Thirlwall Inquiry Sets Suspicion Rules for Neonatal Wards

Lady Justice Thirlwall published a 70-page summary and three-volume report on 15 September 2026 that blamed a complete failure of safeguarding at the Countess of Chester Hospital. Errors, she said, were made by nurses, doctors and managers. Some babies, she found, would have lived if the hospital had acted sooner.

The inquiry did not test whether Lucy Letby, 36, committed those crimes. It still told every NHS neonatal unit to treat suspicion of deliberate harm as enough to move a nurse and call in others.

The Inquiry Refused to Reopen the Verdicts

Letby is serving 15 whole-life terms after convictions for murdering seven babies and the attempted murder of seven others, one of them twice, on the Chester neonatal unit between June 2015 and June 2016. Two bids for leave to appeal were dismissed. She has always said she is innocent.

Thirlwall, an appeal court judge, was appointed after the first trial. Public hearings ran from September 2024 to March 2025. The terms of reference, set on 19 October 2023, asked how the hospital and outside bodies responded when babies collapsed and died, and what would keep infants safer now. They did not ask the inquiry to retry the case.

It is not an investigation into the criminal convictions or the guilt of Letby. The Court of Appeal has dismissed two applications for leave to appeal and it’s not for me to second guess or to cut across the work of the Criminal Cases Review Commission.

Lady Justice Thirlwall, chair’s final statement, 15 September 2026

She recorded each jury verdict, then looked at what staff knew at the time. She described dysfunctional management and governance, a gulf between hospital leaders and clinicians, and a complete failure at all levels to invoke safeguarding. Looking for a clinical explanation, she said, was not wrong. Once suspicion of deliberate harm existed, Letby should have been taken off the ward as a neutral step so an investigation could run without further risk to babies.

Health Secretary Yvette Cooper told the Commons the report was a shocking account of repeated mistakes. She apologised to the 13 families at the heart of the inquiry. She also repeated Thirlwall’s own limit: the inquiry was not there to reopen the verdicts or to cut across the CCRC.

Almost a Year Before Anyone Called the Police

The first death on the indictment, Baby A, came in June 2015. Unexplained newborn deaths are rare, and Thirlwall found that the sudden unexpected death in infancy and childhood process, known as SUDIC, applied to babies who had never left hospital. Staff at the Countess did not treat it that way. Police, she said, should have been called after Baby A.

They were not called until May 2017, almost a year after the last death.

THE MISSED TURNS AT CHESTER

1. June 2015: Baby A dies on the neonatal unit. Thirlwall finds the SUDIC process applied and police should have been contacted.
2. August 2015: Baby F’s insulin test flags insulin from outside the body. A doctor dismisses the result and no safeguarding file is opened.
3. October 2015: After Baby I dies, suspicions become clearer. Thirlwall says Letby should have been moved off the ward then. She was not, and collapses continued.
4. February 2016: Dr Ravi Jayaram, a consultant paediatrician, should have reported what he had seen with Baby K, the report finds.
5. May 2016: Concerns reach senior managers, including the hospital’s safeguarding lead. No safeguarding action is taken. Letby stays on the ward.
6. Late June 2016: Baby P dies. Letby is moved off the neonatal unit. Managers still delay calling the police and commission reviews that never ask whether someone was causing harm.
7. May 2017: Police are contacted, nearly a year after Baby P.

Thirlwall was blunt about the later window. If safeguarding had been raised at the May 2016 meeting, babies O and P should not have died. If Letby had been moved after Baby I, the deaths of O and P and the attacks on babies J, K, L, M and N would, she found, have been prevented. Senior nurses, she said, never accepted that the consultants’ concerns were, or even might be, justified. Neither did senior managers.

The Royal College of Paediatrics and Child Health accepted that its service review could not answer the doctors’ fears and should not have been done. Relationships between consultant paediatricians and senior managers fell apart through 2016 and 2017. Letby’s grievance was handled in a way Thirlwall called deplorable. Some consultants were told to apologise to her and to enter mediation. Plans were being made to put her back on the neonatal unit.

By early 2017, she said, the hospital was far adrift from a responsible response to serious concerns about the safety of babies.

Baby F’s Insulin Result Never Triggered Safeguarding

The insulin paper trail is the cleanest example of a missed stop. In August 2015, Baby F, a twin, returned an insulin level of 4,657 picomoles per litre and a C-peptide of 169 picomoles per litre. A senior clinical scientist at Liverpool Clinical Laboratories flagged a low C-peptide to insulin ratio and suggested the sample be sent on because the insulin looked exogenous, meaning it had come from outside the baby’s body.

Baby F had not been prescribed insulin. A doctor referred to in the report as Dr ZA dismissed the idea of deliberate administration, did not call the laboratories, and did not tell colleagues who already worried about a rise in deaths. Thirlwall said that if those results had been acted on, contact with the police would have been unavoidable. Baby F’s twin, Baby E, had died. Their mother later told the inquiry she had heard her baby scream and had seen blood on his mouth, with Letby there.

