Completed
· Chair: Mr Justice Keith
· Established September 2004
· Commissioned by Home Office
Public inquiry into the murder of 19-year-old Zahid Mubarek at Feltham Young Offender Institution on 21 March 2000 by his racist cellmate Robert Stewart.
More about this inquiry
Public inquiry into the murder of 19-year-old Zahid Mubarek at Feltham Young Offender Institution on 21 March 2000 by his racist cellmate Robert Stewart. Justice Keith found the death entirely preventable and identified multiple systemic failures in prisoner risk assessment, allocation and monitoring. Made 88 recommendations to improve prison safety, risk assessment procedures and staff training.
The elimination of enforced cell-sharing should remain the objective of the Prison Service, and the achievement of this goal should be regarded as a high priority.
The Prison Service should review whether the resources currently available to it might be better deployed towards achieving this goal, without compromising standards in other areas, and should set a date for realising this objective.
If the resources currently available to the Prison Service are insufficient to produce a significant decrease in enforced cell-sharing, central government should allocate further funds to the Prison Service to enable more prisoners to be accommodated in cells on their own.
The Prison Service should retain its practice of placing prisoners who are at risk of suicide or self-harm, but who are not so vulnerable as to require being on suicide watch at all times or accommodated in a safer cell, in a cell with another prisoner who they can talk to in times of crisis. The practice should be extended to women’s prisons to the extent that it is not already happening.
The Prison Service should retain the present practice of not asking new prisoners at reception or in the first night centre or during induction whether they would prefer to be in a cell on their own or to share a cell with another prisoner.
Subject to recommendation 7, the rule that an unconvicted prisoner should not be required to share a cell with a convicted prisoner should always be complied with.
The sole exception to that rule – namely when the unconvicted prisoner consents to share a cell with a convicted prisoner – should be regarded as applying only when the unconvicted prisoner consents to share a cell with a particular convicted prisoner, not with convicted prisoners in general.
All establishments should have a system for ensuring that immediate effect is given to a prisoner’s change of status, by making certain that they do not share a cell with a prisoner of a different status.
The guidelines should proceed on the assumption that the lack of privacy which cell-sharing entails is more likely to be ameliorated if prisoners with a common ethnic and religious background share cells. But that should only be the starting point for the process. All prisoners should be interviewed on their arrival, either in the first night centre or during their induction, to enable them to explain their preferences for the type of prisoner they would prefer to share a cell with.
All decisions about who a prisoner should share a cell with should be made, if possible, by a senior officer. If that cannot be done, the decision should be reviewed by a senior officer at an early opportunity. The suitability of the two prisoners to continue to share with each other should be reviewed at regular intervals, with the prisoners’ personal officers being consulted over the issue.
The Prison Service should publish guidelines to assist officers in handling requests by prisoners to share a cell with a particular prisoner. Practical problems should not be treated as an insuperable hurdle preventing an otherwise suitable move of a prisoner from one cell to another.
The guidelines should require officers to keep prisoners informed of the progress of their requests, and if the request is refused, to notify prisoners of the reason for the refusal, unless security considerations or issues of confidentiality make that inappropriate.
The guidelines should contain guidance on how such requests should be recorded, but there is no need for such requests to be treated as formal applications under the requests and complaints procedures.
Wings holding convicted and unconvicted prisoners together should be kept to a minimum, and should only be used when there is no operational alternative.
The Prison Service should review whether the advantages of holding young offenders on the same wing as adult prisoners outweigh the disadvantages, and whether the practice should be extended to other establishments.
If the practice of holding young offenders on the same wing as adult prisoners is to continue, the law should be clarified to put its legality beyond doubt.
As soon as practicable, the Prison Service should assess the popularity of the bolted-down furniture made from white wood which is currently being trialled. It should then formulate a policy about the most appropriate form of furniture for use in cells, balancing the need to keep prisoners safe from their cellmates against the need for prisoners to live in cells which have a measure of homeliness, and taking into account prisoners’ preferences and cost.
In any event, the Prison Service should assess the resource and security implications of less frequent but random fabric checks against more frequent and more random full cell searches, bearing in mind that different strategies may be required for different establishments.
Senior officers should ensure that their staff know how to carry out fabric checks. They should ensure that officers realise that in order to check the walls of a cell, it will be necessary to look behind or under the furniture next to them. If that can only be done by moving the furniture, the furniture should be moved. If in the course of doing so they discover that the furniture is broken, they should check whether any parts of it could be used as a weapon.
