Lucy Letby: Serial killer or a miscarriage of justice?

Letby illustration
Letby illustration

When Lucy Letby was convicted of the murders of seven newborns and the attempted murders of six more in August 2023 she was described by the judge as having led a “cruel, calculating and cynical campaign of child murder”.

The detectives who led the case said Letby had operated in plain sight and had abused the trust placed in her in the most unthinkable way.

No one could comprehend what had led a nurse with a previously unblemished career to embark on a campaign of such unimaginable horror, attacking and killing the most vulnerable members of society.

Jump to:

A damning shift pattern

The problem of prematurity

Too much air

A strange drop in blood sugar

A struggling baby unit

Witness for prosecution

Murder or Coincidence?

Last week, Letby was convicted of another attempted murder – a charge on which the original jury had been unable to reach a verdict – confirming her status as Britain’s most prolific baby serial killer.

But despite the guilty verdicts and the fact her application to appeal was roundly rejected by judges in May, troubling questions have started to emerge about some of the scientific evidence used to convict Letby.

After the trial, numerous blogs and armchair detectives suggested she could be innocent in interventions which were largely dismissed as conspiracy theories.

The grieving families of the victims have also spoken of their anguish that anyone would question the verdict or show sympathy to Letby.

However, experts from universities including Edinburgh, Harvard and Bristol, and members of the Royal Statistical Society (RSS) are now questioning the way crucial evidence was presented in court.

One of the scientists whose paper was cited in the original case has suggested that his work was misinterpreted.

Others have gone as far as to suggest that, rather than being a calculating killer, Letby is a victim of NHS failings.

Here, The Telegraph examines the scientific concerns in the arguments used to convict Letby.

One of the most damning pieces of evidence presented to the jury was a chart revealing that Letby was always on duty at the Countess of Chester Hospital when babies collapsed or died.

The table, which covered a 13-month period between June 2015 and June 2016, showed that each of the 38 other nurses was in attendance on just a handful of occasions when suspicious incidents occurred.

In contrast, a seemingly conclusive stark black line of dots placed Letby at the scene of all the incidents. Opening the prosecution’s case Nick Johnson KC said that “by a process of simple elimination” only one person could be responsible.

Dr Alexander Coward, a former maths lecturer at the University of Oxford and University of California Berkeley, who reviewed the case, said: “At first glance it’s like ‘Oh my God, at the times all the babies died, Lucy Letby is right there’

“If that was all the data it would be pretty compelling. You’d say something fishy is going on.”

Yet experts say there are major problems in this approach. To begin with, it is unsurprising that Letby is present on all these occasions given that these are the cases for which she is under investigation.

The chart excludes deaths and collapses which occurred at the neonatal unit when she was either not present or where there was little evidence to suggest she was responsible.

Freedom of Information (FOI) data from the unit shows differing results, but there may have been up to 17 infant deaths during the same period, with Letby only charged with seven.

Even without the deaths for which she was tried, the mortality rate at the neonatal unit was much higher than it should have been during 2015 and 2016.

The statistician Peter Elston, who runs the statistics and investment blog Chimp Investor has estimated the probability of so many extra deaths occurring through normal variation to be about 1 in 290.

It suggests that, above and beyond Letby, there was something else going on in the neonatal unit that was increasing the death risk for babies, Mr Elston says.

David Wilson, emeritus professor of criminology at Birmingham City University, an expert in serial killers, said shift patterns were always a shaky form of evidence.

“The weakness of that sort of statistical analysis was really as plain as a pikestaff,” he said.

“What the defence never did was challenge the fact there were other incidents during that time period when Letby wasn’t on duty, and, in fact, there were [at least] nine other neonatal deaths on the ward during that period.

“So the prosecution present the table and it looks like the common denominator is Lucy Letby, but they don’t present all the collapses or deaths during that period because that doesn’t necessarily suit the argument that they are trying to make.”

He added: “Statistics and the law both speak English but they speak it in a very different dialect.”

There are other reasons to be cautious about the shift patterns. Dr Coward has shown it is possible to take an identical number of nurses and incident rate at a hospital with 170 shifts and place any nurse at the same number of events as Letby, simply through randomness.

