Peter Bryan killed three people in three separate events. In March 1993 he killed a young woman in Chelsea, west London. In February 2004 he killed a man in Walthamstow, east London. Two months later, as a patient in Broadmoor Hospital, he attacked a fellow patient, who died of his injuries in June 2004. Each case ended in manslaughter on grounds of diminished responsibility, because each time the courts accepted that a severe mental illness had substantially reduced his responsibility.
The legal record is the center of the case. Bryan was sent to a secure hospital in 1994 under a restriction order, was released to a supervised hostel in 2002, and was in the care of mental health services when he killed on both occasions in 2004. Two independent reports were later commissioned by NHS London to examine his care, and both found failings at several levels, though the inquiry into the Walthamstow killing concluded that it could not reasonably have been predicted. In 2005 the Old Bailey sentenced him to life with no possibility of release; the Court of Appeal in 2006 replaced that with a minimum term of 15 years while saying that it was hard to conceive he would ever be released.
Peter Bryan was born in London on October 4, 1969, the youngest of seven children. His parents had come to England from Barbados in the 1950s and had left four of their older children in the care of a relative there. The family moved to Newham when he was four or five, and it stayed their home. His mother worked and relied on childminders.
The independent inquiry into his care, which drew on his own account, described an unsettled school career. He attended several schools, was suspended once and was reprimanded for bullying. By his own account he was part of a gang from about the age of 12 and first used drugs at that age. He was later found to be dyslexic, in 2002. Reports differ on when he left school, at 14 or 15. His only conviction before 1993 was for possessing a controlled drug, for which he received a conditional discharge at Snaresbrook Crown Court on October 28, 1992.
After school he took a Sunday job on a clothes stall in Petticoat Lane market, which led to work in shops owned by the family of the woman he would later kill. He was paid in cash, on an irregular basis, and for a year at 19 he attended West Ham College, where he passed basic English and mathematics examinations. He had no contact with psychiatric services before 1993. He was later found to have been suffering from paranoid psychosis, or paranoid schizophrenia, at that time.
Nisha Sheth was 20 when she died in March 1993, in Chelsea. Bryan had worked for her family. The inquiry into his care did not examine her death, and beyond her age and her family's connection to him, little about her life has been made public. A member of her family was also wounded in the same event.
Bryan was arrested in hospital on March 20, 1993. He was held on remand at Brixton Prison from April 30, 1993. In October 1993 a consultant forensic psychiatrist assessed him and recommended admission to Rampton Hospital, the high-security hospital in Nottinghamshire; a Rampton consultant assessed him on November 18, 1993 and recommended admission under section 35 of the Mental Health Act 1983. He was admitted on December 17, 1993.
At the Old Bailey on February 25, 1994 he pleaded not guilty to murder but guilty to manslaughter on grounds of diminished responsibility. He also pleaded not guilty to wounding with intent, involving a member of Nisha Sheth's family, and guilty to the lesser offense of wounding. On March 4, 1994 he was convicted of both and sent to Rampton under sections 37 and 41 of the Mental Health Act 1983: a hospital order, with a restriction order that left decisions about his release in the hands of the Home Secretary and the tribunal system and allowed him to be recalled to hospital if he was discharged.
Bryan stayed at Rampton from December 1993 until July 2001. The inquiry reconstructed the period from the hospital's records. Staff recorded threatening and abusive behavior early on, usually brief. His responsible consultant concluded that he had a paranoid psychosis made worse by stress. Bryan had never been given antipsychotic medication before he reached Rampton; a depot antipsychotic was started in May 1994.
In August 1994 the consultant decided that Bryan showed no symptoms of mental illness and directed that the medication be gradually withdrawn. His behavior deteriorated, and in July 1995 an oral antipsychotic was prescribed. In mid-1996 the notes record that he asked to come off medication, and the consultant observed that he was less disinhibited while taking it but disliked it. In April 1997 two psychologists produced a 36-page report that proposed a drug-free or drug-reduced period and observed that it was questionable to what extent he was being managed by his medication. At a treatment planning meeting in September 1997 staff noted that his condition had worsened on two earlier occasions when medication had been stopped.