An establishment’s written cell-searching strategy should require each cell to be fully searched at least once in every three months. Senior officers should ensure that full cell searches are taking place as regularly as the strategy requires. Returns showing which cells have been fully searched, and when, should be submitted monthly. The department which is supposed to scrutinise the returns should do so speedily, and notify the senior officer of the unit and its principal officer if that is not happening.
The violence reduction strategy should be used as a vehicle to encourage prisoners to feel that they have a stake in making their prison safe – in particular, by encouraging prisoners to think that they have let other prisoners down if they resort to violence, and by letting prisoners have a say in the running of their prison through prisoner councils. Every prison should be required to have a functioning prisoner council made up of elected representatives which meets at regular intervals.
Information about a prisoner’s convictions and outstanding charges held on NOMIS should include a short statement of the facts of each offence or charge.
If a convicted prisoner is not due to be released when they complete their sentence because they were remanded in custody awaiting trial on other charges, NOMIS should flag up the date on which their change of status is due to occur.
Although there are no plans for security information to be entered onto NOMIS in its first phase, Security Departments should enter any information which can be shared with the majority of staff when the first phase of NOMIS becomes operational.
NOMIS should include a facility for an alert to appear if information is held by the Security Department on prisoners which could affect their management but which is too sensitive for wider dissemination. An officer at the grade of senior officer or above should be able to ask for that information, and the request should be considered by the governor with line management responsibility for the Security Department. The governor should be able to refuse the request, or grant it on condition that the senior officer does not reveal the information to anyone, or on condition that the senior officer can tell their wing staff about it on the understanding that it is not to go any further.
Information overload should be avoided by enabling officers to get to the information they need quickly and to bypass the information they do not need. The technology should be up to date, and sufficient terminals should be provided to ensure that staff have ready access to NOMIS at all times.
The training which staff receive on NOMIS should not merely address how to log on, enter information and retrieve it. It should reinforce the need for any information which is to be entered to be accurate, comprehensive and unambiguous. It should also reinforce the need for all staff to be aware of the background and offending history of the prisoners in their charge, as well as their previous behaviour in prison. Staff should learn that the system will be useless if they do not use it properly.
To avoid prisoners leaving court without being accompanied by bail information reports or pre-sentence reports, the probation officer should ensure that a copy of the report is available for the escort contractor, and there should be someone in court whose responsibility it is to ensure that the dock officer gets it. Escort contractors should have a list of the documents which should accompany prisoners when they leave court, but if a prisoner leaves court without all the documents, the court should ensure that they are sent on to the prison without delay.
There should be someone in court whose responsibility it is to ensure that the dock officer gets copies of the warrant authorising the prisoner’s detention, the list of their convictions, and the indictment or charge sheet.
When a judge asks for any remarks which they make to be brought to the attention of the Prison Service, the court should assume responsibility for commissioning a transcript of what has been said and sending a copy to the prison.
In the light of such legal advice as the Courts Service receives, it should publish a policy on the disclosure to the prison of medical or psychiatric reports on a prisoner submitted to the court.
The list of a prisoner’s convictions sent to the prison should include a short statement of the facts of each offence and the charges the prisoner is facing.
The Police National Computer should be linked to the whole of the prison estate. In the meantime, any intelligence the police may have about prisoners which could affect their management in prison should be sent to the police liaison officers for the establishments at which the prisoners are being held. A decision can then be made whether the intelligence can be disseminated widely within the prison or given to a governor for their eyes only.
Whenever a prisoner is transferred to another establishment, the receiving establishment should be told what the reason for the transfer is. If the transferring prisoner is a particularly problematic one, the receiving establishment should be warned beforehand.
To ensure that all files accompany prisoners on their transfer to another establishment, they should be ticked off at the reception of the sending establishment against a checklist. Prisoners should not be allowed to leave unless all their files have been ticked as present, except with the permission of a governor. Staff on reception at the receiving establishment should notify the department responsible for chasing up files which do not arrive with a prisoner, entering the action they have taken in a “missing file book”. Consequential action and the eventual receipt of the files should also be entered in the book.
Prisoners should not be admitted to a wing without their current wing file, save with the permission of a governor or the approval of the night orderly officer. Important entries in it should be made in red.
