The Lucy Letby Case May In Time Be The Story Of The Century

Is this a huge miscarriage of justice?

By Peter Elston

AS the retrial of Lucy Letby in relation to one of the ‘no verdict’ counts gets underway in Manchester, is it possible the case will soon be revealed as one of the worst miscarriages of justice in British legal history?

I’m an amateur not professional statistician, though I am a fellow of the Royal Statistical Society. I have a maths background and, during my 30-year career in the financial industry, applied a statistical approach to investing. I am also an accredited geologist (this involved an optional geostats module and anyway, like all sciences, geology requires a grasp of statistics) and I’m a fellow of The Geological Society. I also, like most people, have an abhorrence of injustice. Those are the relevant credentials that explained my early interest in the Lucy Letby case.

A few years ago I started to invoke in my investment articles the similarities between the use of statistics in the justice system (and miscarriages of justice in particular) and the world of investing such that, when Lucy Letby’s trial started in October 2022, I was drawn to look at it.

Lucy Letby had been charged with murdering seven babies and attempting to murder a further 10 on 15 occasions between June 2015 and June 2016 on the neonatal unit at The Countess of Chester, where deaths had previously averaged around three per year. This is a rate that was slightly below that nationally when adjusted for the number of births (there were also three non-fatal attacks on one of the alleged murder victims that were not attempted murder charges, taking the total number of attacks to 25).

Letby was alleged to have used a range of methods, such as injection with air (into bloodstream, stomach), smothering, some form of sabotage, feeding an excessive amount of milk (many will understand if you do this a baby just vomits), thrusting of a tube into back of throat to inflict injury, inflicting injury on liver and poisoned with insulin.

If convicted, she would have been an unusual serial killer; most serial killers stick to one method of attack. And her batting average of 0.28 (only seven of 25 attacks were successful) would have been particularly poor; who’d have thought vulnerable pre-term babies were so hard to kill?

Moreover, in the case of the two babies allegedly poisoned with insulin, they both made full recoveries from what were minor hypoglycaemic episodes, despite the prosecution arguing they received enough of the substance to down an adult (and it was strange Letby injected the insulin into bags of dextrose which is its antidote so is neutralised by it).

Anyway, in early October 2022, I dug out a Royal College of Paediatrics and Child Health (RCPCH) service review commissioned by The Countess in July 2016. I also came across hospital mortality data that had been obtained through a Freedom of Information request. And I went ‘Oh my God! The RCPCH review listed what seemed like countless hospital failings such as doctors’ reluctance to escalate or transfer, issues with umbilical venous catheter (UVC) insertion, understaffing, non-compliant number of ward rounds by paediatricians, etc. The review also cited low birth weights, low gestational ages and high acuity, though a doctor, one who had been saying Letby was a killer, said these did not explain the increased mortality (I suppose he would say that).

As for the mortality statistics, they showed there were 10 deaths during the period with which Letby was not charged. Another serial killer perhaps? What are the chances?

It is also just conceivable that the reason deaths fell after June 2016 was not that Letby had been removed from the ward as argued in the media but that the hospital had decided at the same time to no longer admit particularly vulnerable babies (they still don’t).

And I don’t know who last year told the BBC’s Judith Moritz that, “Since Letby left the hospital’s neonatal unit, there has been only one death in seven years”. According to the hospital mortality data there were seven and that was just in the two years between July 2016 and July 2018 (the data ended in October 2018).

I published an article on my blog on October 12, 2022 titled Miscarriages of Justice in Killer Caregiver Cases. Near the beginning of the article, I wrote:

This week the trial of nurse Lucy Letby got underway in Manchester. Letby is accused of murdering seven babies in her care and of attempting to murder 10 others. I am not suggesting she is innocent. It is possible that there is incontrovertible evidence against her as was the case with Harold Shipman. However, what I do know is if there isn’t, this is yet another case in which there is scope for bad statistics and a biased investigation to be presented to the jury. And therefore for another miscarriage of justice.

I ended the article with:

Caregivers who have either been wrongly accused or convicted, or where there is for good scientific/statistical reason for suspicion of such, include Lucia de Berk, Daniela Poggiali, Jane Bolding, Sally Clark, Susan Nelles, Ben Geen, and Collin Norris. If it turns out there is no direct evidence against Lucy Letby, let’s hope her name doesn’t join the list.

Ten months and zero direct evidence later, on August 18, 2023, Lucy Letby was found guilty on seven counts of murder and seven counts of attempted murder and sentenced to 14 full life terms (she was found not guilty on two counts and in the other six cases the jury was unable to return verdicts). She lost her leave to appeal application last month in London and tomorrow her retrial on one of the six ‘no verdicts’ will begin in Manchester (was 14 full life terms not enough?)

