HAYLEY CANHAM (R on the application of) v THE DIRECTOR OF PUBLIC PROSECUTIONS

The Claimant is the ex-partner of Robert Fenlon, and the mother of his daughter. Mr Fenlon took his own life while in prison custody at HMP Woodhill on 5 March 2016. The Claimant challenges the decision of the Defendant, the Director of Public Prosecutions, not to prosecute any one of three individual suspects (known as A, B and C), the Ministry of Justice (“MoJ”) or the National Offender Management Service (“NOMS”) in relation to Mr Fenlon’s death. That decision was made on 19 October 2020 by Ben Southam, senior specialist prosecutor who was acting on the Defendant’s behalf (the “Decision”).

The Decision was a review of an earlier decision by Colin Gibbs, senior specialist prosecutor, dated 5 April 2019. Dissatisfied with Mr Gibbs’ decision not to prosecute any one of the five suspects, the Claimant exercised her victim’s right of review (under the “VRR scheme”); Mr Southam was the reviewing lawyer.

Pre-action correspondence followed the Decision. The Claim Form was issued on 8 January 2021. Permission for judicial review was granted by Swift J at an oral renewal hearing on 27 May 2021.

The matter was heard on 27 October 2021. At the hearing, the Claimant was represented by Mr Menon QC and Mr Nichols. The Defendant was represented by Mr McGuinness QC. I am grateful to all counsel and to the solicitors who instructed them for the assistance they have given to the Court.

This is a sad case. The loss of Mr Fenlon’s life is a tragedy for his family and those who knew him. This case proceeds in the face of evidence about many other prisoners who have taken their lives at HMP Woodhill. Whatever the legal rights and wrongs of this case, which I shall go on to discuss, I acknowledge at the outset the extremely serious context in which the issues in the case arise.

A detailed chronology of events leading up to the death of Mr Fenlon is set out in the witness statement of Jo Eggleton, solicitor for the Claimant, dated 20 July 2021. What follows is a summary of the key events in that chronology.

Mr Fenlon was born on 5 September 1980. On 15 October 2015 he was remanded in custody on a charge of burglary. He arrived at HMP Woodhill with standard documents, including a record which indicated he was withdrawing from opiates. He had a reported history of mental health issues.

On 26 February 2016, Mr Fenlon passed a note under his door stating that he was having thoughts of self-harm. An officer retrieved the note.

In line with guidance issued by the MoJ and NOMS in relation to management of prisoners at risk of harm to self, to others and from others (Prison Service Instruction or PSI 64/2011), the officer opened a file as part of the Assessment, Care in Custody and Teamwork process (“ACCT”). Within an hour, another officer opened an Immediate Action Plan (“IAP”). Under that IAP, directions were given that Mr Fenlon should not share a cell and should have two conversations per session and five observations per night. Access to a phone and to a prison listener trained by the Samaritans was explained to him. Medical intervention was not considered to be appropriate.

On 27 February 2016, Mr Fenlon lost his job as a painter at the prison. It seems that he was wrongly suspected of having taken a paint brush. He was reported to be spending more time alone in his cell because he had lost his job, he was becoming reclusive and depressed. At about 10am that day, an ACCT assessment took place. Mr Fenlon was offered a move to another wing which he declined. He said he had been waiting to see mental health services for 5 months and thought seeing them would help. At 11am a case review took place; Mr Fenlon said he had feelings of paranoia but did not have feelings of self-harm or suicide at that time.

On 28 February 2016, Mr Fenlon said he felt under threat and wanted to stay in his cell; he felt paranoid and said he would rather take his own life than be killed in prison. He asked to speak to C, who was a senior healthcare assistant. He met C at 3.31pm. C referred Mr Fenlon to the GP.

On 29 February 2016, Mr Fenlon reported that he had slept on the floor because of anxiety attacks triggered by his mattress.

On 2 March 2016, a scheduled ACCT review took place. Mr Fenlon told the officers in attendance that he had no present thoughts of self-harm. The next review was scheduled for 8 March 2016, with risk being assessed as low.

On 3 March 2016, A, who was a senior officer at the prison, entered Mr Fenlon’s cell with another officer. They found him hanging from a ligature made of bedding attached to the window. The ligature was cut and Mr Fenlon was sat on his bed. Razor blades were removed from his cell. A noted that Mr Fenlon was very distressed and when the noose was cut red marks around Mr Fenlon’s neck were revealed. He was seen by C and another healthcare trainee. There is disputed evidence about whether an ACCT review was conducted. Notes were found in which Mr Fenlon said that he could not cope and he was on the verge of doing something daft. The level of risk was raised as a result of “attempted hanging”. Observations were increased to two per hour. Mr Fenlon attended a court hearing that day via video-link from the prison; he was accompanied to this hearing by A. He spoke to a prison listener that afternoon.

On 4 March 2016, Mr Fenlon told another officer that he was having bad thoughts again. He spoke to the Samaritans by phone. He later told an officer that he was feeling suicidal again. At around 1pm, an officer saw that Mr Fenlon had created a noose which was attached to his outer window bars. Officers entered his cell and removed the noose. At about 2pm, B, a senior officer at the prison, conducted an ACCT review, which resulted in the risk remaining unchanged, and the frequency of observations remaining at two per hour. The next scheduled review, due for 8 March 2016, was brought forward to 5 March 2016. Mr Fenlon said he was unsure if he would attempt to harm himself again.