“You could make a chart like that for any nurse in any hospital,” he said.

“All you would have to do is only focus on the times when something went wrong when that nurse was on duty and ignore all the times that nurse was on duty and everything was fine.

“You don’t need a PhD in statistics or maths to know that this is dreadful. This illustrates what you can do with cherry-picked data.”

A similar use of statistics led to the wrongful conviction of Lucia de Berk, a Dutch paediatric nurse convicted of seven murders and three attempted murders in 2004. De Berk was exonerated in 2010 and her case led to reforms of the Dutch legal system.

Partly prompted by this, in 2022, the RSS produced a paper entitled “Healthcare serial killer or coincidence?” to help legal teams present data correctly.

It recommended that investigators and prosecutors consult professional independent statisticians who could give instructions to the jury on how to interpret the data. This did not happen in the Letby trial.

It also advised that confounding factors that could distort the figures should be made clear. Again this did not happen.

Letby was described as a young keen nurse, who would often work over-time. In the spring of 2015, she gained a specialist intensive care qualification, meaning she was allowed to care for the sickest babies on the unit, increasing the likelihood she would be present for deaths or collapses.

Prof Peter Green, of the School of Mathematics at the University of Bristol, one of the authors of the RSS report, said he was concerned by the way the shift chart was compiled.

Speaking in a personal capacity, he said: “The spreadsheet duty roster is almost a textbook example which I would give to my students of how not to collect and present data.”

In 2022, the RSS report warned that evidence such as shift patterns cannot stand alone but must be backed up by other proof, such as motive or clear method of murder.

In the case of Harold Shipman, who gave his patients lethal amounts of drugs, police discovered that several of the victims’wills had been changed to leave assets to the GP.

Burkhard Schafer, professor of computational legal theory at the University of Edinburgh said: “In the Letby case, the prosecution did also present direct evidence, at least for some of the deaths – a theory of how they think she committed the murders.

“This would have been fine if that evidence had been really strong. In this case though, the direct evidence that I saw was much weaker

“My worry would be now that the inference the jury drew went in the wrong direction, and that the weak statistical evidence ‘shored up’ or compensated for any concerns with the other evidence.”

Throughout the trial, the prosecution insisted that the babies who had died or been harmed at the Countess of Chester were in relatively good health.

Infants were variously described as “in good condition”, “stable”, “all observations normal” “doing well” and even “excellent”.

Yet it is clear that many of the babies were born desperately early, had extremely low body weight and, from birth, were beset with numerous complications.

The majority were admitted to the intensive care unit, and most required help breathing with many suffering from medical accidents which had nothing to do with Letby.

Take Baby C, a little boy who lived just four days.

He had been born at 30 weeks weighing less than 2lb after problems with the blood flow to the placenta meant he was only half the size he should have been. One nurse described him as “the smallest baby I have ever seen”.

The little boy was also suffering from pneumonia and breathing distress and an X-ray taken the day before his death showed air in his stomach indicating that potentially he had a blockage in his bowel.

After he collapsed and died, a post-mortem examination at Alder Hey Children’s Hospital, in Liverpool, concluded the death was natural, exacerbated by the lack of blood flow in the womb. A coroner supported this finding.

Despite this, the prosecution claimed his death was caused by Letby pumping air into his stomach.

Letby has been convicted of another attempted murder – a charge which the original jury had been unable to reach a verdict
Letby has been convicted of another attempted murder – a charge which the original jury had been unable to reach a verdict - Chester Standard

Michael Hall, a consultant neonatologist at University Hospital Southampton NHS Foundation Trust, was instructed as an expert for the defence but was never called. He said he was troubled by what he considered to be the prosecution’s “exaggeration” of the health of the infants.

“It’s my opinion, the prosecution medical expert witnesses exaggerated the degree of wellness of those babies to a significant extent,” he said. “I would have thought it would have had a significant influence on the jury.

“One example of this is Baby A. His vital signs were displayed to the court on several occasions and it was clear that he was receiving respiratory support.

