Late on 17 February, a document briefly appeared on the website of Lady Justice Thirwall’s public inquiry into the Lucy Letby murders. By morning it had vanished.
UnHerd can now reveal that this was a statement to police by Dr Astha Soni, a paediatrician at the neonatal unit where Letby worked. And it raises serious doubts about the claims that Letby attempted to murder two babies by poisoning them with insulin. Yet the document was not disclosed to the defence before her first trial began, nor during her unsuccessful bid to appeal. Furthermore, we have discovered a second critical omission, involving the outbreak of a dangerous virus in Letby’s neo-natal ward.
As Thirlwall’s inquiry proceeds, troubling concerns over Letby’s seven convictions for murder and eight for attempted murder — she is now serving 15 whole life sentences — have only swelled. The UK-based consultant neonatologists Svilena Dimitrova and Neil Aiton have submitted detailed reports on three of the babies’ deaths that refute the evidence given at Letby’s trial, and a separate panel of international experts led by Canadian neonatologist Professor Shoo Lee concluded last month there was “no medical evidence to support malfeasance” in the deaths Letby was convicted of causing. Their assessment: the babies died from “either natural causes or bad medical care”. It all points to a troubling question: is Lucy Letby truly guilty beyond reasonable doubt?
Dr Soni’s secret, undisclosed police statement concerns “Baby Y” — born just after midnight on 2 November 2015, some months into the rise in neonatal mortality at the Chester unit that the police were investigating. Letby was never formally accused of trying to harm him, but the prosecution experts who gave evidence at her trial believed she poisoned him with insulin, just as she did Baby F and Baby L, whom she was convicted of trying to murder. Since Soni’s statement undermines that analysis, it raises worrying questions about the experts’ evidence in those other cases.
Baby Y was a big baby, born two weeks after his due date following a long and difficult labour. He was finally delivered by an emergency caesarean, and transferred to the hospital’s neonatal unit early on 3 November. This happened, Soni said in her statement, after the baby experienced “dusky episodes”, struggling to breathe and turning blue. In the second of these episodes, Soni told police, the baby was not only blue but unconscious, with “his eyes rolled up” and his torso “floppy”. Once on the unit, he was given oxygen, but “didn’t pick up straightaway”. Half an hour later, Baby Y suffered what Soni called a “seizure” — turning blue and making “no respiratory effort”.
When Dr Soni saw the baby again later that night, she said, he was hypoglycaemic. Despite increasing doses of dextrose, his blood sugar level remained stubbornly low. His diagnosis, Soni said, was “congenital hyperinsulinism” or CHI, a genetic condition in which babies produce excess insulin. This diagnosis was supported by endocrine experts at Alder Hey hospital in Liverpool whom the Chester doctors consulted, the statement went on. It is also possible that he was suffering from “transient hyperinsulinism”, a non-genetic, common condition with natural causes whose symptoms would have been identical. After more seizures Baby Y was given several drugs and slowly improved, but outpatient treatment for “persistent hypoglycaemia associated with seizures” had to be continued for over seven months.
Medical experts for the Letby prosecution saw it differently. In a statement to police in 2021, seen by UnHerd, Professor Peter Hindmarsh of University College London, who was called as an expert witness by the prosecution over the cases of Baby F and Baby L, commented on a test on a blood sample taken from Baby Y on 3 November. It showed that while the boy’s insulin level was high, the concentration of C-peptide, a substance produced in tandem with insulin, in his blood was low. According to Hindmarsh, this meant Baby Y’s hypoglycaemia must have been caused by “exogenous” insulin administered by someone in the neonatal unit.
Other prosecution experts agreed. Dewi Evans, the retired paediatrician who had played a central role in the police inquiry and also gave evidence in respect of most of the babies Letby was convicted of harming, also reviewed Baby Y’s medical records. He told police in a statement that he thought his first three seizures had been caused not by CHI, but the “trauma” of his difficult birth, agreeing with Hindmarsh that he had later been given “insulin from an external source”. A third prosecution expert, Dr Sandie Bohin, backed this position too.
