From the New Yorker:
Nearly a year after Operation Hummingbird began, a new method of harm was added to the list. In the last paragraph of a baby’s discharge letter, Brearey, who had been helping the police by reviewing clinical records, noticed a mention of an abnormally high level of insulin. When insulin is produced naturally by the body, the level of C-peptide, a substance secreted by the pancreas, should also be high, but in this baby the C-peptide was undetectable, which suggested that insulin may have been administered to the child.
The insulin test had been done at a Royal Liverpool University Hospital lab, and a biochemist there had called the Countess to recommend that the sample be verified by a more specialized lab. Guidelines on the Web site for the Royal Liverpool lab explicitly warn that its insulin test is “not suitable for the investigation” of whether synthetic insulin has been administered. Alan Wayne Jones, a forensic toxicologist at Linköping University, in Sweden, who has written about the use of insulin as a means of murder, told me that the test used at the Royal Liverpool lab is “not sufficient for use as evidence in a criminal prosecution.” He said, “Insulin is not an easy substance to analyze, and you would need to analyze this at a forensic laboratory, where the routines are much more stringent regarding chain of custody, using modern forensic technology.” But the Countess never ordered a second test, because the child had already recovered.
Brearey also discovered that, eight months later, a biochemist at the lab had flagged a high level of insulin in the blood sample of another infant. The child had been discharged, and this blood sample was never retested, either. According to Joseph Wolfsdorf, a professor at Harvard Medical School who specializes in pediatric hypoglycemia, the baby’s C-peptide level suggested the possibility of a testing irregularity, because, if insulin had been administered, the child’s C-peptide level should have been extremely low or undetectable, but it wasn’t.
The police consulted with an endocrinologist, who said that the babies theoretically could have received insulin through their I.V. bags. Evans said that, with the insulin cases, “at last one could find some kind of smoking gun.” But there was a problem: the blood sample for the first baby had been taken ten hours after Letby had left the hospital; any insulin delivered by her would no longer be detectable, especially since the tube for the first I.V. bag had fallen out of place, which meant that the baby had to be given a new one. To connect Letby to the insulin, one would have to believe that she had managed to inject insulin into a bag that a different nurse had randomly chosen from the unit’s refrigerator. If Letby had been successful at causing immediate death by air embolism, it seems odd that she would try this much less effective method.