardless of the underlying cause. The forthcoming policy framework will reinforce this message and will emphasise the importance of acting quickly when a prisoner is seen to be more violent, including making them temporarily high risk while the situation is reviewed.

You have raised also the risk associated with secondary exposure to psychoactive substances. This is a matter that we have also considered and based on the evidence available, we do not believe this currently presents a significant risk.

Testing has been carried out to assess the possibility of secondary exposure, focusing on staff. This included post‑shift urine testing and personal and residential air sampling. No psychoactive substances were detected in the urine tests or personal air samples, and only trace levels of contaminants were found in residential air samples.

From this, the conclusion is that secondary exposure is highly unlikely to result in any adverse effects. We recognise that the composition of psychoactive substances in prisons continues to change, and we are not complacent. However, at present we do not consider secondary exposure to pose a risk to the cell mates of those using psychoactive substances.

Finally, you have raised concerns about the levels of violence and drug use at HMP Belmarsh and its impact on the safety of prisoners.

Reducing violence in prisons is a key priority, and we are working hard to make prisons as safe as possible. We know that prisoners in overcrowded cells are 19% more likely to be involved in an assault, and we are tackling overcrowding and aim to build 14,000 new prison places by 2031.

Since Mr Ghuman’s death the operational capacity at HMP Belmarsh has been reduced by 136 places. This has enabled the removal of triple-occupancy cells, resulting in improved decency standards and enhanced safety across the residential accommodation.

A number of measures to strengthen safety, reduce violence, and improve the management of prisoner movements have been implemented. A cohorted regime is now in place, ensuring prisoners are managed in fixed groups to minimise unnecessary mixing across residential units. All internal movements are subject to risk assessment to prevent contact between individuals who may present a threat to one another. This approach includes separate internal movement routes, visits scheduled strictly by residential unit, adjustments to corporate worship groups, and cohorted access to activities such as work and education.

First Night Centre processes have also been strengthened and a robust assessment procedure is now in place for all individuals entering custody, whether via court or transfer. These assessments ensure that information is current and accurate, that any risks to others are clearly identified, and that critical information is promptly shared with all relevant stakeholders. The First Night Centre now works closely with the local Safety team, which further improves the flow of information and early risk identification.

Additionally, security and conflict management arrangements have been enhanced through the establishment of a dedicated Conflict and Gang Team. This team is responsible for investigating and managing issues that could lead to violence, and ensures that prisoners are only moved to locations where they can be safely accommodated.

The Safety Intervention Meeting has also been reinforced and now takes place regularly with consistent attendance from key stakeholders. This forum enables timely information sharing and supports effective decision-making around individuals who may require additional monitoring or intervention.

We also continue to work closely with health partners to address substance misuse. This includes the implementation of Incentivised Substance Free Living Units in 88 prisons, including HMP Belmarsh, where prisoners commit to behavioural compacts, regular drug testing, and access to enhanced opportunities. We have embedded 54 Drug Strategy Leads in key establishments, and appointed 17 Group Drug and Alcohol Leads who are now embedded to provide strategic oversight. The Adult Health, Care and Wellbeing Core Capabilities Framework was introduced in May 2025, and through the Enable Programme, MoJ, HMPPS and NHS England are accelerating specialist training on drugs, alcohol dependency, and trauma-informed care. Additionally, a comprehensive redesign of Foundation Training for new prison officers is underway, including mandatory modules on drug and alcohol misuse for all staff.

Supported by these national initiatives, HMP Belmarsh has reviewed and implemented a number of measures to strengthen drug-misuse management, recovery support and overall safety within the establishment. A comprehensive Drug Strategy is embedded across the

prison, focusing on both the supply and demand of illicit substances, while also maintaining a strong emphasis on support, detoxification, and long-term recovery.

A key development is the implementation of an Incentivised Substance-Free Living (ISFL) unit. This dedicated wing is designed to promote recovery and abstinence, with prisoners signing a formal compact committing to monthly drug testing. Those engaging with the ISFL scheme are able to access additional incentives, including extended domestic periods, increased gym access, and enhanced wing-level provisions. This approach has had a positive impact on stability across the prison and has contributed to reductions in drug misuse.

Mandatory Drug Testing (MDT) continues to operate robustly at HMP Belmarsh, with 5% of the population subject to random testing each month. Alongside this, enforcement and support measures are in place to manage the risks associated with substance misuse. Prisoners found under the influence or in possession of illicit substances are subject to adjudication and appropriate disciplinary action, while structured support is provided by Phoenix Futures and keyworker-led interventions. Monthly multi-disciplinary reviews ensure that care and recovery plans are individually tailored.

Governance arrangements have also been strengthened. A dedicated Drug Strategy Lead oversees the coordination of activity across security, healthcare, residential units, and external partners. Monthly multi-disciplinary meetings are held to monitor trends, review incidents, and manage emerging risks. In addition, recovery and rehabilitation provision remains a central focus. The partnership with Phoenix Futures continues to deliver treatment and harm-reduction support. Lived Experience courses and staff training are in place to improve prisoner engagement and ensure continuity of care on release. These measures have supported high levels of community recovery engagement post-custody.