On 5 March 2016, Mr Fenlon was extremely paranoid. He spoke to a listener. He was observed on several occasions that morning by prison officers. At 10.40am an officer noticed that his cell observation panel was blocked. Officers entered the cell to find Mr Fenlon hanging from a noose tied to the window. He was cut down and CPR commenced but he could not be resuscitated.

Much of the argument in this case has centred around the way the ACCT process was managed in Mr Fenlon’s case. I have already referred to PSI 64/2011. That guidance is issued by MoJ and NOMS jointly. It is a lengthy document running to 70 pages. It would have been available to all the officers and healthcare staff involved with Mr Fenlon in the last days of his life, and it plainly informed their actions.

The executive summary records that the document “sets out the framework for delivering safer custody procedures and practices to ensure that prisons are safe places for all those who live and work there”. The guidance requires that prisoners who are at risk of harm to self are managed using ACCT procedures, which are outlined in chapter 5.

Turning to Chapter 5: the ACCT process requires that any member of staff who receives information or observes behaviour which may indicate a risk of suicide/self-harm must open an ACCT by completing the Concern and Keep Safe form (Chapter 5, p 26). Within an hour of the ACCT being opened, a manager must talk to the prisoner and complete the IAP to ensure the prisoner is safe from harm and must inform healthcare, including the mental health in-reach team where appropriate, arrange for an ACCT assessment to take place and organise the first case review, and ensure that the prisoner has been offered the opportunity to talk to a listener and/or the Samaritans (Chapter 5, pp 26-27). The ACCT assessment must be undertaken by a trained ACCT assessor and must involve an interview with the prisoner within 24 hours of the ACCT being opened; the first review must be held within 24 hours of the ACCT being opened, ideally immediately after the ACCT assessment interview. The review is multi-disciplinary and is usually attended by the prisoner; a case manager should be appointed (Chapter 5, pp 27-28). The outcome of the review should be a CAREMAP giving “detailed and time-bound” actions aimed at reducing the risk to the prisoner. This document must reflect the prisoner’s care needs, level of risk and triggers for distress, and should cover a range of considerations such as level of supervision and cell sharing (Chapter 5, pp 28-29).

Chapter 6 deals with constant supervision and notes, in the overview section at the beginning, that “constant supervision must only be used at times of acute crisis and for the shortest time possible”. That is because the process of being constantly supervised by a member of staff can be de-humanising which may increase risk (Chapter 6, p 33).

Mr Fenlon’s death by suicide was not an isolated incident at HMP Woodhill. Between May 2013 and the date of Mr Fenlon’s death, twelve prisoners had taken their own lives, all found hanging in their cells.

Over time, a number of reports and inquiries have been produced, examining the deaths of prisoners by suicide at HMP Woodhill. Many of them consider the ACCT process and the way PSI 64/2011 was implemented, its strengths and weaknesses. The Senior Coroner for Milton Keynes has produced a number of prevention of future deaths reports, following inquests into the many suicides at HMP Woodhill.

The Claimant noted that a number of these reports preceded Mr Fenlon’s death, and exposed problems with the ACCT process. The reports in this category, most of which were put in evidence in this judicial review, are:

A Report by the Prisons and Probation Ombudsman Nigel Newcomen CBE dated May 2013 (the “PPO Report”),

A report by Her Majesty’s Senior Coroner dated December 2015, following the inquest into the death of another prisoner, Daniel Byrne (the “SC Report”),

The report by HM Chief Inspector of Prisons following an unannounced inspection of HMP Woodhill, dated January 2016 (the “HMCIP Report”).

The Safer Custody Review dated 18 February 2016 undertaken by the NOMS National Safer Custody Delivery Manager (the “Safer Custody Review”),

A number of reports by the Senior Coroner for Milton Keynes (Mr Tom Osborne) to prevent future deaths, including the report dated 8 August 2014 into the death of Sean Brock (“PFD reports”).

The Claimant notes the reports into deaths at HMP Woodhill which post-dated Mr Fenlon’s death, including:

A joint review at the request of the Governor of HMP Woodhill, the local NHS England Commissioners and the healthcare provider at the prison, dated 31 March 2016 (the “Joint Review”),

The report by HM Chief Inspector of Prisons following an unannounced inspection of HMP Woodhill, dated April 2018 (the “second HMCIP Report”).

The report by Stephen Shaw which was commissioned in February 2017 and reported in May 2017 on the prevention of self-inflicted deaths and self-harm at HMP Woodhill (the “Shaw Report”),

An MoJ report issued in 2019 but based on research in 2015 (the “MoJ report”).

Some of these investigations and recommendations were examined by Mr Southam in the Review Note. He noted in addition a CQC inspection in September 2016 and criminal proceedings at the Old Bailey in relation to the death of Ryan Harvey on 8 May 2015; the criminal proceedings resulted in an acquittal of the custodial manager who was charged. He considered a number of PFD reports on individual deaths, and the PPO reports on a number of other individuals also and a PPO annual report for 2016-17. Mr Southam summarised the criticisms of the ACCT made by Mr Shaw and the PPO at § 4.14 of the Review Note. I shall return to this paragraph later to deal with a challenge by the Claimant which arises out of it.

The Decision was conducted according to guidance contained in the Code for Crown Prosecutors, issued under section 10 of the Prosecution of Offences Act 1985. The guidance requires a prosecutor to adopt a two-stage test, known as the Full Code Test. The first stage is consideration of the evidence to decide whether there is a realistic prospect of conviction (the “evidential stage”). If the case does not pass the evidential stage, it should not go ahead however serious or sensitive it is. The test is whether a properly directed jury is more likely than not to convict. The prosecutor is required to assess the evidence objectively, including the impact of any defence and any other information which might be put forward at trial. The second stage is whether a prosecution is required in the public interest (the “public interest stage”).