The inquiry summarized the hospital years as long periods of settled behavior alongside spells of threatening conduct, poor boundaries with women and a fluctuating mental state, with deterioration following each withdrawal of medication. A referral to the John Howard Centre, a forensic unit in north London, was made in 1999, and the transfer took two years.
A mental health review tribunal heard his case on March 28, 2001. It found that he was no longer ill enough to warrant detention and that he was entitled to a conditional discharge, and it adjourned the decision until suitable arrangements could be made. The inquiry noted that later interpretation of this decision had major consequences.
On July 12, 2001 Bryan was transferred to the John Howard Centre on six months' trial leave. The inquiry criticized how it was done. He went directly onto Colin Franklin Ward, a locked pre-discharge ward, instead of the unit where patients from high security were usually admitted first. The panel thought this set the course for his later management. He stayed about six months, against an expected 18 to 24 months for patients moving from high security, and no single document gathered what Rampton had learned about him. The inquiry also found that a detailed psychologist's report was not used in treatment plans or in the tribunal process.
On February 12, 2002 he was conditionally discharged to Riverside House, a nine-bed supervised forensic hostel in Stoke Newington, north London, staffed around the clock. The panel had no strong criticism of the decision to discharge him. As a condition, the John Howard Centre team remained responsible for his care for six months, and his consultant there became his supervising psychiatrist in the community, with an experienced approved social worker as his social supervisor and a forensic community psychiatric nurse as care coordinator.
The arrangements then changed. The first social supervisor handed the case to a social worker who had about five months' experience, no mental health training and no approved status. The panel said the social worker should never have been asked to take responsibility for someone like Bryan and placed the fault with senior management; neither the psychiatrist nor the nurse knew how little experience he had. On September 17, 2002 the supervising psychiatrist role passed to a general adult psychiatrist in his first consultant post, with a caseload of about 70 and no experience of forensic patients, who accepted it with great reluctance. Bryan was allocated to him only because his parents lived in that psychiatrist's catchment area. The inquiry found a lack of effective management of Bryan in the community.
Bryan complained often about his medication and about the hostel. Early in 2003 his antipsychotic was changed to another drug. On April 28, 2003, five days before the psychiatrist left for five weeks of leave, the dose was halved. The panel found no explicit reason recorded, and considered that the reduction tipped the balance and that a subtle deterioration began around June 2003.
Toward the end of 2003 staff recorded more erratic behavior, paranoid complaints and disturbed sleep, which the panel said should have been assessed together instead of one by one. On November 19 a random urine test was positive for a non-prescribed drug; Bryan disputed it, a repeat test gave the same result, and no further test is recorded. In late November he began a cleaning job without the Home Office being told, and was allowed to manage his own medication around the same time. A test on December 29 was again positive, and a second result was inconclusive.
On January 9, 2004 the social worker wrote to the Home Office asking to move Bryan to a hostel supervised only from 9 a.m. to 5 p.m., despite the concerns about relapse. Later that month Riverside staff reported damage to his room, a theft allegation with little evidence, a police stop and search, and drunkenness. On January 22 his room was found in disarray, and the hostel manager ended his self-medication. On January 26 the John Howard Centre told the social worker that Bryan had applied to a tribunal for an absolute discharge, which would have ended the conditions and the power of recall. The hearing was set for March 1, 2004. His solicitor had first seen him about legal aid on November 21, 2003 and had not expected the application to succeed at that stage.
On January 27 the social worker acknowledged that he had the option of recalling Bryan to hospital under section 42(3) of the Mental Health Act, but accepted Bryan's plausible explanations. On January 28 Bryan returned from night work and said he had used a drug; no drug test followed. On February 2 an emergency meeting was held with the psychiatrist, the social worker, the hostel manager and Bryan. The inquiry recorded that the meeting was dominated by his passport application. The psychiatrist had advised an increase in medication on January 29 but did not make one; he told the family doctor that the changes might indicate early relapse and planned a review in three weeks, which the panel thought too long.