The statistics pertaining to the case, let alone the medical evidence that invoked highly unlikely methods of attack contradicting six postmortems that cited ‘natural causes’ and other evidence such as the Post-It notes, written by Letby when she must have been under extreme duress, were questionable. It all led me to call the case a ‘three-act tragedy’.

The scene of the first act was The Countess itself, where very quickly consultants associated incidents with Lucy Letby after the deaths of Babies A, C and D in June 2015. It is not clear why they associated the deaths with Lucy Letby rather than a whole host of risk factors, as follows.

In the case of Baby A, the mother was just 31 weeks pregnant and with twins; she had a rare condition, antiphospholipid syndrome, that causes blood clotting; her blood pressure had shot up requiring general anaesthetic (dangerous for babies) and an emergency C-section that increases infection risk; there were delays in transfer from the maternity to the neonatal unit; Baby A did not receive fluids, according to a nursing note, “for a couple of hours” because the UVC had twice been placed in the wrong position; there was no record of glucose feed for 23 hours, a catastrophic event for a preterm neonate with no glycogen reserves; another note read “doctors busy”; a long line placed by a junior doctor too close to the heart (obviously dangerous) which was removed (also dangerous) by the same doctor; and the pathologist noted “crossed pulmonary arteries” and a “strong temporal relationship” between the long line insertion and the collapse. It is no wonder the pathologist cited cause of death as “unascertained”, meaning a natural cause but not clear which.

Baby C weighed just one pound twelve ounces and was born to a mother with a dangerous placenta condition. Within four days of being born, Baby C had developed acute pneumonia. The pathologist cited ‘natural causes’.

Baby D’s mother went to The Countess after her waters broke (risk factor for infection) but was sent home, without antibiotics that would have reduced infection risk and told to wait. More than 24 hours later she noticed her baby was making fewer movements so returned to the hospital and still wasn’t given antibiotics, later saying she “felt forgotten by the staff”. Sixty hours after her waters had broken, she had a C-section, giving birth to a baby girl who was dusky and limp and who was not immediately given antibiotics (doctors later conceded she should have been). The pathologist found pneumonia in Baby D’s lungs and wrote that the infection was probably present at birth.

Many of the other babies who were allegedly attacked by Lucy Letby over the next 12 months also had issues of one sort or another. It is not clear why jurors were repeatedly told by prosecution witnesses that the babies were “stable” when that was clearly not the case.

The scene of Act Two was the police investigation, thoughtlessly (or perhaps not) named Operation Hummingbird (this was the name of Hitler’s purge of his enemies within his ranks). Very shortly after Cheshire Constabulary announced in May 2017 that it was investigating deaths and non-fatal collapses at The Countess, a long-retired paediatrician, Dewi Evans, sent an email to his friend ‘Nick’ at the National Crime Agency saying: “Incidentally, I’ve read about the high rate of babies in Chester and that the police are investigating…If the Chester police had no one in mind I’d be interested to help. Sounds like my kind of case.” (the defence tried to argue, not unreasonably, that this constituted touting for business).

Two months later Evans is given clinical notes by Cheshire Police relating to incidents it was investigating (there does not appear to have been the usual process of tendering through an expert witness database). And five months after that he declares that crimes have been committed, in the knowledge no doubt that he would later be paid by the CPS to testify at trial, a conflict of mammoth proportions (the more crimes, the longer the trial, the more money).

Interestingly, before his email came to light, Evans had testified at trial that he had been approached by police rather than the other way round. And jurors also heard that in a previous and unrelated trial the judge had called his evidence “worthless”. Moreover, when the prosecution presented Dr Evans’ credentials, it noted he was “properly registered” with his regulatory body, the General Medical Council; perhaps it was fearful of saying “properly licensed”, something altogether different, given he had not been for the first two years of the police investigation (as of eight days ago, he is again not licensed to practise, perhaps because of the numerous complaints that have been submitted against him in recent weeks).

While Evans had his head in clinical notes, the police searched Lucy Letby’s homes and unearthed the sinister code in her diary, “LO”, which it said corresponded with dates on which she had attacked babies (the sinister code later turned out to be “LD”, an acronym commonly used by nurses short for “long day”). Photos found on Lucy Letby’s phone of bereavement cards she had sent to parents were deemed to be the trophies of a serial killer rather than records a kindly nurse might want to keep to remind her of her patients.

One or two sentences on Post It notes were considered confessions, while the many others that were not incriminating were ignored. And anyway, the handwritten notes, according to renowned criminologist Prof. David Wilson, appeared more like the outpouring of a young person under extreme stress who had just lost her dream job and been accused of the most horrendous of crimes.

Then comes Act Three and the trial at Manchester Crown Court and six prosecution medical experts who between them ignored six ‘natural causes’ original postmortem reports, instead invoking various unlikely methods of attack (there were countless other prosecution witnesses, while the defence only called Lucy Letby herself and, to corroborate her testimony about the frequent sewage issues on the unit, a plumber).