The CPS considered the possibility of bringing charges against A, B, C, MoJ and NOMS. On 5 April 2019, the CPS (by Colin Gibbs) informed the Claimant of the CPS’ decision that no individual or organisation should be prosecuted for any offence arising out of events connected to Mr Fenlon’s death. The CPS concluded that the evidential stage was not passed in relation to any charge which was under consideration in relation to any of the five suspects. The charges under consideration were: gross negligence manslaughter, misconduct in public office, corporate manslaughter, breaches of section 3 and 7 of the Health and Safety at Work Act 1974 (HSWA 1974). Mr Gibbs made a total of 13 prosecutorial decisions, all of them not to prosecute.

By letter dated 16 May 2019, Deighton Pearce Glynn, solicitors for the Claimant, sought a review under the VRR scheme. That letter noted the “fundamental injustice in Mr Gibbs’ reasoning: no individual is to be prosecuted because of the general poor state of Woodhill at the time and yet the Ministry of Justice/NOMS won’t face prosecution either because they were too removed from the decisions about [Mr Fenlon]’s care.”

In keeping with internal procedures at the CPS, the matter was passed to a review lawyer from a different CPS area, Mr Southam. He works with the special crime and counter terrorism division based in York. He conducted a review of Mr Gibbs’ decision. He obtained the opinion of Senior Treasury Counsel, Timothy Cray QC, before finalising the Decision. Senior Treasury Counsel agreed that the Full Code Test was not met in relation to any one of the suspects.

On 19 October 2020, Mr Southam issued the Decision to Deighton Peirce Glynn. He concluded that there was insufficient evidence to provide a realistic prospect of conviction against any of the individuals, or either of the organisations (MoJ and NOMS). He explained the Code for Crown Prosecutors, summarised the facts, and then set out the ingredients of each of the offences under consideration. He considered the acts and omissions of each individual and the two organisations. He concluded that “there is insufficient evidence to provide a realistic prospect of conviction of any of the suspects”.

The Decision therefore upheld the 13 decisions made by Mr Gibbs, although the reasons for those decisions varied to some degree.

There is common ground as to the approach the Court should take to judicial reviews of prosecutorial decisions, subject to an argument by the Claimant that a modified approach applies when Article 2 of the European Convention on Human Rights is in issue. I will deal with that argument below. In this section, I record the common ground.

A number of cases have considered the approach the Court should take to the review of a decision by the DPP not to prosecute. The starting point, agreed by both parties, is R v DPP ex p Manning [2001] QB 330 at [23]:

The Claimant particularly emphasises the last sentence of this extract and urges this Court not to adopt a standard of review which is set too high. The Defendant particularly emphasises an earlier part of this extract which explains the prosecutor’s role in assessing how the case against the defendant “would be likely to fare” in the context of a criminal trial taking account of all the evidence and the likely defences.

The Claimant relies on a number of cases which have followed Manning, including R (Webster) v Crown Prosecution Service [2014] EWHC 2516 (Admin) which confirmed this approach at [16] and R (L) v DPP [2020] EWHC 181 (Admin) for the proposition at [11] that no margin of discretion can save a decision which is wrong.

In L v Director of Public Prosecutions [2013] EWHC 1752 (Admin) the President of the Queen’s Bench Division (Sir John Thomas) emphasised that challenges to decisions not to prosecute will only succeed in very rare cases, for the good and sound constitutional reason that decisions to prosecute are entrusted to the prosecuting authorities; it is for that reason that the courts adopt a “very strict self-denying ordinance”, although they will of course put right cases where an unlawful policy has been adopted or where there has been a failure to follow policy or where decisions are perverse, noting that those are likely to arise in exceptionally rare circumstances, see [5]-[7]. The process of internal review following a victim’s request for a review meant that it would be even more difficult now to bring a successful challenge to the decision not to prosecute, see [10].

Mr McGuinness reminds me that the Court in Manning drew on the classic statement by the Divisional Court in R v Director of Public Prosecutions ex p C [1995] Cr App Rep 136 that:

because of some unlawful policy (such as the hypothetical decision in Blackburn not to prosecute where the value of goods stolen was below £100); or

because the Director of Public Prosecutions failed to act in accordance with her own settled policy as set out in the Code, or

because the decision was perverse. It was a decision at which no reasonable prosecutor could have arrived.”

He says that the Manning / ex p C approach has been emphasised in many other cases such as R (Pepushi) v CPS [2004] EWHC 798 (Admin) at [49], Sharma v Brown-Antoine [2007] 1 WLR 780 at [14(5)], R (Bermingham and Others) v Director of SFO [2007] QB 727 at [63] and S v Oxford Magistrates Court [2016] 1 Cr App Rep at [14].

He notes that the most recent statement of the court’s approach is to be found in the judgment of the Lord Chief Justice in R (Monica) v DPP [2018] EWHC 3508 (Admin):

Particularly where a CPS review decision is exceptionally detailed, thorough and in accordance with CPS policy, it cannot be considered perverse: L’s case 177 JP 502, para 32.

A significant margin of discretion is given to prosecutors: L’s case, para 43.

Decision letters should be read in a broad and common sense way, without being subjected to excessive or overly punctilious textual analysis.

It is not incumbent on the decision-makers to refer specifically to all the available evidence. An overall evaluation of the strength of a case falls to be made on the evidence as a whole, apply prosecutorial experience and expert judgment.”