In order to improve the response to an opioid overdose, naloxone is now available across all residential units, and staff have been trained to administer it.

Thank you again for bringing your concerns to my attention. I trust that this response provides assurance that action is being taken to address the matters raised.

Report sections

Investigation and inquest

On 7 May 2020 an inquest was opened into the death of Mr Sundeep Ghuman and adjourned pending Police investigation. Following the Police investigation a criminal trial took place, and in 2022 one of Mr Ghuman’s cellmates, [“M”], was found guilty of murder. Following the resumption of the inquest, the final hearing took place from 20 October 2025 until 7 November 2025. The medical cause of death was 1a. Head injury. In section 3 of the Record of Inquest the jury recorded the following: “At the time of his death Sundeep Ghuman was a prisoner at HMP Belmarsh on House Block two. Sundeep was in a triple cell and on 7th February 2020 two prisoners were placed in the cell, 'M' and his cousin 'T'. 'M' was a racist. On 18th February 2020 between 17:24 and 17:30 one of the prisoners , 'M' violently assaulted Sundeep with a table leg causing unsurvivable head injuries. Prison Officers attended the cell and called the emergency services at 17:32. Sundeep was taken to the Royal London Hospital, he arrived at 19:52 and was met by the Trauma Team. Sundeep died on 19th February 2020 at 18:10 in the Royal London Hospital. The cause of Sundeep's death was from a head injury.” The jury’s conclusion at the inquest was a narrative conclusion. The conclusion in section 4 of the Record of Inquest was as follows: “(1) Sundeep Ghuman was unlawfully killed by a cell mate, ‘M’ while a prisoner at HMP Belmarsh. (2) (a) The cell sharing risk assessment, (C.S.R.A) of M at HMP Belmarsh was not carried out appropriately by the prison on 5-6 November 2019 due to the following reasons: - The P.S.I 2015 [Prison Service Instruction (PSI) 20/2015] provided that if evidence is found of racially motivated offence of reports in community or custody that is a strong indicator the prisoner should be high risk but can share with own ethnicity. Racist alert was active and was evidence of racism and should lead to high risk can share with own ethnicity unless further sources checked, e.g. intelligence and the rationale should then be recorded in the CSRA. No evidence of this was undertaken by the prison officer completing the reception assessment. Previous CSRA from [HMP] Highdown stated, ‘standard risk but racist share with own ethnicity and corresponding NOMIS entry, confirmed this and this was not accurately reflected in Belmarsh CSRA reception assessment, which only references standard risk, transferred in from Highdown. (2)(b) If ‘M’s’ CSRA had been carried out appropriately on 5th November 2019 that would have avoided ‘M’ being placed in Sundeep’s cell and Sundeep’s death would have been avoided. (3)(a) In requiring Sundeep to share a cell with ‘M’ there were failings in the assessment by the prison service in their assessment of risk to Sundeep as a result of the risk presented by ‘M’s’ racism as the CSRA was not carried out appropriately and this materially contributed to his death as absent the failing it is likely that Sundeep would not have been in a cell with ‘M’ on 18th February 2020. (3)(c) In requiring Sundeep to share a cell with ‘M’, there were failings by the prison in their assessment of a result of any additional risk associated with the use of a triple rather than a double cell as additional potential risks such as a greater number of people means more scope for tension and any potential imbalance in relationships amongst the three prisoners was not considered. However the evidence does not support that prison officers were expected to consider such additional risks with the use of triple cells, the P.S.I did not cater for the use of triple cells. We cannot say that absent the failing, Sundeep would probably not have been in a cell with ‘M’ on 18th February 2020 or that it is reasonably possible that they would have not been in a cell together. (4)(a) There was a failing by the prison that made a material contribution to Sundeep’s death due to the lack of efficiency of the CSRA process, including the training of prison officers. The prison operated the ‘S1’ system which demonstrates CSRA PSI process not properly understood and applied. On balance, prison officers had not recorded adequate training on CSRA process other than through shadowing (main for of training) and those officers providing training through shadowing did not understand the requirements of the P.S.I. the risk was heightened as the CSRA process did not require the second day check or any other assurance of the reception assessment leading to a single point of failure.”