If prisoners arrive on a wing without a copy of the cell-sharing risk assessment form, they should be placed in a single cell until the form is found or a new one completed.
All establishments should have a procedure for notifying wings in writing that a prisoner is currently charged with, or has been convicted of, an offence under the Protection from Harassment Act 1997.
Each Security Department should establish a proper system for vetting security files to ensure that they are read by the end of the working day following their arrival, that any relevant information is relayed in writing to the wings, and that a record is kept of who vetted them, with the date and time.
Since security information is not to be included in the first phase of NOMIS, there remains a need for a national database for security information. Establishments should have the 4x4 system converted to the 5x5 system if that has not already been done, and the systems should be networked across the prison estate.
Where prisoners have served a sentence before, the Security Department should always obtain their security files from the establishment from which they were discharged.
Documents held in an establishment’s administrative section should be vetted by staff from the Security Department, and each establishment should have a protocol for that work along the lines of the system for vetting security files.
The senior officer on a wing should ensure that information arriving on the wing about a prisoner is recorded in the wing observation book. Particularly important information should be in red. When coming on duty, staff should read any entries in the wing observation book made since they were last on duty.
The discharge report which used to be prepared on prisoners on their discharge should be completed as a matter of course, and a copy included in their main prison file. It should be accessible on NOMIS when NOMIS becomes operational.
The Prison Service should publish a model procedure dealing with how establishments should bring Prison Service Orders and other instructions, whether national or local, which affect the management of prisoners to the attention of staff. The model procedure should be regarded as having been adopted by any establishment which does not produce one of its own.
Governors should ensure that any relevant comments or recommendations in external reports about their establishments which have implications for the safety of prisoners be brought to the attention of the workforce.
Every establishment should appoint an officer not below the grade of governor to be responsible for overseeing the flow of information. Such an officer should ensure that systems are in place for the transfer of information within an establishment and that the systems are being followed. They should take action when they find that they are not, and should review the arrangements periodically to ensure that best practice is being maintained.
The handbook giving guidance on how to complete the prisoner escort record should be revised to make it clear that it is not merely of use while a prisoner is in transit, but it is also intended to inform the prisoner’s management in the receiving establishment. It should give clear guidance to staff as to when a box should be ticked. Staff who are tasked with completing the form should be instructed on how to complete it by senior officers in their department. They should each be provided with a copy of the handbook, and they should be reminded of the need to spell out the reasons for a particular box being ticked.
Staff who are tasked with completing form XF001 should be instructed on how to complete it by a senior officer. In particular, they should be reminded that they are only assessing the risk prisoners pose to other inmates. They should not automatically assess prisoners as a high risk simply because they claim, for example, to be prone to lose their temper, but should ignore such claims if they believe them to be untrue, and they should guard against being over-defensive.
The instructions for completing the form should give duty governors guidance on how to exercise the options available to them when dealing with prisoners who are both at risk of self-harm and a risk to their cellmate.
The first review of the initial assessment should take place within one week of the initial assessment, and should take place in every case. It should be a multi-disciplinary review, with representatives from the prisoner’s wing, healthcare and the team responsible for implementing the establishment’s violence reduction strategy all contributing to it. The documents set out in paragraph 59.15 should be considered, and the participants should have been briefed on the contents of the prisoner’s security file if there is one.
Wing officers should be reminded of the need to call for a review of an assessment when the necessity for one is triggered by some occurrence which might affect the prisoner’s emotional well-being.
The officer responsible for monitoring the processes for implementing measures outlined in risk minimisation plans should also be responsible for monitoring their actual implementation.
OASys should be used by all establishments to identify the risks posed by prisoners to staff and other inmates, and the risks which are identified should be fed back to the wings.
OASys assessors should have access to information relating to prisoners’ behaviour while serving previous sentences and to information held in security files.
All adult prisoners serving sentences of less than 12 months’ imprisonment who had an OASys assessment before being sentenced should have that assessment reviewed during the induction process. The review need only address the risk of harm which the prisoner poses to staff and other inmates. If resources permit, adult prisoners serving sentences of less than 12 months’ imprisonment who did not have an OASys assessment before being sentenced should have one during the induction process.
The risk management model for the management of prisoners to whom MAPPA applies and who have been identified as posing the greatest risk to the public should be adopted by all establishments. Prisoners who have been identified as being a high risk to their cellmate on any review of the cell-sharing risk assessment, or posing a very high or high risk to staff and other inmates on the OASys assessment, should be referred to the inter-departmental risk management team envisaged by the model.