This line of authority is of long-standing. It emphasises the need for the Courts to stand back – the more so since the VRR scheme was introduced – and to permit a significant margin of discretion to the decision-maker, without setting the bar too high. Intervention is warranted if the decision is bad in law. Mere disagreement with the merits of the decision is insufficient.

Deighton Pierce Glynn, solicitors for the Claimant, sent a pre-action protocol letter outlining grounds of challenge on 19 November 2020 in relation to five of the decisions not to prosecute. The CPS (by Mr Southam) provided a pre-action protocol response letter on 3 December 2020, resisting the grounds and providing some further explanation.

The Claim was issued on 8 January 2021. Summary grounds resisting the claim were filed on 26 January 2021. The Claim Form challenged the same 5 decisions as intimated in pre-action correspondence, namely:

the decisions not to prosecute A and/or B for gross negligence manslaughter,

the decision not to prosecute C for a breach of s 7 of the HSWA 1974, and

the decisions not to prosecute either of the MoJ or NOMS for corporate manslaughter.

After permission was granted by Swift J on 27 May 2021, the Defendant filed evidence in the form of a witness statement from Mr Southam dated 24 June 2021, exhibiting a copy of Mr Southam’s review note dated 17 December 2019 updated on various occasions to a final date of 25 June 2020 (the “Review Note”).

The Claim Form sets out five grounds of appeal in relation to each of the five individual decisions under challenge (see paragraph 42 above):

Each decision was incompatible with the procedural duty in Article 2. The question whether a prosecution should take place is one for the Court. It is not, in this context, a Wednesbury irrationality challenge. The remaining grounds are in the alternative.

Each decision failed to take into account relevant considerations.

Each decision gave irrational weight to certain relevant considerations.

Each decision contained material errors of law.

Each decision was irrational.

The Claimant’s skeleton adjusts the list of challenges, adding a new purported ground of material error of fact said to be in response to the Secretary of State’s disclosure, and relegating the Article 2 challenge to a submission that this Court should apply a heightened standard of review.

The grounds substantially overlap, although they differ in their detail depending on the individual decision under challenge. It appears to me that there are four issues to be resolved:

Assuming (without deciding) that Article 2 is engaged on the facts of this case, is the Court required to take a different approach to the review of the decision to prosecute than if Article 2 was not engaged?

Was the decision not to prosecute A or B for gross negligence manslaughter wrong in law?

Was the decision not to prosecute C for breach of s 7 HSWA 1974 wrong in law?

Was the decision not to prosecute MoJ or NOMS for corporate manslaughter wrong in law?

The Claimant, by Mr Menon QC and Mr Nicholls, submit that a different approach is necessary because Article 2 imposes a procedural obligation to carry out a thorough and effective investigation, triggered in this case by the fact that Mr Fenlon committed suicide while in prison and in the care of the state. As they note, he was not the only prisoner at HMP Woodhill to take his own life. There is a serious concern about the safety of prisoners at that prison at that time given the number of suicides over the few years preceding Mr Fenlon’s death (and indeed after his death). They cite Öneryildiz v Turkey (2005) 41 EHRR 20 (Grand Chamber) at [91], [94] and [96] for the proposition that “The national courts should not under any circumstances be prepared to allow life-endangering offences to go unpunished”. They also cite R (Middleton) v HM Coroner for West Somerset [2004] 2 AC 182 at [9], R (Birks) v Commissioner of Police of the Metropolis [2015] ICR 204 at [52], R (on the application of Skelton) v Senior Coroner for West Sussex [2020] EWHC 2813 (Admin) at [91] and Secretary of State for the Home Department v Nasseri [2010] 1 AC 1 at [12]-[14], to support their pleaded proposition that the Court must decide for itself whether Article 2 requires there to be a prosecution, rather than posing a question on review as to whether the Decision is Wednesbury irrational. In this context, they accept that the Court should give “due weight” to the views of the Defendant but they submit that this is an area where the Court has its own relevant knowledge of criminal law and that it should not show “undue deference” to the decision-maker, citing R (Eileen Alexandra Oliver) v Director of Public Prosecutions [2016] EWHC 1771 (Admin) at [9]). These points were largely articulated in the Claimant’s Statement of Facts and Grounds, with the skeleton argument suggesting a “heightened, intensive standard of review” (§ 60), and the matter touched on at the tail end of Mr Menon’s oral submissions without taking me to any of the cases cited.

Mr McGuinness submits that the Claimant is wrong in law; the approach is no different just because Article 2 is or might be engaged or because others had committed suicide while detained at HMP Woodhill. There is a single invariable approach which has been emphasised on many occasions by this Court – see Re C, Manning and the other cases in that line of authority referred to above. He says that the cases relied on by the Claimant, for example, Skelton and Middleton, are of no relevance.

In my judgment, Mr McGuinness is right to say that there is one approach to be adopted by this Court, whether or not Article 2 is engaged by the circumstances of the death, and whether or not the death stands alone or was one of a tragic sequence of suicides within a prison. I was shown no case involving a death at the hands of the state where the Court suggested that a different approach applied on a judicial review of a decision not to prosecute. The Court in Oliver, a case also involving death in custody, cited Manning saying that the principles were not there in dispute (see [5]). Cases like Skelton and Middleton are not relevant because they address the different issue of the scope of an inquest when Article 2 is engaged. Birks was a very different case (concerning police disciplinary proceedings, held not to be necessary in order to meet the state’s Article 2 obligation in that case). Nasseri is a case about Article 3, not Article 2, and it goes to a different point about the scrutiny which attaches to Convention claims.

