Why the peer-reviewed layer matters
The Crown presented the Baby F Roche Cobas result as forensic proof: “you cannot argue with a lab result”. In the period since the August 2023 verdicts, specialist peer-reviewed clinical-biochemistry and endocrinology journals have addressed the forensic-use question directly. The professional-literature verdict — that the Crown’s framing overstated what the assay can establish — is the operational form of specialty consensus against the Crown’s insulin-evidence reading.
The specific points the literature establishes
- Immunoassays are screening tests, not forensic tests. The Roche Cobas system, and all insulin immunoassays, are designed to flag samples for further investigation. Forensic distinction of exogenous from endogenous insulin requires mass spectrometry.
- False-positive rates are non-trivial. Insulin auto-antibodies, heterophilic antibodies, macro-insulin, assay interference, cross-reactivity with certain drugs — each produces false-positive results. The published false-positive literature in the Roche Cobas specifically is cited in the Joint Insulin Report.
- Sample-handling is load-bearing. Gel tubes, delayed centrifugation, ambient-temperature storage each independently compromise a sample’s forensic reliability. Together, as in the Letby samples, they disqualify forensic use.
- C-peptide dissociation is not specific to exogenous insulin. Adrenal suppression, sepsis, liver disease, kidney disease, certain drug exposures, repeated dextrose infusions — all suppress C-peptide in ways that can produce the pattern the Crown interpreted as exogenous-insulin-specific.
- Physiological plausibility matters. Prof. Geoff Chase’s physiological modelling, published in peer-reviewed form, demonstrates the reported Letby insulin values are implausible on the Crown’s own theory of a TPN-bag contamination.
The Royal Liverpool laboratory protocol change
A specific documentary feature surfaced by science4justice.nl (Dr Sarrita Adams) and cited in peer-reviewed literature: the Royal Liverpool laboratory’s own insulin assay protocol changed between 2010 (in force at the time of Letby sample testing) and 2012 (post-protocol). The 2012 protocol explicitly states the laboratory cannot diagnose exogenous insulin. The 2010 protocol did not include that statement. The laboratory itself, in its own post-2012 guidance, acknowledges that results of the type produced in the Letby case cannot diagnose exogenous insulin.
What the literature is not saying
The peer-reviewed clinical-biochemistry literature is not publishing post-2023 work defending the Crown’s insulin-evidence framing. There is no body of peer-reviewed work saying that the Roche Cobas screening result, processed under the Letby chain of custody, can support a criminal finding of exogenous insulin administration.
This evidential absence is itself significant. If the Crown’s framing were defensible on modern laboratory-medicine standards, peer-reviewed literature would be defending it. It is not.
What the Joint Insulin Report consolidated
The May 2025 Joint Expert Witness Insulin Report on Babies F and L, filed with the October 2025 CCRC application, consolidates the peer-reviewed literature into a case-specific report. Its conclusion — the insulin evidence cannot support a criminal finding of exogenous insulin administration — is the operational clinical-biochemistry consensus. See our summary.
Why this matters for the CCRC review
The insulin count on Baby F was the only count with a concrete laboratory measurement. The peer-reviewed literature since 2023 has established, on modern specialty consensus, that the measurement cannot bear the weight the Crown placed on it. On the Cannings principle — reputable medical experts disagree on the interpretation — the count is unsafe. See our Baby F insulin deep-dive.
The Roche Cobas immunoassay in clinical biochemistry
The Roche Cobas insulin-immunoassay platform is widely used in UK clinical biochemistry laboratories for routine clinical (non-forensic) insulin measurement. Its validation profile is documented in extensive peer-reviewed literature, including the canonical clinical-biochemistry texts. The manufacturer’s own documentation specifies the conditions under which the assay produces reliable quantitative results and the conditions under which it does not.
For forensic use — specifically, for distinguishing endogenous from exogenous insulin in a criminal-evidence context — the manufacturer’s documentation specifies confirmatory mass-spectrometry testing as the required protocol. The screening immunoassay alone cannot reliably make this distinction because of multiple known interference mechanisms: the hook effect at high C-peptide concentrations, cross-reactivity with insulin auto-antibodies, sample-handling degradation, and dextrose-treatment-induced C-peptide suppression.
The peer-reviewed literature on assay limitations
The peer-reviewed literature on Roche Cobas insulin-immunoassay limitations is extensive. Dr Adel Ismail’s clinical-biochemistry commentary consolidates the false-positive catalogue and the interference mechanisms from the published literature. Prof. Geoff Chase’s physiological-plausibility modelling addresses the specific Baby F reading (4,657 pmol/L) and demonstrates that this magnitude is incompatible with the Crown’s proposed mechanism (small-volume spike of exogenous insulin into a slow-running TPN bag).
The Joint Expert Witness Insulin Report (May 2025)
The Joint Expert Witness Insulin Report on Babies F and L, published May 2025, consolidates the peer-reviewed literature and applies it to the trial record. The Report’s consensus position is that the Crown’s insulin evidence cannot bear the evidential weight a criminal conviction requires. The Report addresses: the Roche Cobas methodology; the absence of confirmatory mass-spectrometry; the false-positive catalogue; the sample-handling standards; the C-peptide-dissociation non-specificity; the physiological-plausibility modelling; and the Royal Liverpool 2012 protocol acknowledgement.
The Guildford forensic-laboratory standard
The UK forensic-laboratory standard for insulin measurement in criminal proceedings is the Royal Surrey County Hospital laboratory at Guildford, which operates under the Forensic Science Regulator’s code of practice for endocrinology work in criminal contexts. Royal Liverpool, where the Letby insulin samples were tested, is a clinical biochemistry laboratory operating outside the Forensic Science Regulator framework. The forensic-laboratory-standard question is one of the load-bearing chain-of-custody arguments in the post-conviction insulin critique.