It is neither necessary nor desirable to introduce a risk classification, similar to the security classification, which identifies the degree of risk which a prisoner poses to staff and other inmates.
That part of the basic training course for new prison officers which focuses on the development of their inter-personal skills should be reviewed in the light of this report and the lessons to be learned from Zahid’s murder. In particular, two of the key attributes required of the prison officer should be stressed. They are the ability to pick up on what is happening on the wing, which prisoners pose a risk to other inmates, and what prisoners might be worried about, and the ability to earn the respect of the prisoners on the wing so that they are prepared to confide in them without fear of it getting out.
It should be mandatory for all establishments to have a personal officer scheme. That includes busy local prisons, although if time for personal officer work is limited, it should be used constructively.
Personal officers should be assigned to individual prisoners, not to a group of cells. They should be members of a small team, so that when a prisoner’s personal officer is not on duty, the prisoner can approach another member of the team.
The role of the personal officer should be clearly defined in each establishment’s personal officer scheme. The Prison Service should publish a model scheme, which should be regarded as having been adopted by every establishment which does not produce one of its own.
The Prison Service’s policy on whistleblowing should identify the most appropriate way for staff, in an exceptional case, to get confidential advice from an independent outsider and to raise their concerns outside the Prison Service. It should also ensure that the members of staff of companies who contract with the Prison Service, such as escort contractors and those responsible for the running of the contracted-out establishments, have access to the Prison Service’s whistleblowing policy.
Making a false and malicious allegation that wrongdoing has taken place should be expressly stated to be a disciplinary offence, for which dismissal from the Prison Service may be an appropriate sanction.
Research should be conducted on how effective the obligation to report wrongdoing – and the designation of a failure to do so as a disciplinary offence – has been. This should be done by comparing the number of instances of reported wrongdoing before and after these measures were implemented.
A comprehensive review of the quality of care provided to prisoners with mental health problems and its effectiveness should be conducted once the changes introduced since Zahid’s murder have had a chance to work.
The first reception health screen questionnaire should be revised so as to trigger a referral to a mental health professional on the healthcare team even if the prisoner has only self-harmed in prison. A referral should also be triggered where the prisoner’s behaviour is such that the healthcare officer completing the questionnaire considers it desirable.
The responses of those members of staff who have attended the mental health awareness training course should be analysed to determine whether the course can be improved. The number of front line staff attending the course should be increased, resources and staff deployment permitting.
Profiled time should be set aside for staff to read the booklet explaining the main components of the mental health awareness training course. The booklet should include advice given on the course about how prisoners with particular disorders should be managed on the wing. That advice should be published and made freely available.
The measures which should be taken to minimise the risk which a mentally disordered prisoner on ordinary location poses to staff and other inmates includes: • not placing such a prisoner in a shared cell • if such a prisoner is to share a cell, carefully selecting their cellmate • ensuring that, whatever difficulties there may be in operating a proper personal officer scheme, such a prisoner has a personal officer who is fully aware of their background and who makes a particular effort to get to know them and keeps an eye on their state of mind • checking the correspondence and searching the cell of such a prisoner more frequently and carefully than would otherwise have been the case • ensuring, again regardless of the difficulties which might be faced in providing a good regime for all prisoners in the establishment, that such a prisoner is appropriately occupied with, for example, work, education or offending behaviour programmes • keeping a closer watch over material such as films, to which such a prisoner has access, and exercising control over their suitability • checking on NOMIS or with the Security Department about the existence of any useful intelligence about such a prisoner and what is known about their previous behaviour in prison.
The Prison Service should prepare a readable guide, which explains the circumstances in which personal information about a prisoner should be disclosed by healthcare staff to officers on the wing. The guide should contain practical examples of situations where disclosure should or should not be made.
The Prison Service and the CRE should address the key areas for development identified by the Inspectorate in Parallel Worlds in managing the action plan, Implementing Race Equality: A Shared Agenda for Change.
The diversity training which prison officers receive as part of their basic training should stress the need for them to put themselves in the position of BME prisoners and see things from their point of view. They should be told about the techniques they can use to develop that skill.