On February 6 an allegation that Bryan had indecently assaulted a 17-year-old girl was reported to the hostel. The nurse and social worker did not see him until Monday, February 9, and neither the psychiatrist nor the Home Office was told until February 10. The panel was firmly of the view that waiting was not good practice and said he should have stayed at Riverside House over the weekend. On February 10, after threats were made against him by others and the police were informed, he was admitted as an informal patient to Topaz Ward, an open mixed-sex acute ward in east London, mainly for his own safety. The panel found it totally inappropriate to use the ward as a safe house, given that it was mixed and he had just been accused of assaulting a girl.
On the ward he was kept under observation, was not allowed to leave, and was given antipsychotic medication each night. He appeared settled, friendly and compliant, and showed no psychotic symptoms. On February 17 he was given permission to leave the ward for the first time.
About three hours after leaving the ward Bryan killed Brian Cherry in Cherry's home in Walthamstow, east London. Cherry was 43 according to the reports of the 2005 sentencing; the inquiry gives his age as 45. He had lived alone in a council flat in Walthamstow for four or five years and before that with his parents, whom he visited most evenings. The inquiry describes him as a man much loved and missed by his family, and records that his brother called him "a sweet man" and a brother-in-law said he was "just a big softie really" and "didn't deserve what happened to him." Press reports of the sentencing describe Cherry as a friend of Bryan.
He was arrested, held at Pentonville and then at Belmarsh, and the inquiry noted that his condition deteriorated, with several serious incidents at Belmarsh.
The killing came twelve days before the tribunal hearing at which Bryan was to ask for an absolute discharge.
Belmarsh referred Bryan to Broadmoor Hospital on March 23, 2004. The independent inquiry into the Broadmoor killing found that the assessments before admission were limited. He arrived on April 15, 2004, was placed in seclusion, and on April 19 was moved to general observations and allowed to mix with other patients on Luton Ward, with staff checks about every 15 minutes.
Richard Loudwell was 59. He had been transferred to Broadmoor from HMP Belmarsh on January 15, 2004, after killing an elderly woman in Kent in December 2002. A separate inquiry by Verita for Medway Primary Care Trust and Medway Council records that Loudwell pleaded guilty to manslaughter on grounds of diminished responsibility at Maidstone Crown Court on April 22, 2004, and was detained at Broadmoor under an interim hospital order. That inquiry examined the mental health, probation, police and social care services in his case before his offense and criticized the community team and the failure to give him enhanced support on discharge in 2002. At Broadmoor, the NHS London inquiry found, he was bullied by other patients for the whole of his three months on the ward.
On April 25, 2004 Bryan attacked Loudwell on Luton Ward. The attack took place in a part of the ward that staff could not see into while staff were elsewhere on the ward. Loudwell was seriously injured. He died on June 5, 2004 without regaining consciousness. His sister later said that the family had not forgotten why he was detained, that another family still lived with the effects of what he had done, and that they had expected both the public and Loudwell himself to be kept safe once he was in custody.
Bryan was indicted for the murder of both men. On March 15, 2005 at the Central Criminal Court he pleaded guilty to two counts of manslaughter on grounds of diminished responsibility, and the prosecution accepted the pleas. Five reports from four psychiatrists were before the judge. It was agreed that he had no defense of insanity, was fit to plead and had diminished responsibility. The diagnosis was schizophrenia, described by three of the four psychiatrists as paranoid schizophrenia with unusual features, including an absence of visible symptoms during relapse. One psychiatrist, Dr Martin Lock, told the court that Bryan was probably the most dangerous man he had ever assessed, and that Bryan had been able to appear calm and normal. Another judged him to be extremely dangerous, with no remorse.
The judge, Giles Forrester, accepted that Bryan's responsibility for each killing was substantially impaired by mental illness. He found no exceptional circumstances. He imposed concurrent life sentences and directed that the early-release provisions should not apply. In his remarks, as the Court of Appeal later quoted them, he said that he had come to the conclusion in the circumstances of this quite exceptional case that Bryan must never be released from prison. He considered a hospital order and rejected it; the Court of Appeal noted that the mandatory provisions of section 109 of the Powers of Criminal Courts (Sentencing) Act 2000, which applied because of the 1994 manslaughter conviction, in any case made a hospital order unavailable. The prosecutor was Aftab Jafferjee. Bryan's counsel, David Etherington QC, told the court that he was the victim of a terrible illness. A report of the hearing says the court heard that the mental health system had let the public down.
