Chapter 34. Insulin

On a few occasions over decades nurses have killed patients by poisoning them with insulin. UK examples include the cases of Beverly Allitt, Victorino Chua and Colin Norris. It was one of the methods of killing discussed by the RCPCH report team during their investigation at the Countess.
There are two separate issues: first, access to insulin; second, how tests are sought and how results are reported.

Storage and access to insulin

Insulin is not a controlled drug and not subject to the same very tight processes and procedures. No one in this Inquiry has suggested that it should be. All accept, however, that there should be effective systems to prevent and detect unauthorised access to and use of insulin.
On behalf of Family Group 1, it was submitted that access to the administration of insulin on neonatal units should be restricted and more effectively controlled. It should not be left to individual hospitals to institute controls over access to insulin. If this happens, it will result in inconsistent or non-existent safety standards.¹ It was submitted that CCTV should cover drug dispensary areas within the hospital, so dispensing insulin can be monitored at all times to deter and catch malevolent use.²
Family Group 1 also submitted that the relevant unit in DHSC (that takes over policy responsibilities for the administration of insulin from NHS England) should mandate that hospitals implement standard safety measures, such as electronic access and records. This will ensure that everyone who accesses insulin in hospitals can be identified and their actions can be checked where necessary.³
On behalf of Family Groups 2 and 3, it was submitted that the storage areas should be monitored to ensure that those who access them can more easily be traced.⁴
NHS England suggested automated access to drug dispensaries by the submission of biometric data, such as fingerprint or retina scans, rather than CCTV covering those areas.⁵
On behalf of the Countess, support was expressed for the expansion of systems designed to encourage best practice in the administration of medications and the storage of drugs on neonatal units.⁶
Mr Dzikiti, principal witness for CQC, acknowledged that it would be beneficial were there to be electronic access to medicines, but made it plain that, as before, CQC did not take a position on this or any other recommendations.⁷

The guide to safe insulin use in neonatal services

The guide, issued in January 2026, sets out a delivery checklist for safe access, storage, handling, prescription, administration and disposal of insulin.⁸ This includes individually identifiable and auditable access control (probably swipe card access), locked medicine fridge or cupboard (the expectation of an automated drawer with digital access on units within six months to two years). It also covers professional practice – the need for a local policy and clinical guidance, appropriately trained staff who complete specific neonatal insulin training every three years, mandating two-person checking for insulin preparation and administration, with at least one checker having completed neonatal nurse specialty training (QIS), a documented standardised approach for preparing and administering insulin infusions, and working towards implementing ‘closed loop’ administration systems. It states there should be “appropriate pharmacy staffing support as per NPPG [Neonatal and Paediatric Pharmacy Group]/ BAPM standards and … network pharmacists as recommended in GIRFT Neonatology national specialty report”.⁹

Technology

There is broad agreement about this issue. Technology allows us to be tracked in every aspect of our lives – transport, health, work, school, shopping, entertainment. It can easily be used to track who has access to which cupboard in a hospital and when. It requires only the installation and proper operation of the right technology. I recommend that digital devices be used to: restrict access to authorised people; and record access to insulin storage units. I acknowledge that steps are being taken to achieve this. It should be mandatory for all neonatal units to meet the expected requirements for access control and storage of insulin on neonatal units set out in the NHS’s Getting It Right First Time guide within six months.¹⁰
Until such time that access requires the provision of biometric data, there is always a risk that a swipe card or similar device may be swapped, inadvertently or deliberately, so each Trust should install a CCTV camera focused on the storage cupboard or unit. The camera should store recordings for at least 28 days. A visible camera is a very powerful deterrent and I see no argument against it. It is no more invasive than the digital tracker, the requirement to provide biometric data, or the cameras monitoring the entrances to hospitals or Accident and Emergency departments or other wards.
The provision of biometric data by someone accessing an insulin storage unit will combine restriction and record of access, obviating the need for CCTV. If that were in place, then such a system would comply – in spirit and practice – with my recommendation.

High insulin low C-peptide results: Laboratory handling guidance

Sir Stephen Powis was asked, directly, if there was a need for “regulation or some sort of direction to be given to laboratories by NHS England about how they should react in circumstances where a low C-peptide is found corresponding with a high insulin level … and a corresponding direction to doctors”. He took the point and undertook to discuss it with the Chief Pharmaceutical Officer.¹¹ This is now addressed in the laboratory handling guidance published in October 2025, which I welcome.¹² The document contains “best practices for establishing standards around the measurement, reporting, and timely communication of insulin and C-peptide test results”. This includes “[t]he critical identification and communication of red flag scenarios where hypoglycaemia occurs with elevated insulin but low C-peptide levels”. It is aimed at Trusts, biochemists/pathologists/scientists, endocrinologists, paediatricians, neonatologists, clinical and medical directors and pharmacists. The key recommendation is “[m]echanisms must be put in place to ensure that the red flag scenario of hypoglycaemia, elevated insulin and low C-peptide can be rapidly and readily identified”.¹³ To that I would add ‘and immediately acted upon’.
This 12-page document sets out the red flag scenario on three occasions. There is no doubt that the significance of high insulin low C-peptide in the presence of hypoglycaemia should be well understood. When it is found, something must be done and each person in the chain from patient to laboratory and back has responsibility for doing something. The guidance is detailed and ends by directing that local protocols be set up. I suggest that serious consideration be given to producing a single step-by-step protocol for use as a template when high insulin low C-peptide is found. The protocol can be adjusted to local conditions, if necessary, but the fundamental requirements will always be the same. That would be a much less labour-intensive way of producing guidance for use in all Trusts.
I recommend that all guidance, local and national, applying to the high insulin low C-peptide situation mandates immediate discussion between laboratory and treating clinician (which must be recorded) and between laboratory head and the medical director of the hospital (which must be recorded). Senior managers must inform NHS England and CQC. In the absence of a complete explanation for the finding, a senior manager must contact the police. Where a member of staff is suspected of being responsible, they should be removed from all patient contact pending investigation.
On behalf of the Countess, Mr Andrew Kennedy KC supported the evidence of Dr Brearey, who suggested that a nationally agreed process for the testing and reporting of insulin and C-peptide results should be developed. Mr Kennedy KC suggested that asking whether a patient had received insulin at the time of the insulin/C-peptide test would be enough to establish the significance of any result.¹⁴ This is an important point. I am encouraged to see it taken up in the guidance issued in October 2025.¹⁵ I recommend the current guidance be made mandatory and of national application.