I find the most direct answer to Mr Menon’s point in Öneryildiz v Turkey itself. That case confirms that Article 2 does not provide the right for an individual to have third parties prosecuted or sentenced for a criminal offence, see [91]-[92] and this extract from [94] (emphasis added):

In this case, there has been a thorough police investigation. There will be an inquest. Civil proceedings have been issued and are currently stayed pending the completion of this judicial review and (I anticipate) the inquest. The Article 2 investigative obligations of the state will need to be satisfied by means of these various strands. The task of the Court on this judicial review is not to assess whether the investigative obligation has been met or to determine whether there has been a Convention breach. It is to review Mr Southam’s conclusion, based on the findings of the investigations to date, that criminal proceedings are not justified. The approach to such a review is established by the Manning line of authority.

The Decision summarised the elements of the offence of gross negligence manslaughter drawing on cases such as R v Adomako [1995] 1 AC 171. No dispute arises on that summary, and the five elements described were (a) the suspect owed a duty of care to Mr Fenlon; (b) the suspect negligently breached that duty of care; (c) it was reasonably foreseeable that the breach of that duty gave rise to a serious and obvious risk of death; (d) the breach of that duty caused the death of Mr Fenlon; and (e) the circumstances of the breach were “truly exceptionally bad” and so reprehensible as to justify the conclusion that it amounted to gross negligence and required criminal sanction.

In relation to each of A and B, Mr Southam was satisfied that the first four elements, (a) to (d), were met. (He was not so satisfied in relation to C. In her case, he concluded that although she owed a duty of care which she breached, in circumstances where death was foreseeable, her breaches did not cause or contribute to his death.) However, he did not consider that the fifth element of ‘grossness’ was met on the facts in relation to A or B.

Mr Southam concluded that A’s principal failings were his failure to read the ACCT file, especially given that Mr Fenlon was a prisoner of whom A had no prior knowledge and that A placed too much reliance on what Mr Fenlon told him, so that A’s response was to raise the risk level and increase observations where he should have considered more radical action to protect Mr Fenlon, for example, raising the risk level to high, removing ligature points, requesting constant watch or making an urgent referral to the mental health team. On the other hand, A had alerted the Custody Manager to what happened and the Custody Manager concurred with A’s actions. Mr Southam concluded that A had made mistakes, but those mistakes did not fall so far below the standard required as to be gross, they were not “truly exceptionally bad”.

Mr Southam considered the failings by B. B also failed to read Mr Fenlon’s ACCT file, which he could have done on returning from lunch and before he saw Mr Fenlon, especially given that Mr Fenlon was a prisoner of whom B had no prior knowledge. He failed to increase the level of risk even though a noose was found in Mr Fenlon’s cell on 4 March 2016 where Mr Fenlon had tried to take his own life the day before. Mr Southam noted that if B had read the ACCT file he would have noted the concerns logged by other officers. He failed to check the CAREMAP created a week earlier which would have alerted him that one of the issues was mental health and that Mr Fenlon was waiting for an appointment with mental health services. He failed to check on progress of the actions in the existing CAREMAP, nor did he create a new one. B did not put in place more regular or continual observations, he left the observations at 2 per hour. He did not contact the mental health team. He did arrange for the next ACCT review to be brought forward to 5 March 2016 (from 8 March 2016). Mr Southam thought B’s culpability was higher than A’s, but B’s failings were not so “truly exceptionally bad” as to amount to gross negligence.

Mr Menon advances a number of challenges to the decision not to prosecute A and/or B for gross negligence manslaughter. He first argues that the identified breaches were very serious, were causative of death, and demonstrated indifference on the part of the officers to the obvious risk of death. He complains that neither A nor B has offered a credible explanation for these breaches. There has been a failure to take these considerations into account adequately. Alternatively, putting all these factors together, he submitted that the breaches were truly exceptionally bad. They met the test of grossness.

Secondly, he said that the Decision fails to give proper reasons for the conclusion that the breaches fell short of being grossly negligent. (This is the point which was not pleaded as a ground of challenge.)

Thirdly, he says that the Decision is defective (errs in law) in its failure to state in terms that grossness is supremely a jury question.

Fourth, he argues that the Decision is silent on the legal principle that multiple breaches by the same individual can combine to establish grossness and that there is an error of law evident on the face of the Decision in the failure to recognise this aspect of the law on gross negligence manslaughter. This criticism of the Decision was first made by the Claimant in her pre-action protocol letter dated 19 November 2020, citing Adomoko and Misra [2005] 1 Cr App Rep 21 at [2]-[16] and [23]). That elicited a pre-action protocol response letter dated 3 December 2020, in which Mr Southam answered this criticism and stated that in reaching his decision on A and B, he had considered whether the breaches were “singularly or cumulatively” truly exceptionally bad; the pre-action protocol response letter stated that “the Defendant agrees that each mistake need not be a gross breach, but that grossness can be established cumulatively having regard to all the circumstances”. Mr Menon challenges these statements, saying that they come two months after the Decision was made and they are too late to exculpate Mr Southam.

Fifth, he argues that Mr Southam erred in fact, because he based his conclusions on the Shaw Report of 2017 and referred to the second HMCIP Report. Both reports post-dated these events and neither dealt with matters at the time of Mr Fenlon’s death. The particular point of criticism is that the Shaw report identified a culture of risk aversion at the prison with ACCTs opened too readily, a situation which Mr Menon says did not exist in March 2016 when Mr Fenlon died but developed subsequently.

Mr McGuinness resists these various criticisms. He addressed them in roughly reverse order, as do I.