However, we spoke to two practising neonatal consultants who say all Baby Y’s “dusky episodes” and seizures could have been caused by naturally occurring hyperinsulinism, either transient or CHI — which, they say, is far more common than the prosecution experts appeared to believe. They say what matters most is that Baby Y had a confirmed diagnosis of CHI and still needed treatment long after leaving the Chester unit.
Shoo Lee’s expert panel has come to a similar conclusion. Its summary report, now being examined by the Criminal Cases Review Commission (CCRC), says naturally-occurring hypoglycaemia affects up to 40% of newborn babies, and that the relative levels of insulin and C-peptide seen in babies F, L and Y are also “not uncommon”. The timeline here is important too, because Baby Y’s first three seizures all took place before Letby ever encountered him.
“Baby Y’s first three seizures all took place before Letby ever encountered him.”
If Soni’s statement had been given to Letby’s defence team, they might have called her as a witness — and explored why she thought Baby Y was suffering from CHI rather than a deliberate insulin overdose. That, in turn, might have enabled them to question the conclusions reached by the prosecution experts around babies F and L. In those cases, too, they claimed that the combination of high insulin and low C-peptide levels pointed to deliberate insulin poisoning.
We asked the Crown Prosecution Service why it failed to disclose Soni’s statement, but it refused to answer, stating only that Letby had been convicted and had lost her appeal. But its former chief, Lord Ken Macdonald KC, suggests the statement should likely have been shared. “The legal test is whether evidence is capable either of undermining the prosecution case or assisting the defence, and if it is, it should be disclosed before trial,” Macdonald says. “It sounds to me that the material you’ve found does meet that standard, and so should have been made available to Letby’s defence.”
Beyond the questions around Soni’s deleted statement, our investigation has uncovered a second key omission. Neither of the juries in Letby’s two trials was informed that a dangerous respiratory syncytial virus (RSV) outbreak was ravaging the neo-natal ward. This breakout was so severe that it forced the unit’s closure to new admissions — on 18 February 2016, the day after Baby K, one of the babies Letby was convicted of trying to murder, was moved from Chester to another hospital, where she later died.
This is despite the fact that staff at the hospital were clearly concerned about the outbreak. UnHerd has seen a police statement by Dr Bohin, dated 2021, which refers to a “management synopsis report” on the virus, a spreadsheet listing all the babies who contracted it, along with references to several similar documents about the outbreak.
Until the eve of her first trial, Letby was charged with murdering Baby K, but the prosecution asked the judge to alter the indictment to attempted murder. It was confident Letby had dislodged the child’s breathing tube — but, to quote the prosecution counsel, “we cannot and do not say” that what she did caused the girl’s death. He did not mention the virus sweeping through the neonatal unit when Baby K was born. For her part, Bohin made the dangers of RSV clear. “Its effects are usually mild, but in preterm infants RSV is responsible for significant mortality and morbidity,” she told police in a statement. “The virus is spread easily and can live for prolonged periods on hands and hard surfaces. There is no cure.” Baby K’s medical records do not say whether she was tested for RSV, but she had been in respiratory distress, and if she did contract the virus at Chester, it may have contributed to her death.
Meanwhile Shoo Lee’s expert panel is sceptical of the prosecution’s argument that Letby tried to kill Baby K by dislodging her breathing tube, suggesting instead that the hospital had used the wrong size tube, causing air to leak.
Beyond the medical questions here, these documents could have implications for Letby’s future. The CCRC and the Court of Appeal can consider fresh evidence that might, in theory, have been available at the time of a trial — if they are satisfied that to do so would be in the interests of justice. But they also have to consider whether there is a reasonable explanation why it was not used by the defence, and there is no stronger reason than a failure by the prosecution to disclose evidence to start with. This is ultimately why the cases of Baby Y and Baby K may yet be so significant.
Later this month, Lady Justice Thirwall will be asked by Mark McDonald, Lucy Letby’s lawyer, to put her inquiry on hold until the Criminal Cases Review Commission (CCRC) decides whether to order a new appeal. If it does, he will say, it would undermine the whole inquiry, which has been based from the outset on the presumption that Letby is guilty. In the face of the growing volume of fresh evidence, the verdicts against her are beginning to look distinctly unsafe.