“But his breathing rate was clearly abnormal for almost 24 hours, or at least intermittently, certainly for the last 12 hours before he collapsed.”

The experts said Baby A could be considered “stable” because he did not require oxygen ventilation and was doing “so well” that medical staff decided to start giving him some feeds.

But the other babies who died were also vulnerable and initially, nobody considered their deaths to be suspicious, merely a natural outcome of prematurity, illness and even sub-optimal care.

Baby B, the twin of Baby A was born at just 31 weeks, blue and floppy and needing resuscitation, to a mother suffering from an auto-immune disease that can increase blood clots, and both babies initially needed help breathing.

Likewise, Baby D was delivered by C-section and within 12 minutes of birth had lost colour and become floppy in her father’s arms, later showing signs of respiratory distress and an infection.

She had been born early when her mother’s waters broke early putting her at risk of health problems with breathing, feeding and infection. Her mother should have been given antibiotics but was not, an error for which the hospital was criticised.

After death, the coroner ruled Baby D had died from pneumonia with acute lung injury.

Baby E was born at just 29 weeks and needed help breathing. He was also suspected of having necrotizing enterocolitis (NEC), a life-threatening disease with a mortality rate as high as 50 per cent.

After his death the consultant ruled Baby E had been a “high-risk infant” and certified the cause of death as NEC and prematurity.

Baby F, the twin of Baby E, also needed resuscitation at birth. Both were suffering from a rare condition where twins share a placenta in the womb (monochorionic pregnancy) that can cause uneven blood flow and uneven blood volume that can threaten the growth and survival of one or both twins.

Baby G was born weighing just 535g (1lb 2oz). According to Imperial College’s neonatal mortality calculator, the chance of survival for a baby of that weight is just 44 per cent.

Baby I was also born at just 27 weeks, while Baby J was born with NEC and a perforated bowel. Baby K also only weighed 692g (1lb 8oz) after being born at just 25 weeks while Baby N suffered from haemophilia.

Imperial College figures show that one in five babies born under 1,000g will die.

And there were other factors that made them more vulnerable.

Six of the babies were twins or triplets. Twins are three and a half times more likely to die as newborns and triplets five times compared with single births. Conditions which impact one twin or triplets are likely to affect the other.

A study in 2005 of deaths in neonatal units in London by King’s College found that birthweight below 1,500g and length of gestation below 32 weeks was found in around half of all preterm deaths. The vast majority of the babies in the Letby case had one or both of those risk factors.

Several of the babies also experienced medical accidents while on the ward including four incidents where umbilical venous catheters (UVCs) were placed wrongly or fell out, a mistake that can sometimes lead to deadly clots.

In the case of Baby A, staff failed to give fluids for four hours while Baby H had breathing difficulties from birth and the prosecution accepted she had “suboptimal” treatment in the unit.

As well as delays in intubation, she was left with butterfly needles in her chest which may have punctured lung tissues causing air to collect in the chest. Letby was found not guilty on one count of her attempted murder and the jury failed to reach a verdict on a second count.

The unit was extremely busy and often did not meet British Association of Perinatal Medicine standards, which stipulated that babies in intensive care should have one-to-one care.

An original investigation into the deaths by The Royal College of Paediatrics and Child Health (RCPCH), found that, as well as staff shortages, there had been “higher activity and lower admission birthweight than average during the period corresponding to an increase in mortality”.

There were also a significant number of stillbirths at the maternity unit in 2015.

The Royal College’s review found that there was “no definite causal correlation” between any of the cases flagged although it did recommend that new UVC guidance should be issued.

Likewise, all the deaths examined by pathologists and a coroner were deemed to be natural causes or unascertained.

The prosecution made much of the fact that the increased deaths stopped after June 2016, when Letby left the unit, but, as the jury also heard, in the same month the neonatal unit was downgraded due to concerns about the standard of care. It was no longer allowed to care for high-risk babies. Two new consultants were also added to the roster to relieve pressures.

The defence did make the case that the babies were very premature, and the jury carefully considered the expert evidence, eventually deciding it was more likely that Letby killed or attempted to murder the infants than that their deaths were from natural causes.