Insulin is not a rare drug on these wards. A January 2026 NHS Getting It Right First Time guide to safe insulin use in neonatal services recorded that 88 of 152 neonatal units in England, 58 per cent, used intravenous insulin at least once in 2024-25, across more than 8,000 days of use. Every intensive-care neonatal unit in that set had used it. The drug is given for high blood glucose in premature babies, and sometimes for high potassium. The same guide sits beside separate laboratory advice, issued in October 2025, on how to handle insulin and C-peptide requests when a baby’s blood sugar crashes.

Thirlwall wants that laboratory advice made mandatory across the country, so a result like Baby F’s cannot be shrugged off again. Until biometric locks are in place, she wants cameras pointed at the fridges and cupboards where insulin is kept, with recordings stored for at least 28 days.

Seventeen Rules Headed for Every Neonatal Unit

The report’s local findings are about one hospital in Chester. The remedies are national. Thirlwall issued 17 recommendations on neonatal safety, covering cameras, insulin, bereavement, contracts, data, SUDIC, medical examiners, managers and the Care Quality Commission. Cooper said the government would consider the whole report and already agreed with several of the most visible items, including cameras, stronger insulin storage, a manager barring scheme, and a national bereavement pathway.

The Countess unit now runs 13 cots as a local neonatal unit for babies born from 32 weeks, after a 2016 downgrade that followed the collapses. Cooper said services for women and children at the hospital have improved since the events the inquiry examined. The new rules are written for every trust, not for Chester alone.

THE DEADLINES THIRLWALL SET FOR THE NHS

Cooper has asked officials to draw up plans for cot cams in neonatal units so parents can watch their babies when they cannot be on the ward. Thirlwall said parents of very young babies would find that reassuring, and that it would deter the rare person who wants to cause harm. Had such monitors existed in 2015, she said, the attacks on babies A, B, C and D would probably have been detected. She made no criticism of their absence then. They were not in use at the Countess or elsewhere.

A camera on a cot will show who is at the bedside. It will not show pneumonia, a bowel obstruction, or a thrombosis. Those are the sorts of medical explanations a later expert panel put on several of the same collapses. The kit Thirlwall wants is built to watch people, which is a direct answer to a crime theory, and only an indirect answer to a unit that missed test results and never put baby deaths on the board agenda.

That board gap is one of the quieter findings, and one that does not depend on any jury. During the years the inquiry examined, the board and medical director reviewed adult deaths. They received no reports about the deaths of babies and children. Cooper quoted Thirlwall’s line that this was a serious failure of governance that no one on the board seems to have noticed, after a reorganisation that removed the voice of children and babies from the top table.

How the Suspicion Protocol Will Work

By 31 March 2027, NHS England must issue a one-page Suspicion of Deliberate Harm Protocol. Managers must act on a good-faith concern at once. It does not matter whether the person who hears the concern believes it, and it does not matter that the person raising it is not sure. Suspicion is enough. Moving the worker off the ward is likely to be the first step, and that move is to be treated as a neutral act. Safeguarding steps follow, and in many cases the police become involved. Only a concern that is obviously irrational or malicious, with that judgement written down, can be left alone.

WHAT THE ONE-PAGE RULE REQUIRES

• Belief is irrelevant: Managers cannot wait until they are sure, and they cannot reject a concern because they personally doubt it.
• Patients move first: The suspected worker is likely to be taken off the ward while an investigation runs.
• Police in many cases: Child safeguarding includes the local authority designated officer, and police involvement is built into the sequence.
• Contracts will say so: From March 2027, following the protocol is a term of NHS employment, including for agency and bank staff.
• Training for boards: All staff and board members, including non-executives, must be trained on suspicions of deliberate harm by colleagues.

Thirlwall said she wanted this to be beyond argument for the next manager who claims confusion.

Never again should any manager assert that they did not know what to do.

Lady Justice Thirlwall, chair’s final statement, 15 September 2026

On a ward where collapses are part of the work, that rule cuts two ways. It is the instruction Chester lacked when consultants were raising names and executives were commissioning reviews that never asked the harm question. It is also a standing invitation to send a colleague to the police on a suspicion that has not been tested, in units that already run on fear of the next unexpected death. A protocol that says only good-faith concerns count still has to be applied, in the moment, by the same professional groups Thirlwall found locked in what she called tribalism: senior nurses certain the consultants were wrong, consultants certain a nurse was dangerous, managers protecting the hospital’s name.

A revised NHS safeguarding framework was published in April 2026. Cooper has asked the chief nursing officer to review that framework and its training again in light of the report. Freedom to Speak Up arrangements, she and Thirlwall both noted, did not carry this load at Chester. Putting the duty into contracts is meant to force the issue down to every employee, not leave it with a guardian.