Circumstances of the death

Mr Sundeep Ghuman was a prisoner at HMP Belmarsh where, at the material time, he was sharing a cell with two other prisoners, [“M”], and another prisoner who was a cousin of [“M”]. Mr Ghuman was a British Asian man [“M”] had a history of racist and violent behaviour. There were active NOMIS alerts (NOMIS being the National Offender Management Information System, an operational database used in prisons in England and Wales at the time to manage offenders) for [“M”] in respect of racism and violence. The alert for racism had been added some years before on the basis that [“M”] was a member of a racist gang, the Racist Army of Woolwich and that they carried out racist assaults in south London. There were further intelligence reports thereafter in respect of racist behaviour in custody by [“M”] and involvement in incidents and threats of violence. The evidence from HMPPS policy officials was that a prisoner who posed a risk by reason of racism should in fact be marked as “high risk” on a CSRA, but permitted to share with his own ethnicity. On arrival at HMP High Down on 8 October 2019, the CSRA for [“M”] had recorded that he was “standard risk” but that due to his racism, [“M”] was to share with prisoners of his own ethnicity only. [“M”] was then moved to HMP Belmarsh on 5 November 2019 where reception staff did not impose the same restriction, but marked him as “standard risk as per High Down” without any qualification. They did not apply the “S1” marker used at HMP Belmarsh to indicate a racist prisoner who is only suitable to share with his own ethnicity. On 7 February 2020, staff at HMP Belmarsh moved [“M”] and his cousin from the cell and houseblock where they had been residing due to their disruptive behaviour (including “self isolating” due to drug debt to other prisoners) and they were placed into Mr Ghuman’s cell. No consideration or assessment of their suitability to share together was made by officers in charge of the move. No consideration was given to the information on NOMIS or the intelligence reports in respect of [“M”], and staff involved were unaware of his alert for racism or apparent history of involvement in racist violence. On 18 February 2020, Mr Ghuman was violently assaulted in his cell with a table leg by [“M”]. Mr Ghuman was taken to the Royal London Hospital, where he died in hospital of a head injury on 19 February 2020. On 16 May 2022 [“M”] was found guilty of Mr Ghuman’s murder and later sentenced to life in prison.

Copies sent to

Oxleas NHS Foundation TrustPractice Plus GroupCGL (‘Change Grow Live’)XXXXXXXXXX, HMPPSXXXXXXXXXX, HMP BelmarshXXXXXXXXXX, HMP Belmarsh

Inquest conclusion

“(1) Sundeep Ghuman was unlawfully killed by a cell mate, ‘M’ while a prisoner at HMP Belmarsh. (2) (a) The cell sharing risk assessment, (C.S.R.A) of M at HMP Belmarsh was not carried out appropriately by the prison on 5-6 November 2019 due to the following reasons: - The P.S.I 2015 [Prison Service Instruction (PSI) 20/2015] provided that if evidence is found of racially motivated offence of reports in community or custody that is a strong indicator the prisoner should be high risk but can share with own ethnicity. Racist alert was active and was evidence of racism and should lead to high risk can share with own ethnicity unless further sources checked, e.g. intelligence and the rationale should then be recorded in the CSRA. No evidence of this was undertaken by the prison officer completing the reception assessment. Previous CSRA from [HMP] Highdown stated, ‘standard risk but racist share with own ethnicity and corresponding NOMIS entry, confirmed this and this was not accurately reflected in Belmarsh CSRA reception assessment, which only references standard risk, transferred in from Highdown. (2)(b) If ‘M’s’ CSRA had been carried out appropriately on 5th November 2019 that would have avoided ‘M’ being placed in Sundeep’s cell and Sundeep’s death would have been avoided. (3)(a) In requiring Sundeep to share a cell with ‘M’ there were failings in the assessment by the prison service in their assessment of risk to Sundeep as a result of the risk presented by ‘M’s’ racism as the CSRA was not carried out appropriately and this materially contributed to his death as absent the failing it is likely that Sundeep would not have been in a cell with ‘M’ on 18th February 2020. (3)(c) In requiring Sundeep to share a cell with ‘M’, there were failings by the prison in their assessment of a result of any additional risk associated with the use of a triple rather than a double cell as additional potential risks such as a greater number of people means more scope for tension and any potential imbalance in relationships amongst the three prisoners was not considered. However the evidence does not support that prison officers were expected to consider such additional risks with the use of triple cells, the P.S.I did not cater for the use of triple cells. We cannot say that absent the failing, Sundeep would probably not have been in a cell with ‘M’ on 18th February 2020 or that it is reasonably possible that they would have not been in a cell together. (4)(a) There was a failing by the prison that made a material contribution to Sundeep’s death due to the lack of efficiency of the CSRA process, including the training of prison officers. The prison operated the ‘S1’ system which demonstrates CSRA PSI process not properly understood and applied. On balance, prison officers had not recorded adequate training on CSRA process other than through shadowing (main for of training) and those officers providing training through shadowing did not understand the requirements of the P.S.I. the risk was heightened as the CSRA process did not require the second day check or any other assurance of the reception assessment leading to a single point of failure.”

Similar PFD reports

Shared signals

Related inquiry recommendations

Report details

Reference: 2025-0625
Date of report: 15 December 2025
Coroner: David Manknell
Coroner area: London Inner South

Responses identified

Responses identified 1 of 2

1 response not yet linked

Organisations named in PFD reports are normally expected to respond within 56 days. Deadline: 9 Feb 2026.

Sent to

HMP Belmarsh

Ministry of Justice

Source links

Judiciary.uk page → Report PDF →

Data: judiciary.uk under OGL.