Why this count matters more than the others
Most counts on the indictment rest on expert interpretation of clinical signs. The Baby F insulin count is different: it has numerical laboratory measurements. If those measurements are a sound forensic test, the count is evidentially stronger than any other. If they are not, the count collapses. The Panel’s answer is that they are not — and, separately, that two of the factual premises the Crown built around them do not match the clinical record.
Step 1: The clinical context, on the Panel’s summary of the record
Baby F was a twin, born at 29 weeks and 5 days weighing 1.434 kg, delivered by emergency Caesarean section for absent end-diastolic flow, with borderline intrauterine growth restriction. He had mild respiratory distress syndrome, and hyperglycaemia that required insulin treatment.
On 5 August 2015, at 0130, he developed sepsis and hypoglycaemia, and was treated with antibiotics and an intravenous glucose infusion. Over the next 17 hours his blood glucose stayed low — in the range 0.8 to 2.4 — despite repeated boluses of 10% dextrose.
Step 2: The long line tissued, and the glucose stopped
At 1000 his long intravenous line was noticed to have tissued, with extensive swelling and induration of the right groin, thigh and leg. Intravenous fluids were stopped from 1000 to 1200 while a new long line was inserted. At 1200 the infusion bag was changed. At 1900 the dextrose infusion was increased to 15%, and the hypoglycaemia resolved.
That sequence — a septic, growth-restricted preterm infant left without his glucose infusion for several hours — is the Panel’s explanation for why the hypoglycaemia was so prolonged.
Step 3: What the Crown alleged
It was alleged that Baby F was given exogenous insulin through his infusion bag. The allegation rested on four propositions: that the hypoglycaemia was inexplicably prolonged; that his blood glucose rose from 1.3 to 2.4 while the dextrose infusion was stopped between 1000 and 1200; that his blood sugar rose after the infusion bag was changed; and that he had high insulin with low C-peptide, which indicates exogenous insulin.
Step 4: The Panel finds two of those premises are wrong on the record
The first is the “rise without dextrose” point, which was the Crown’s most striking claim: sugar going up while the glucose was switched off. The Panel says the blood sugar did not rise from 1.3 to 2.4 while the infusion was stopped, because the blood sugar was 1.4 at 1146 — still within the stopped period. The 2.4 reading was measured after 1200, when the intravenous line had been restarted.
The second is the “rise after the bag change” point. The Panel says the blood sugar rose after 1900 not because the infusion bag was changed, but because the dextrose was increased to 15%.
Step 5: The bags would all have had to be poisoned
The infusion bags were prepared in the pharmacy, stored on the unit, and changed at 1200. The Panel’s point is structural: if insulin poisoning of the bags were the explanation for a hypoglycaemia that ran across that bag change, then multiple infusion bags would have had to be contaminated, not one.
Step 6: The hypoglycaemia was managed badly, and that made it worse
The Panel is critical of the treatment. When the hypoglycaemia persisted despite a 10% dextrose infusion, a higher glucose infusion should have been given earlier. Worse, repeated boluses of 10% dextrose worsen hypoglycaemia: they cause surges of blood sugar, which trigger surges of insulin secretion, producing a yo-yo pattern of sharp rises and falls in insulin and blood sugar. Baby F received repeated 10% dextrose boluses across those 17 hours.
Step 7: Preterm infants are not small adults — the normative-standards problem
The Panel relies on the bioengineering analysis by Chase and Shannon, annexed to its report, which reports that preterm infants have different insulin and C-peptide normative standards from adults. The reference range against which Baby F’s results were read as abnormal is not the range that applies to him.
Step 8: Why the Panel says exogenous insulin is unlikely
The Panel gives six reasons, taken together, for concluding that exogenous insulin is unlikely to be the cause of the hypoglycaemia:
- The C-peptide was not low for a preterm infant — it sits at the 20th to 45th percentile.
- Potassium levels were normal. Insulin decreases potassium.
- Glucose levels should have been lower still if exogenous insulin had been used.
- The insulin/C-peptide ratio was within the expected range for preterm infants.
- Insulin autoimmune antibodies, which are common in preterm infants, bind to insulin and increase measured insulin levels.
- The immunoassay test is unreliable, because interference factors such as sepsis and antibiotics can give false-positive insulin readings — and Baby F was septic and on antibiotics.
The Panel’s conclusions on Baby F
- Baby 6 had prolonged hypoglycaemia because of sepsis, prematurity, borderline intrauterine growth restriction, lack of intravenous glucose when the long line infiltrated for a prolonged period of several hours, and poor medical management of the hypoglycaemia.
- Baby 6’s insulin level and insulin/C-peptide ratio do not prove that exogenous insulin was used, and are within the norm for preterm infants. Preterm infants, and especially those with illness and drug treatments such as antibiotics, have different normative standards compared with healthy adults and older children.
Source: International Expert Panel — Summary Report (February 2025), “Baby 6”, and the Chase and Shannon analysis at its Annex. Read the Panel report (PDF). For how to read it, see how to read the Panel report.
What this means for the indictment
The insulin count was the count that was supposed to be the anchor: the one place where the Crown could point to a number rather than to an interpretation of clinical signs. On the Panel’s reading, the number does not do the work asked of it — the results are within the norm for a preterm infant, and are exactly what a septic preterm infant on antibiotics would be expected to produce on an immunoassay. What remains on the rest of the indictment — the air-embolism counts, the air-in-stomach counts — is interpretive in a way the insulin count was said not to be.
The Panel’s report has not been tested in court and no court has ruled on it. It forms part of the post-conviction material now before the CCRC. Lucy Letby’s convictions currently stand.
What the Panel’s report adds
The Panel reported in February 2025, after the 2023 convictions. None of the following was before the jury in this form:
- That the blood sugar did not rise while the dextrose was stopped — it was 1.4 at 1146, and the 2.4 came after the drip was restarted at 1200.
- That the sugar rose after 1900 because the dextrose was increased to 15%, not because the bag was changed.
- That the bags were pharmacy-prepared and stored on the unit, so multiple bags would have had to be contaminated.
- That the C-peptide was not low for a preterm infant, and that the potassium was normal.
- That insulin autoimmune antibodies are common in preterm infants and inflate measured insulin, and that sepsis and antibiotics can produce false-positive insulin readings on immunoassay.
- That repeated 10% dextrose boluses make hypoglycaemia worse, and that a higher glucose concentration should have been given earlier.