The only surprise perhaps is that the jury did not return guilty verdicts on all 22 counts, though it seems understandable it did not believe the main witness in the case of Baby K, Dr Ravi Jayaram, who said, among other things, that the baby could not have removed its nasogastric tube because she had been sedated (she hadn’t been, as was recorded in clinical notes).

The closest the prosecution came to presenting forensic trace evidence was in the form of two blood test results which two biochemists testified could only have been the result of exogenous insulin having been administered. The only problem with these assertions was that the test guidelines clearly stated, “Please note that the insulin assay performed at Royal Liverpool University Hospital is not suitable for the investigation of factitious hypoglycaemia. If exogenous insulin administration is suspected as the cause of hypoglycaemia, please inform the laboratory so that the sample can be referred externally for analysis”.

The reason for this is that the test, a quick and dirty one, was not necessarily measuring insulin. There are cross reactivities, made clear by manufacturer Roche to users of its product, that meant the test could pick up not just insulin but also insulin analogues such as insulin antibodies and proinsulin, hence the need in the event of an unusual reading for more detailed and expensive mass spectrometry analysis that would separate components.

This analysis was never done, perhaps because the two babies had by the time the results were sent back to the hospital made full recoveries (itself instructive). To top it all, the high ratio of insulin to another hormone, C-peptide, that prosecution witnesses said was unusual and thus proved poisoning turns out to be usual in pre-term neonates.

According to the article by Rachel Aviv about Letby case published recently in The New Yorker, Evans had flagged to the police 25 more incidents that he thought were suspicious. Among them was a third insulin case that Evans said bore the same high insulin to c-peptide hallmark and thus was a clear case of poisoning. Michael Hall, a retired neonatologist at University Hospital Southampton who worked as an expert for Letby’s defence, was surprised and disturbed to learn of it from Aviv and wondered why it had not been part of the trial. Perhaps there was an alternative explanation for the test result (which would have been relevant for the other two cases). Or perhaps Letby wasn’t on duty at the time. “Whichever way you look at this, that third case is of interest,” Hall told Aviv.

Regarding duty patterns, Jurors were shown by prosecutors on several occasions during the trial a roster table purporting to show that Lucy Letby had been on duty for all 25 incidents, a construct that would have been compelling evidence to them of her guilt. Except it did not include the 35 other cases that Evans had reviewed. Nor did it take account of the fact that many of the other 37 nurses worked part time. Nor the fact that Lucy had an intensive care qualification. Nor that she was saving to buy a house and an enthusiastic nurse, so accepted requests to do overtime. These were all factors that made it more likely she would be involved in some way with babies on the unit than other nurses.

Furthermore, in relation to Letby only, the prosecution used the phrase “there or thereabouts” rather than “on duty”. Letby worked 60-hour weeks, so was on duty 35 per cent of the time, closer to 40 per cent if adjusted for the “there or thereabouts”. She was charged with 40 per cent of the deaths during the 13 months. And 40 per cent of the non-fatal collapses. In other words, what you’d expect if the incidents were non-malevolent and spread randomly across shifts.

Every miscarriage of justice involves key elements that come together to charge or, worse, convict an innocent person, whether a mother (e.g. Sally Clark, Kathleen Folbigg), a nurse (e.g. Lucia de Berk, Daniella Poggiali), a doctor (e.g. David Sellu), subpostmasters (900 of them), or the many others (The University of Exeter miscarriages registry lists 466 cases of proven miscarriageand those are just the ones that involved a conviction).

I attended Lucy Letby’s appeal court hearing recently, including the announcement by the judges that her leave to appeal had been refused. On the way to the bus stop afterwards I passed the building that was hosting the Post Office and Infected Blood inquiries and wondered about the cognitive dissonance that allowed most to be appalled by these but at the same time to dismiss the mere possibility that Lucy Letby could be innocent.

Many of course would have been influenced by the numerous “evil witch” articles published after the trial (why these articles, all still available online, were not considered prejudicial to the retrial about to get underway in Manchester is beyond me).

Then last week I watched the BBC2 documentary about Andrew Malkinson and was reminded of the appalling revelation that the CPS was aware in 2009 of a DNA profile from the victim’s clothing that did not match the man who had already been in prison for six years. And who would remain in prison for another 11 years, despite the new and overwhelming evidence. (And whose case the Criminal Cases Review Commission, a supposedly independent body led by a former chief executive of United Biscuits, refused to refer to the Court of Appeal. Not once, but twice.)

Is it just me, or is there a serious problem?

• This article is supported by Rachel Aviv’s recent long investigative piece about the Letby case published in The New Yorker. Aviv’s article drew on numerous opinions of experts that raised questions about the case. And it underwent a meticulous fact checking process before it was published.