Commenting on the verdict, criminologist Prof Wilson said: “I did not feel that [Letby] was adequately defended on several of the key prosecution notes; however the jury reached a defensible verdict.

“If I was asked: was justice done? I’d have to conclude ‘Probably’, I would go no further than that.

“When I was waiting for the verdicts to come in, if they had come back as 22 acquittals I would not have fallen off my chair.”

One of the key planks of the prosecution case was that Letby injected air into the veins or stomachs of babies in her care, causing irrecoverable collapse. The nurse was said to have murdered or attempted to murder 10 babies by this method.

The suggestion was first made by Dr Ravi Jayaram, a neonatal consultant on the Chester unit, and supported by prosecution defence experts Dr Dewi Evans and Dr Sandie Bohin, a consultant paediatrician at Medical Specialist Group.

The theory relied on a journal paper from 1989, authored by Dr Shoo Lee, an eminent Canadian neonatologist, which detailed 53 cases of air embolism in newborn babies.

Several of the babies who died or collapsed had “flitting patches of pink” over mottled or blotchy blue tinged skin, which the prosecution experts claimed was a clear sign of air embolism.

Following Letby’s conviction, Dr Lee gave evidence during the application to appeal, warning that the skin discoloration was not unique to air embolism, and said that none of the babies appeared to have suffered from the condition.

Dr Lee said that air embolism would present as pink blood vessels standing out against a background of blue skin, and appear significantly different to a general discoloration or mottling of the skin.

“There is only one diagnostic of an air embolism – pink vessels against a particular background,” he told the court.

“When air bubbles go into blood vessels they oxygenate the blood and the vessels appear pink on a blue background. Other [types of] discoloration cannot be used to diagnose air embolism.

“If you do not see pink vessels on blue background you can see that [although] there’s a problem, you can’t say it’s air embolism.”

Commenting on the case of Baby A, Dr Lee said the symptoms described by Dr Bohin at Letby’s trial were not indicative of an air embolism, and said the doctor had been wrong to draw such a conclusion simply because they had excluded other causes.

He said: “Air embolism should never be diagnosed by exclusion. The rash she [Dr Bohin] described is not diagnostic of air embolism. Air embolism is very specific.”

Moving on to Baby D, Dr Lee said experts were wrong to conclude that skin mottling was indicative of air embolism, warning that such discoloration was generic and could have been caused by a circulatory collapse.

Dr Lee’s paper also described embolism occurring when oxygen was delivered with force through ventilation.

It is a different scenario to that which the prosecution claim happened – air being injected into the babies through tubes.

Dr Hall, the defence paediatric expert, said: “What was being alleged was that air was being injected into the veins. Now air contains 78 per cent nitrogen, which doesn’t get absorbed across membranes very readily.

“Relatively recently I did eventually find a description of what happens when a neonate inadvertently receives an injection of air into a peripheral vein and it didn’t report transient skin changes, what it reported was that the baby’s back went blue for several hours.”

He added: “Both the barristers and the medical witnesses talked in terms of ‘we’ve now established that these skin discolorations are clearly due to air embolism’ but there is little evidence for that.

“Babies who are very sick can get variations of skin discoloration.”

Police gather evidence at Letby's home
Police gather evidence at Letby's home - Peter Byrne

Several studies have also shown that the resuscitation process itself can trigger a fatal air embolism even if one has not been present before.

In 2015, Dr Abid Qazi, a former NHS paediatric surgeon now working in Pakistan, published a case report about the death of a baby from air embolism.

The baby, who was feeding well through a nasal-gastric tube and receiving intravenous fluids, suddenly collapsed with similar blue-purple skin changes to some of the babies in the Letby trial.

Dr Qazi said there had been no obvious cause for the collapse but it was later determined the baby did die from an air embolism. The team concluded that the fatal embolism had occurred during the resuscitation process itself.

When asked to review the case of Baby A by The Telegraph, Dr Qazi said he was doubtful about the conclusions drawn by the prosecution experts.

“I’m very sceptical about the diagnosis,” he said.