Bryan appealed against the whole-life direction only. On January 31, 2006 the Lord Chief Justice, Lord Phillips of Worth Matravers, sitting with Mr Justice Cresswell and Mr Justice Openshaw, varied the sentence. The judgment, reported under the initials P.B., accepted that the harm was immense and the circumstances terrible. It held that his culpability was very considerably diminished by severe mental illness and that the sentencing judge had not adequately reflected this, treating features that were symptoms of the illness as increasing the gravity of the offenses.
The court said the question was how long justice demanded custody to punish offenses committed while ill, putting aside the risk of future dangerousness, and it noted that the longest sentence in earlier cases of manslaughter on diminished responsibility was nine years and that none had involved a series of killings. It found that determinate sentences of 15 years for each offense, served consecutively, would have been appropriate, and halved the total, giving a minimum term of 15 years. No credit was allowed for time on remand, because the court treated him as being on license, given the earlier hospital order, when the second offense was committed. The court stressed that the minimum term did not mean release at the end of it: his mental condition would be kept under review and he would remain in custody while he remained a risk. It added that it was hard to conceive that he would ever be released.
NHS London, then the strategic health authority for the capital, commissioned the independent inquiry into the Broadmoor killing in 2005, and a second into his care before the Cherry killing. The report into Bryan's care and the Cherry killing was chaired by Jane Mishcon, a barrister specializing in clinical negligence, with Dr Tim Exworthy, a consultant forensic psychiatrist, Stuart Wix, a forensic nurse consultant, and Mike Lindsey, a former deputy director of social services; Professor Tom Sensky advised the panel and Derek Mechen of Verita coordinated. The report states that it was handed to NHS London in early October 2008, and NHS London decided to anonymize the professionals named in it. A separate inquiry examined the care of Bryan and Richard Loudwell at Broadmoor. A solicitor for the Loudwell family said the family had not been involved in the inquiry. The reports were published in September 2009; the solicitors' account gives September 3, 2009.
The panel's central conclusion was that no individual professional's failure directly caused the killing and that only Bryan was responsible for it. It found instead a systemic failure: a general adult psychiatrist and a social worker with no mental health training should not have been asked to take responsibility for a patient with his history, and senior management allocated them. The community team worked as individuals, not as a team, so concerns about his behavior were handled one by one and their cumulative weight was never assessed. Bryan, the panel said, was not fully open with his carers and was manipulative; his written complaints and demands absorbed staff time that should have gone to assessing his mental state.
The panel listed specific failings: the transfer directly to Colin Franklin Ward and the short stay at the John Howard Centre; the failure to use the psychologist's report; the reduction in medication without a robust reason shortly before the supervising psychiatrist's leave; the weak response to the positive drug tests and to self-medication and unreported work; delay and drift after the allegation of February 6, 2004, including the failure to consider recall under section 42(3); and the use of Topaz Ward as a safe house.
On the question of whether the killing could have been prevented, the panel was cautious. His illness presented atypically: he did not show the classic signs of schizophrenia and appeared normal even when unwell, so the risk was hard to assess, and the panel said this was recognized only with hindsight. It concluded that it was unlikely anything other than a recall to hospital under section 42(3) could have prevented the killing, that such a recall would not have been easy to arrange, and that it would probably have been temporary and might only have postponed a later homicide. The panel's main lesson was that the management of restricted patients should be placed only in the hands of professionals with sufficient experience and training. It stated that its criticisms were shaped by hindsight and by access to every document and witness, which those involved at the time did not have.