Mr Southam referred to the PPO Annual Report for 2016-17 and the Shaw Report at § 4.14 of his Review Note as part of a section on previous investigations of deaths at HMP Woodhill. He notes that the authors of both reports considered the ACCT needed to be reconsidered and states that the combination of poor staffing levels and the risk averse approach to opening ACCTs left officers inadequate time to complete ACCTs or read entries before conducting reviews. This comes as part of the Decision dealing with background, not part of Mr Southam’s reasoning on whether to prosecute A or B. It is not material to decisions made.

However, Mr Southam comes back to these two reports later in the Review Note, in the section which does address whether A or B should face charges for gross negligence manslaughter. In relation to A, he said at § 6.6: “As with the findings of the PPO and Stephen Shaw, [senior officer] A placed too much reliance on what Mr Fenlon said to him about future risk, rather than taking a holistic approach and considering all the risks … he under-assessed the level of risk…”. Of B he said at § 6.9: “Again [senior officer] B appears to have fallen in to the same error as [senior officer] A by placing too much reliance on what Mr Fenlon said to him, with the inherent dangers that creates, rather than considering that as one factor together with all the other available information…”. In my judgment, these are justified comments on the particular facts and evidence before Mr Southam, albeit drawing on aspects of what had been said in two reports which had examined the deaths at HMP Woodhill at different times. There is no error of fact here. Mr Southam was focussing on the particular events which led to Mr Fenlon’s death, as he was required to do.

On the point about aggregation of breaches, Mr McGuinness accepts that the Decision is not explicit in setting out the aggregation principle; but he says that aggregation was plainly in the mind of Mr Southam, who wrote of “breaches” and “decisions” by A, saying “they” were mistakes but “they” were not so truly exceptionally bad as to be grossly negligent. So, says Mr McGuinness, it is obvious that Mr Southam was considering A’s breaches in combination, or aggregated. Mr McGuinness makes a similar point in relation to B, noting reference to “a number of failings” but concluding that “they” were not so truly exceptionally bad as to amount to gross negligence. Mr McGuinness says that when that language from the Decision is put alongside the assertion in the pre-action response letter that Mr Southam did consider each officer’s breaches “singularly or cumulatively” and that he understands that grossness can be established “cumulatively with regard to all the circumstances”, the Court should be in no doubt. No error of law has been made.

I accept Mr McGuinness’ submissions on this point. On a fair reading of the Decision, it is clear that Mr Southam considered all the breaches by each officer, in combination as well as individually. The reassurance on this point given in the pre-action protocol response letter comes as no surprise and is consistent with what Mr Southam had written in the Decision. It is reasonable, in this case, for the Court to take account of what was written in the pre-action response letter, the point of which was to clarify criticisms or queries raised; and this was one such criticism on which Mr Southam had a clear answer. I am satisfied that Mr Southam considered grossness in the context of all the circumstances of the case and did not make an error of law.

Turning next to the alleged lack of awareness that grossness of the breaches was a question for the jury, I am satisfied that Mr Southam, an experienced prosecutor, was well aware of that. He had referred in his Decision and his Review Note to Adomako, a case which makes the point that grossness is ultimately a jury question.

At times, Mr Menon appeared to suggest that once the first four elements of gross negligence manslaughter were established, only the jury could determine the fifth element relating to grossness, and it was not for Mr Southam to take a view at all. That is obviously incorrect. Mr Southam needed to consider, in line with the Code for Crown Prosecutors, whether the jury were more likely than not to find that the breach or breaches, if established on the evidence, were indeed gross, as that term has been interpreted in case law and assuming a proper direction (he cited Adomako and R v Honey Rose [2017] EWCA Crim 1168). It was for Mr Southam to make that assessment.

That assessment was in the end a matter of impression. It did not require lengthy analysis or explanation. The Decision set out the breaches and their causal impact; it recited the case law on what gross meant in this context; it was permissible to state shortly whether, taking account of all the circumstances, either officer had been grossly negligent. The conclusions were adequately reasoned.

Finally, I deal with the Claimant’s challenge to the merits of Mr Southam’s decision, at times put as a failure to take account or sufficient account of various considerations, at other times as irrationality. The point made by the Claimant is that the failures of A and/or B were so bad as to be criminal, and so a jury was likely to have concluded, and so Mr Southam should have decided. On this point the Claimant’s case is weak. Mr Southam adequately identified the key breaches of duty by A and B. He applied the legal test of what amounted to grossness, taking account of all relevant facts and matters, and weighing the evidence appropriately. His conclusion that the mistakes made by A and/or B were insufficient to meet that high threshold was open to him.

So far as the Decision related to gross negligence manslaughter in the context of the actions and omissions of A and B, it was lawful.

Section 7 of the HSWA 1974 is headed “general duties of employees at work” and provides as follows:

The Health and Safety Executive (“HSE”) has published guidance on prosecuting individuals, known as Operational Circular (or “OC”) 130/8. It commences with reference to the Code for Crown Prosecutors which HSE inspectors are invited to apply. The document tells HSE inspectors about the principles to follow and the factors to take into account when considering prosecuting individuals under section 7, and other provisions, of the HSWA 1974. Paragraph 10 states:

Appendix 1 deals with the prosecution of individual employees under section 7. At paragraph 1, it is noted that a prosecution under section 7 requires that three elements can be proved: (i) that a person is employed, (ii) that a person is ‘at work’ in the course of his or her employment and (iii) either that that person did not take reasonable care for someone’s health and safety or did not co-operate with their employer so far as was necessary to enable their employer to comply with a statutory demand or requirement. Paragraph 2 of Appendix 1 states:

The Decision addressed section 7 of the HSWA 1974. Mr Southam noted breaches of duty by A, B and C. He noted that the prison was under-staffed at the time, the officers were cross-deployed dealing with prisoners they were unfamiliar with and that there was not continuity of case manager for ACCTs and they were dealing with many more ACCTs than were advisable. Then, he says:

The Review Note shows that Mr Southam considered the position under section 7 in relation to all three individual suspects. Mr Southam thought that C had a responsibility for Mr Fenlon but she was in a different category from A or B because although she had a part to play in the ACCT process she was not responsible for it. He concluded that there were management failings: little evidence of support available to those involved in ACCT processes, the prison was seriously understaffed, and they were cross-deployed dealing with prisoners they did not know, without continuity of ACCT case managers. He said:

Mr Menon criticises C for failing to make an urgent healthcare referral in circumstances where C had attended Mr Fenlon’s cell on 3 March 2016 and helped to cut him down from where he was hanging and Mr Fenlon had told C that he could not say he would not try to harm himself again. Further, he says that the management failures identified by Mr Southam do not touch on the position of C: she was not in charge of the ACCT process, she was not cross-deployed, she was not (so far as the evidence discloses) working in an environment affected by understaffing. He says that there is at least a reasonable prospect of conviction under the HSWA 1974 in her case.

Mr McGuinness disputes that. He says that the prosecutor must assess how the case against a particular defendant would be likely to fare and this involves assessment of the strength of the evidence and likely defences (Manning, [23]). He notes § 5.2 of the Review Note which records evidence that the prison only had 3 out of 6 MHT personnel available that day, all of them with appointments that morning, and § 5.7 of the Review Note which records that C was covering two wings that morning. He points to the explanation given in the pre-action response letter where Mr Southam says this:

He says that Mr Southam correctly took into account the HSE guidance and considered C’s actions in the context of her employer’s provision and support of her.

I confess to finding Mr Southam’s conclusion that there was insufficient evidence to support a prosecution under section 7 at first sight puzzling given his earlier conclusion that C was in breach of duty owed to Mr Fenlon. But on reflection I think Mr McGuinness is right to say that Mr Southam was entitled to take account of the wider context, including employer’s provision, when considering the prospect of a successful conviction under s 7 HSWA.

It would have been better if Mr Southam had included his full reasons about C in the Decision or at least in his Review Note. It is unsatisfactory that he expanded his reasons about C in the pre-action protocol letter. But in the end I conclude that it was open to Mr Southam to conclude that a prosecution against C was unlikely to succeed, for the reasons set out in the Decision, the Review Note and his pre-action protocol response letter which I accept as truthful. Mr Southam was required to consider the likelihood of conviction. Taking account of her likely defence (a denial of breach) and in the context of shortcomings at the prison including lack of staff and mental health specialists, and her own need to cover two wings that day, the prospects of a successful prosecution in her case were doubtful. C played a relatively small part in the sequence of failures leading to Mr Fenlon’s death and it was reasonable to conclude that this prosecution would not have fared well in front of a jury. That assessment was for Mr Southam to make, and it cannot be stigmatised just because others might not agree with it: see Manning [23], again.

The Decision set out the components of corporate manslaughter as follows:

That the way an organisation’s activities are managed or organised:

Caused a person’s death; and

An organisation can only be guilty if the way its activities are managed or organised by its senior management is a substantial element in the breach of the company’s duty of care”

This is a summary of section 1(1) and (3) of the Corporate Manslaughter and Corporate Homicide Act 2007 (the “2007 Act”). No issues arise as to the accuracy of the summary.

Mr Southam confirmed that both MoJ and NOMS were considered to be qualifying organisations, whereas HMP Woodhill was not because it was an executive agency of the MoJ. He noted that each organisation owed a duty of care to Mr Fenlon. This is not challenged. He noted that the MoJ and NOMS managed or organised their systems of protection of prisoners by maintaining the ACCT guidance with the “clear purpose of prioritising prisoner safety and seeking to put in place a system whereby those prisoners at risk of self-harm or suicide were appropriately supported”. He concluded that Mr Fenlon died because “the guidance was not adhered to rather than by the absence of measures being put in place by the senior managers.”

He said that he had considered whether staffing levels influenced the decisions of A, B and C. He referred to three reports (the PPO, the Shaw Report, and the report of HMCIP, the second I believe) which had identified staff shortages across the prison as a problem; he noted that the three individuals had alluded to staff shortages in their interviews, without raising that explicitly as an excuse for their mistakes; and he noted that the PPO report and the Shaw Report had questioned whether the ACCT process was fit for purpose. But how staff were deployed at HMP Woodhill was a matter for the Governor of that prison, and there was no indication that lack of staff was a significant factor in the decisions made by those individuals dealing with Mr Fenlon.

He noted the fact that the four ACCT reviews of Mr Fenlon were conducted by different officers who were cross-deployed and did not know Mr Fenlon, leading to a lack of continuity which may have been attributable to staff shortages. But he did not think it was possible to say whether a different decision may have been reached if there had been a consisted allocated case manager. He concluded that “the lack of consistent case manager, whilst inadvisable, cannot be said to have caused or significantly contributed to Mr Fenlon’s suicide.” This was to conclude that causation, one of the necessary ingredients of corporate manslaughter, was not made out.

He held that there was a “clear system in place” to manage prisoners at risk of self-harm, introduced by MoJ and NOMS, with training delivered at HMP Woodhill to implement that system. The ACCT document itself mirrors that guidance. But the evidence demonstrated that the guidance was not followed in this case.