Families Were Left in the Dark for Years

Almost all of the parents gave evidence. Their babies cannot be named because of Crown Court orders. Thirlwall asked everyone commenting on Chester to remember that these are real people, still living with collapses and deaths from 10 and 11 years ago, and that families must not become collateral in the public fight over Letby’s guilt.

Parents were not told for years that there were concerns their babies might have been harmed on purpose. Their consent was not obtained before medical records went to outside experts and other organisations. They were not informed about reviews. Only in 2018 did they learn that the collapses and deaths might have been the work of a nurse they had trusted. Many concluded that protecting the hospital’s reputation mattered more to senior managers than keeping them informed. Thirlwall called the lack of consideration reprehensible.

Cooper put one parent’s words on the Commons record.

I cannot emphasise enough the impact of this on our whole family, who we are as people, parents, work, life spouses, children. It has cast a shadow of sadness over every part of our lives.

Mother of Baby J, quoted by Health Secretary Yvette Cooper, House of Commons, 15 September 2026

Bereaved parents, Thirlwall said, should never again be handed a leaflet with no support, or ignored. The National Bereavement Care Pathway for neonatal death is to be in every trust by 31 August 2027. Cooper said all trusts are already signed up and that she will keep the requirement in the NHS planning framework.

Duty of candour failed in the other direction too. The inquiry found repeated failures of governance and candour by the trust, and what Cooper called an exercise in spin to steer away from a police referral. Executive presentations to the board, in the report’s account, dressed the problem rather than naming it. Chief executive Tony Chambers, the findings say, added to the delay in contacting police. Director of nursing Alison Kelly, who was also the safeguarding lead, knew suspicion of harm to a baby required action and did not act.

A 14-Expert Panel the Inquiry Would Not Hear

The same week the report landed, Letby’s case remained where Thirlwall said it belonged: with the Criminal Cases Review Commission, not with her. A review of Letby’s convictions is underway. A preliminary application arrived on 3 February 2025. Expert reports, main submissions, a waiver of privilege and further papers followed through 2025 and into January 2026. The CCRC’s job, its chair has said, is not to decide innocence or guilt. It is to ask whether new evidence or a new argument means there is a real possibility a conviction would not be upheld.

In February 2025, Dr Shoo Lee, professor emeritus of paediatrics at the University of Toronto, presented a 14-member international panel’s reading of the medical files. Lee’s 1989 paper on air embolism in newborns had been used by the prosecution at trial. He told a London briefing that the panel found no medical evidence of murder or of air injected into veins, and that death or injury in each case was due to natural causes or poor medical care.

We did not find any murders. In all cases death or injury were due to natural causes or just bad medical care. In our opinion, the medical evidence doesn’t support murder in any of these babies.

Dr Shoo Lee, professor emeritus of paediatrics, University of Toronto, London briefing, 4 February 2025

The panel included senior neonatologists from Imperial College London, the Karolinska Institute and Harvard Medical School. A separate group of experts, instructed for the CCRC application, has challenged the insulin readings that underpinned two of the convictions. Thirlwall did not hear that alternative account. She had, at the start of the inquiry, declined to pause or widen the terms to take in the safety of the convictions, and she was still declining on 15 September 2026.

WHAT WE KNOW

• The verdicts: Letby’s convictions stand. Two appeals failed. The CCRC has not referred the case back to the Court of Appeal.
• The hospital record: Insulin results were not treated as safeguarding. SUDIC was not used for babies who never went home. The board did not see baby deaths. Police were called in May 2017.
• The new rules: Cot-camera plans, insulin locks, a one-page suspicion protocol and contract changes are on a 31 March 2027 clock.

WHAT IS UNCONFIRMED

• The CCRC outcome: No referral decision had been published by 19 September 2026, and further papers can still be added to the file.
• How trusts will apply suspicion: The protocol does not yet exist. Whether “good faith” will stop weak or vengeful referrals is untested.
• Whether cameras change outcomes: Thirlwall’s detection claim about babies A to D is a judgement on a crime theory, not a trial of camera systems on live wards.

Chester failed in ways that would be a public health problem on any theory of these deaths. A board that never saw baby deaths, a laboratory flag that never became a police call, and parents left uninformed for years are failures of candour and of basic monitoring. The 17 recommendations mix those repairs with tools designed for a killer on the unit: livestream cameras, biometric insulin locks, and a rule that suspicion of a colleague is enough to bring in the police.

Nurses will work under that mix from 2027 whether or not the CCRC ever sends Letby’s file back to the Court of Appeal. The inquiry that would not test the verdicts has already written the next set of rules for the ward.

Disclaimer: This article is news reporting and analysis of a published public inquiry and related criminal proceedings. It is for information only and is not medical advice, legal advice, or a view on any person’s guilt or innocence. Readers who need advice about neonatal care, safeguarding, or a criminal case should consult a qualified doctor, safeguarding professional, or solicitor before acting. Figures, deadlines and case statuses are those given in the inquiry report, government statement and CCRC notices cited here, and they may change as reviews and NHS plans move on.