The training of any officer responsible for investigating complaints of racism should stress that corroborative evidence of a complaint, though desirable, is not essential. Officers should be reminded of the need to guard against falling into the trap of seeing as decisive the existence of a possible racially neutral reason for treatment which would otherwise be discriminatory.
Prisoners should not be used under any circumstances to assist in the investigation of complaints of racism, or to act as an intermediary between the complainant and the investigating officer.
The definition of a racist incident adopted by the Stephen Lawrence Inquiry should be used to identify what constitutes a complaint of racism, so that a complaint of racism is one where the action complained of is perceived to be racist by the victim or any other person.
The Prison Service, the Inspectorate and the CRE should consider whether there is a need for the investigation of complaints of racism and other serious complaints to be carried out by an independent body, or at least to be carried out with a strong independent element built into the process.
The Prison Service and the CRE should investigate the desirability and feasibility of race relations liaison officers being recruited from outside the Prison Service.
Without suggesting in any way that the Prison Service should be regarded as institutionally infected with religious intolerance, thought should be given by the Home Office to recognising the concept of institutional religious intolerance, along the lines of the definition of institutional racism adopted by the Stephen Lawrence Inquiry.
The Muslim Adviser to the Prison Service should consider how the role of prison Imams can be expanded – without in any way compromising their religious role – so as to make them readier to assist with the non-religious needs of Muslim prisoners.
The Home Office should promote legislation to add each prison to the list of bodies required to publish race equality policies or race equality schemes under the Race Relations Act 1976.
The Zahid Mubarek Inquiry examined the circumstances surrounding the murder of 19-year-old Zahid Mubarek by his cellmate Robert Stewart at Feltham Young Offender Institution on 21 March 2000. Stewart, who had documented violent and racist tendencies, attacked Mubarek with a table leg hours before Mubarek's scheduled release. The inquiry, chaired by Mr Justice Keith, made 88 recommendations addressing cell-sharing procedures, management of prisoners with racist attitudes, and information-sharing between criminal justice agencies.
The inquiry's findings led to documented changes in prison procedures. The Offender Management Act 2007 incorporated provisions strengthening cell-sharing risk assessment duties. The National Offender Management Service introduced enhanced procedures for identifying and managing prisoners displaying racist behaviour. A revised Cell-Sharing Risk Assessment tool was rolled out across the prison estate, requiring senior officer involvement in cell-sharing decisions.
HM Inspectorate of Prisons' thematic review in January 2009 found that cell-sharing risk assessment procedures had been improved following the inquiry. However, the same review identified that racist bullying remained a significant problem and that information-sharing between establishments continued to be inconsistent. The inquiry's primary objective - the elimination of enforced cell-sharing - has not been achieved.
The inquiry contributed to broader examination of racism within the criminal justice system, alongside the Commission for Racial Equality's 2003 formal investigation into racism in the Prison Service. Its recommendations continue to influence prisoner management protocols and risk assessment procedures within the UK prison system.
Reforms attributed to this inquiry · AI-generated• Cell-Sharing Risk Assessment (CSRA) tool introduced across the prison estate, with procedures strengthened through provisions in the Offender Management Act 2007 • Enhanced procedures for identifying and managing prisoners with racist attitudes established by the National Offender Management Service • Revised information-sharing protocols between criminal justice agencies regarding prisoner risk factors • Formal requirement for senior officer involvement in cell-sharing decisions • Mandatory recording systems for racist incidents in prisonsUnfinished business · AI-generated• Elimination of enforced cell-sharing remains unachieved - the inquiry's primary objective (recommendation ZAHI-1) • HM Inspectorate of Prisons (2009) found information-sharing between establishments remained inconsistent despite new protocols • HM Inspectorate of Prisons (2009) reported racist bullying continued as a significant problem in prisons
Strengthened the Prison Service's duty to assess and manage risks in cell-sharing, informed by the Zahid Mubarek Inquiry's findings.
Implementation reviewed by
HM Inspectorate of Prisons (thematic review)(Jan 2009)
Thematic review of race equality in prisons. Found that cell-sharing risk assessment procedures had been improved following the inquiry, but that racist bullying remained a significant problem and information-sharing between establishments was inconsistent.
Influence & connections
Led directly toPrison safety and racial equality
The inquiry led to reformed cell-sharing risk assessment procedures and contributed to the broader anti-racism agenda within the criminal justice system, alongside the CRE's formal investigation into Prison Service racism (2003).