He had regard to section 8 of the 2007 Act which directs juries to consider whether there was evidence of accepted practices which encouraged or tolerated breaches; but he concluded that the guidance was clear, but not adhered to; and if there were issues from earlier deaths to address, they were for HMP Woodhill which was not an organisation to which the 2007 Act applied.

He summarised:

The Review Note addressed corporate manslaughter in greater detail. In that document, Mr Southam thought it was “abundantly clear” that MoJ and NOMS had introduced comprehensive policies to safeguard prisoners who were at risk of self-harm or suicide, and training had been delivered to HMP Woodhill which mirrored that guidance: “The issue in this case, like a number of earlier suicides at HMP Woodhill, is that officers did not follow the guidance … senior managers have a right to expect that when experienced officers are trained in the ACCT process that they should comply with it” (§ 6.42). He noted that in assessing whether there was a gross breach by MoJ or NOMS, the jury would have regard to the factors set out in section 8 of the 2007 Act but thought that any such practices were for the prison to address; so far as MoJ was concerned, it had commissioned a report from Mr Shaw to consider the issues at HMP Woodhill (§ 6.44). In conclusion, he did not believe that there was a realistic prospect of conviction of the MoJ or NOMS for corporate manslaughter (§ 6.45).

Mr Menon challenges the decision not to prosecute the MoJ or NOMS for corporate manslaughter in a number of ways. First, he says that conclusion is irrational because there were glaring errors in the management of Mr Fenlon. The most obvious flaw was the lack of continuity of case manager: Mr Fenlon’s case was reviewed by four different officers within the ACCT process; if there had been a single case manager, as there should have been, different measures would have been put in place to protect Mr Fenlon, such as constant watch or an earlier referral to the mental health team. Secondly, Mr Menon advances his case on error of law (the aggregation point) in the context of corporate manslaughter also. Thirdly, he says that Mr Southam failed to take account of the important consideration of resourcing by the MoJ and NOMS, and without proper resourcing, safe ACCT management cannot occur. Fourth, he raises the error of fact (reliance on post-dating investigations such as the Shaw Report) in the context of corporate manslaughter also. Fifth, he says that Mr Southam has failed to give adequate reasons in the Decision.

Mr Menon also makes a wider point that Mr Southam’s conclusions are paradoxical. The individual suspects are avoiding criminal liability on the basis that there were systemic failings which affected their ability to care for Mr Fenlon adequately, for example, staff shortages; yet the organisations are avoiding criminal liability because they blame individuals for the errors made in this case and say that systemic failings were not the cause.

Mr McGuinness resists all of these points.

I have already dealt with a number of these points in the context of the second issue relating to gross negligence manslaughter. I am not persuaded that Mr Southam failed to recognise that breaches could be aggregated when considering grossness (second point), so there is no error of law in that regard. So far as reliance on the post-dating reports are concerned (fourth point), I am satisfied that Mr Southam referred to each in that part of the Decision and Review Note dealing with corporate manslaughter in a manner that was appropriate, to make the valid point that the issues identified on the facts of this case find an echo in these reports which post-date Mr Fenlon’s death. I am not persuaded that there was any error of fact in this regard. As to the adequacy of the reasons given (fifth point), Mr Southam explained at some length why he does not consider either organisation breached its duty to Mr Fenlon; his reasons are adequate.

The resources issue (third point) is presented as a failure to take account of a relevant consideration. The problem Mr Menon faces here is the conclusion by Mr Southam, on the evidence, that resources (in terms of lack of staff) were not the reason why the individuals did not follow the ACCT process properly; further, that even if there had been continuity of ACCT case manager in Mr Fenlon’s case, it is not possible to say that the outcome would have been any different. These conclusions were open to Mr Southam.

The irrationality argument, suggesting that there were gross breaches of duty by the two organisations, amounts to a disagreement on the merits. Mr Southam was not in the end critical of MoJ or NOMS: he did not think either organisation had failed in its duty of care to prisoners at HMP Woodhill. They had put in place a clear system to manage prisoners at risk of self-harm. Failings in implementation of that system were at an individual level. If the individuals in turn complained that their breaches were attributable to systemic failings, such as under-staffing or cross-deployment, or a culture of indifference or disregard of the system, then those were matters for HMP Woodhill to address. These are not irrational conclusions. They were conclusions open to Mr Southam, who took account of relevant evidence and matters.

In any event, Mr Southam did not think the failure of continuity of the ACCT manager itself was causative of death, and that posed a further obstacle to prosecution of the organisations. This conclusion was open to Mr Southam.

The paradox argument is irrationality put another way. It is predicated on the notion that someone must be held accountable in the criminal courts for Mr Fenlon’s death. That is not right. In some cases, even cases where there is a tragic outcome as there was here and even in the context of multiple deaths, it may be that no one bears criminal responsibility for what occurred.

Mr Southam was right to consider each suspect in relation to each possible charge, in light of the evidence as it would be adduced at trial against that suspect, and in light of the likely defence(s) to be adopted by each suspect, so as to assess the likelihood of a conviction. It was not his job to decide which, of the five suspects, should face criminal charges; that would beg the question whether any charges at all should be brought. His conclusion, that there was not a realistic prospect of conviction of any one of the five suspects, is not irrational or paradoxical. It was properly open to him.

It follows that this application for judicial review is refused. The Decision, in relation to each of the five suspects, is lawful. It lies within the margin of discretion conferred on the Defendant. It is not vitiated by any error of law, of fact, or any failure to take account of relevant considerations or to accord appropriate weight to those considerations. It is adequately reasoned.