The inquiry into the Broadmoor killing found that no single person was responsible. Shortcomings at many levels of care and within the trust, West London Mental Health NHS Trust, combined to produce Loudwell's death. Observations did not meet the trust's own policy: staff did not always know where patients were, and patients were out of sight during association. Loudwell had no care plan for the risk from other patients in his first week, and the plan written later was poor and largely ignored. His complaints of bullying were not taken seriously; he reported physical abuse to a Mental Health Act Commissioner three days before the attack, but his observation level was not raised.
Bryan, the inquiry found, was not properly risk-assessed. He had no mental state examination in the week before the attack and no medical contact after he left seclusion; he should have been treated as highly dangerous until proven otherwise. The report described weak management at every level, poor morale, a reactive nursing culture and a failure to act on concerns raised by staff and by the Mental Health Act Commission, and called it a collective failure. Security advice was not well integrated into ward practice, and the critical incident review missed that no staff were in the dining room or dayroom during the attack. The Loudwell family's letter went unanswered for nearly three weeks, and the chief executive's reply contained no apology.
An inquest into Richard Loudwell's death resumed on August 31, 2011 at Windsor before a jury, with the Berkshire coroner, Peter Bedford. The jury returned a narrative verdict. It found that the area where the attack took place was not adequately observed, and that Bryan had not been adequately examined when he came out of seclusion and should have been monitored more closely. The coroner said he would make no recommendations, citing the earlier inquiry and the trust's adoption of 82 recommendations. The family's solicitor, Kate Maynard, had said before the hearing that the family looked forward to an open and transparent inquest at which they could hear from Broadmoor staff and management in person.
The Court of Appeal's 15-year minimum term followed the sentence of March 2005, and the judgment said that Bryan's mental condition would be kept under review and that he would remain in custody while he remained a risk to the public. A summary of the sentencing appeal in the Law Gazette recorded that the court thought he was unlikely ever to be released. The inquiries' recommendations were accepted by the trusts concerned; at the Loudwell inquest in 2011 the coroner noted that the Broadmoor trust had adopted 82 recommendations.
Whether the Cherry killing could have been avoided is a matter on which the inquiry was carefully hedged. The panel found no single failure that caused it and a systemic one that made it more likely, and it said that a recall under section 42(3) might have prevented it but might only have postponed it. How his illness should count in sentencing is the question on which the sentencing judge and the Court of Appeal differed: the first treated signs of the illness as raising the gravity of the offenses, the second as lowering his culpability. All three deaths ended in manslaughter on diminished responsibility, and none of the courts or inquiries attributes any further killing to Bryan.
October 4, 1969: Peter Bryan is born in London, the youngest of seven children.
October 28, 1992: conditional discharge at Snaresbrook Crown Court for possessing a controlled drug.
March 1993: Nisha Sheth, 20, is killed in Chelsea; Bryan is arrested in hospital on March 20.
December 17, 1993: admitted to Rampton Hospital.
February 25, 1994: pleads guilty to manslaughter on grounds of diminished responsibility and to wounding.
March 4, 1994: convicted at the Old Bailey; hospital order with restriction order (Mental Health Act 1983, sections 37 and 41).
March 28, 2001: a tribunal finds him entitled to conditional discharge and adjourns the decision.
July 12, 2001: moves to the John Howard Centre.
February 12, 2002: conditionally discharged to Riverside House, Stoke Newington.
April 28, 2003: his antipsychotic dose is halved.
January 26, 2004: told that his application for absolute discharge is set for March 1.
February 6 to 10, 2004: an allegation is reported; on February 10 he is admitted informally to Topaz Ward.
February 17, 2004: given leave for the first time; Brian Cherry is killed about three hours later.
March 23, 2004: Belmarsh refers him to Broadmoor; he arrives on April 15.
April 25, 2004: attacks Richard Loudwell, 59, on Luton Ward.
June 5, 2004: Richard Loudwell dies.
March 15, 2005: pleads guilty to two counts of manslaughter on grounds of diminished responsibility; two concurrent life sentences with early release disapplied.
January 31, 2006: Court of Appeal replaces the whole-life direction with a minimum term of 15 years.
September 3, 2009: NHS London inquiry reports are published.
August 31, 2011: Loudwell inquest resumes at Windsor; the jury returns a narrative verdict.