“Air embolism has been reported in several cases through the natural course of illness and I personally would doubt significantly to diagnose air embolism [on the X-ray evidence].

“I have been closely following the case of Lucy Letby and I believe she has been a victim of the poor NHS system.”

One neonatologist told The Telegraph that it would take a significant amount of air to cause an air embolism and that could not be done simply by using a syringe.

The consultant, who wished to remain anonymous, said he had only come across one case that was noticed during resuscitation.

“That was the only case I remember in my career, in which we could definitely say there’s air embolism because, as we were resuscitating the baby, we were getting air out from the umbilical vessels which were connected to the heart, which means it was in the circulation,” he said

“But apart from that, it’s not easy to prove at all. It’s a very difficult diagnosis to make, and I think it’s a diagnosis of exclusion.

“Certainly a small amount of air in a thin syringe will not make the difference.”

Dr Qazi said the misplacement of the umbilical venous catheters – which occurred in several cases – also had the potential to cause problems.

“Even a well positioned UVC catheter can cause, rarely though, serious life-threatening complications,” he said.

In 2021, Anna Crawford, assistant coroner for Surrey, called for guidelines from the National Institute for Health And Care Excellence (NICE) on the use of the catheters after a newborn baby died when doctors failed to realise the UVC was wrongly positioned.

As well as air being injected into the veins, the prosecution also alleged that in some cases, air or milk had been injected into the stomach. They pointed out that several of the babies had distended abdomens.

But experts said there were several reasons why newborns might have air in their stomachs.

“The care of newborn babies, in particular premature babies, often involves the blowing of gas into the lungs at pressure, and so they are constantly getting gas blown into their stomach,” added Dr Hall.

“They will have a naso-gastric tube going into the stomach to decompress them, but that sometimes works and sometimes doesn’t work terribly efficiently.

“The principle that gas is being blown into these babies in this trial and that it can cause abdominal distention is difficult to dispute.”

Before the death of Baby C, an X-ray was taken that showed an “abnormal build-up of gas in his abdomen”.

“One of the possible causes of that has to be an obstruction of the bowel,” said Dr Hall.

Premature babies are often treated with a process of Continuous Positive Airway Pressure (CPAP) and many of the babies in the case had this kind of breathing support.

A sympathy card from Lucy Letby to the grieving parents of a baby girl
A sympathy card from Lucy Letby to the grieving parents of a baby girl - Cheshire Constabulary/Crown Prosecution Service

A paper published in 2020 by doctors from the Neonatal Intensive Care Unit at Westmead Hospital in Sydney suggested that a condition called CPAP belly syndrome can sometimes occur, with similar symptoms to NEC.

Writing in the International Journal of Clinical Paediatrics, the authors described the case of a newborn suddenly developing a massively distended abdomen and feed intolerance, which led to a “crisis point” with extremely low oxygen levels and blue-purplish skin.

They concluded: “In our experience, the clinical presentation of CPAP belly syndrome in extreme preterm infants can be dramatic, mimicking acute NEC scare, necessitating urgent diagnostic evaluation.”

In the case report, the large amount of gas in the stomach stopped the chest being able to rise, meaning the baby could not breathe and needed immediate intubation.

A similar breathing issue arose in the case of Baby C, but the prosecution experts claimed it had occurred because Letby had injected air into his stomach, stopping his lungs from functioning.

However, in the days before his death, Baby C had suffered a “very distended gas-filled stomach” which the prosecution accepted “can be a consequence of CPAP”.

In the application to appeal, Ben Myers KC, one of the defence barristers, argued that the prosecution experts had insufficient clinical expertise of air embolism, and warned their diagnosis had been given “on the hoof” without any scientific basis.

“None of the experts who gave evidence of air embolisms had the knowledge to do so,” he argued.

The Court of Appeal acknowledged that neither Dr Evans nor Dr Bohin had any significant direct experience of patients with air embolism.

But the judge rejected the suggestion they were not qualified to give evidence on the issue.

They also ruled that if the defence team had wished to call Dr Lee to give evidence during the trial they should have done so at the time.

