Who Beverley Allitt was
In 1991 Beverley Allitt was a newly-qualified State Enrolled Nurse on the children’s ward at Grantham and Kesteven Hospital. Over a 58-day period that year, four children on her ward died and nine more suffered serious harm. Allitt was identified through direct clinical-forensic evidence: anomalous potassium and insulin values on specific patients; a stolen nursing Kardex record that was later found at her flat; eyewitness accounts of her proximity to each event. She was convicted in 1993 of four murders, three attempted murders, and a range of causing-grievous-bodily-harm offences, and sentenced to thirteen life terms.
The 1994 Clothier Inquiry into the case produced detailed recommendations on how hospitals should respond to clusters of unexplained deaths. Its central finding was that Grantham management had been too slow in recognising the pattern and in escalating.
How Allitt shaped the Hummingbird investigation
When consultants at the Countess of Chester began raising concerns about the cluster of unexplained neonatal deaths on their unit in 2015 and 2016, the Allitt case was, for everyone involved, the recent historical reference point. The published Hummingbird materials and the anonymous 150-page Hummingbird whistleblower report (see our summary) together make clear that the investigation, from the day Cheshire Police were contacted in May 2017, was framed as a “another Allitt” case.
That framing had operational consequences. In particular:
- Suspect-first not cluster-first. An Allitt-style investigation starts from an identified suspect and works outward. A cluster-first investigation starts from the unexplained events and works inward, systematically eliminating non-criminal explanations (outbreak, infrastructure, staffing, gestation profile) before narrowing to criminal hypotheses. Hummingbird, by all public accounts, began in the former mode.
- Expert instruction shaped by the frame. Dr Dewi Evans was instructed on the basis of the cluster having already been framed as a possible criminal pattern. His subsequent causation methodology — forensic rather than differential-diagnosis — was consistent with that framing.
- Escalation dynamics. Once an Allitt frame is adopted, independent expert challenge — the natural-causes, systems-failure hypotheses — becomes institutionally harder to pursue. Evidence that would fit a systems-failure explanation (the superbug outbreak, the sewage and plumbing failures, the doctor-shortage picture, the infrastructure Datix record) is there, but it does not become the primary narrative.
Why the Allitt analogy is imperfect
Rev.Dr Phil Hammond and others have written publicly about the ways the Allitt case differs from the Letby case on the fundamental facts:
- Direct forensic evidence. Allitt was convicted on anomalous potassium and insulin levels that were confirmed by forensic-standard testing. The Letby insulin samples were screening immunoassay results without confirmatory mass spectrometry (see evidence: insulin).
- Physical exhibits. In Allitt, a stolen nursing Kardex was physically recovered from her flat. In Letby, the prosecution relied on handover sheets that the nurses were specifically instructed not to dispose of (see evidence: handover notes).
- Eyewitnesses. Allitt was observed at the scene of individual incidents by clinical staff. Letby’s proximity is a function of the shift chart and the eyewitness evidence on Child K, which independent neonatologists regard as explicable without deliberate harm.
- Patient population. Allitt’s patients were older children, some of them admitted for relatively minor problems. Letby’s patients were extremely preterm neonates whose mortality risk was high without any intervention.
- Cluster context. Allitt’s Grantham ward was not experiencing a concurrent outbreak, infrastructure failure, or Level-of-care mismatch. The Countess of Chester neonatal unit was experiencing all three.
What the CCRC application says about the framing
The October 2025 Mark McDonald KC application to the CCRC (see our summary) addresses the investigation-framing point directly. Its argument is not that Hummingbird investigators acted in bad faith but that the frame under which the investigation operated was confirmation-bias-inducing: once the hypothesis “another Allitt” is in place, the evidence that fits the hypothesis is weighted more heavily, and the evidence that does not fit — the systems-failure evidence — is structurally discounted.
Why this matters for conviction safety
If the investigation was framed in a way that inhibited genuine consideration of non-criminal explanations for the cluster, the criminal case that emerged is, by construction, an incomplete treatment of the evidence. Correcting that gap is a legitimate question for the CCRC. This is not a claim that there was no reasonable basis for opening a criminal investigation in May 2017; it is a claim that the investigation that followed did not adequately test its own hypothesis.