Letby was accused of attempting to murder two babies by poisoning liquid nutrition bags with insulin.

There was no direct evidence that the bags had been tampered with, as tests were never carried out on the contents, and the prosecution instead relied on blood samples from the babies.

Insulin regulates blood sugar and too much can cause blood sugar levels to crash dangerously, a condition known as hypoglycemia, which can be life threatening.

In the first case, Baby F had suffered an unexpected drop in blood sugar together with a surge in heart rate in the early hours of Aug 5 2015 when Letby was on shift.

It was the prosecution’s case that Letby had deliberately injected insulin into the bag.

Baby F’s low blood sugar continued even when Letby was not on the ward, and the prosecution claimed that she had also put insulin in a bag in the fridge, which had then been selected by other nurses.

Blood tests from Baby F showed high levels of insulin, yet low levels of another hormone called C-peptide.

C-peptide is produced alongside insulin in the body, so if there is none present that suggests the insulin present has not been made in the body, and so must have come from outside.

A second baby, Baby L, also suffered an alarming dip in glucose levels on April 8 2016. Tests revealed high levels of insulin but on this occasion C-peptide was within the normal range.

However Prof Joseph Wolfsdorf, a specialist in child hypoglycemia at Harvard Medical school, said the results “made no sense”.

“When serum insulin is extremely high (caused by injection), the beta cells of the pancreas stop secreting insulin (and C-peptide). One expects, therefore, to have extremely low or unmeasurable serum C-peptide concentration.

“In one of the babies, serum insulin was extremely high and yet C-peptide was within the reference range, i.e. not extremely low or unmeasurable.

“Therefore, one has to be highly suspicious that the high insulin level was caused by exogenous administration.

“If the sample was obtained at the time plasma glucose was low (hypoglycemia), a C-peptide concentration within the reference range makes no sense.”

Both babies had suffered from blood sugar problems before the catastrophic falls. Just a few hours after birth, Baby F had high blood sugar and was prescribed insulin. Baby L’s blood sugar was found to be low in the hours after his birth and he needed treatment with a sugar infusion.

When the tests were returned for Baby L, insulin levels were at the top of the scale that the equipment was capable of measuring, around 10 to 70 times higher than in a premature baby.

Both babies survived, and initially there was no suspicion of foul play, so the tests were not checked at a proper forensic lab, but instead carried out at a clinical lab of the Royal Liverpool University Hospital.

An NHS guidance note from the Liverpool lab warns that such testing cannot be used to infer if insulin has been administered artificially.

“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 guidance reads.

The samples were never checked by forensic experts and the defence were not able to re-test the blood as it has since been disposed of.

Dr Charline Bottinelli, an insulin expert from the Laboratoire LAT Lumtox, in France, said: “Techniques generally performed at hospital for insulin research are immunoassays. These techniques allow the detection of very low insulin concentration.

“However, it doesn’t allow us to formally identify the type of insulin and to distinguish human insulin from synthetic analogues”

She added: “A high level of insulin in blood could suggest exogeneous (outside) administration but also disease such as insulinoma.”

In the case of Baby L, the prosecution said the insulin readings were “grossly abnormal” yet by the time the tests were returned the infant had recovered.

Dr Bottinnelli said levels of insulin recorded were so high for Baby L that it was “highly probable” that “severe hypoglycemia and dramatic consequences” would have been expected.

The Countess of Chester neonatal unit first noticed an alarming spike in the number of newborn deaths in 2015.

With fingers pointing at Letby, the nurse was removed from her post in 2016 and the Royal College of Paediatrics and Child Health was brought in to investigate the upsurge.

The review team quickly found serious concerns with the unit, uncovering a pattern of “insufficient senior cover and a reluctance to seek advice” from consultants.

Reviewers warned that senior doctors should have been more available, given the vulnerable nature of the premature babies on the ward, while the number of nurses on the unit was frequently less than the recommended levels.

There was a 21 per cent shortfall in nurses between 2014 and 2015 and agency staff were often drafted in to plug the gaps.

In every board meeting throughout 2015 and 2016, the neonatal unit had the most “red” ratings for staffing levels out of all of the departments in the hospital.

Nurses were stretched thin, expected to help out on the maternity unit and give antibiotics to infants on another ward alongside their normal neonatal duties, the Royal College found.

In late 2015, one of the senior paediatricians emailed the hospital’s chief executive, Tony Chambers, to report that staff on the unit were “chronically overworked” and “no one is listening”.

“Over the past few weeks I have seen several medical and nursing colleagues in tears,” she wrote.

Doctors were working shifts of more than 20 hours and the unit was so busy that “at several points we ran out of vital equipment such as incubators,” the consultant warned.

“This is now our normal working pattern and it is not safe. Things are stretched thinner and thinner and are at breaking point. When things snap, the casualties will either be children’s lives or the mental and physical health of our staff.”

An inspection by the Care Quality Commission (CQC) in February 2016, found the neonatal unit was under-staffed and “lacked storage space and resources for the care of patients who required strict infection control measures”.

The hospital knew conditions needed updating, and had launched a £3 million fundraising appeal in 2012 to build a bigger and more modern baby unit, but by 2015, had only managed to raise half of the sum needed.

Rooms in the unit were small with cots placed too close together “making nursing more of a challenge” and it had poor lighting.

The Royal College’s report warned: “There is insufficient storage space resulting in many pieces of equipment being stored in corridors.

“Direct visibility from one area to another is poor, and infants are moved regularly to accommodate acuity – an extra risk in the system.”

When deaths started spiking, the trust carried out post-mortem examinations and reviewed the circumstances in their “mortality and morbidity” meeting.

But the Royal College said it was concerned that the local safeguarding Child Death Overview Panel did not appear to know about the cluster of deaths and that a rapid review had not been carried out for some of the cases within five days.

In cases of unexplained deaths or patient harm, a review is usually carried out within five days so that immediate remedial action can be taken.

Dr Stephen Brearey, the head of the neonatal unit, claimed he had tried to flag serious concerns about the increase of deaths by October 2015.

The Telegraph understands that the staff also tried to alert CQC inspectors to the spike in deaths arising from a patient safety issue in February 2016, saying they had informed hospital management but were struggling to be heard.

According to a source inside the hospital the CQC did not ask for any details about what the patient safety issue was, and despite telling the doctors they would return to discuss it at a later date they never did.

A court sketch of Letby giving evidence during her trial
A court sketch of Letby giving evidence during her trial - Elizabeth Cook

The CQC told The Telegraph that consultants had not raised increased deaths during the February 2016 inspection, and it was not aware of the issue until the end of June.

A spokesman said: “Some consultants (including those working in the neonatal unit) at the trust shared concerns with CQC during a focus group held as part of the 2016 inspection.

“Those concerns related to staffing levels, a lack of support from senior management and consultants who felt there was a culture of bullying and where concerns they had raised with management were ignored.

“We followed up directly with the trust’s medical director that same day to relay those issues so that action could be taken in response.”

The Royal College’s investigation found there had been little attempt to get departments together to “deal proactively with the increased mortality” and the trust appeared so unconcerned by some of the deaths that they were recorded in its risk register as “green-low risk of harm”.

In July 2016, Letby was removed from the ward and the trust downgraded the unit to a Level One, meaning it was only allowed to deal with babies who did not need intensive care, and mostly those born after 32 weeks’ gestation.

There was an immediate improvement and no more babies died unexpectedly after this point.

“Since the temporary redesignation, staff reported feeling calmer and more confident and morale/sickness has improved,” the Royal College’s report said.

“The pressure has reduced and the unit is operating more in line with BAPM (British Association of Perinatal Medicine) staffing standards.”

A neonatal consultant told The Telegraph that the decision to downgrade must have been made long before July 2016, showing a long-term concern that went beyond the spike in deaths attributed to Letby.

“You don’t wake up on a Monday and say, ‘actually, on Friday, we’ll downgrade’,” he said.

“The unit knows it a year in advance at least. And if you want to appoint people in place you have to get your business cases approved or lined up at least six months to a year in advance.

“Which makes me think that there was enough worry there – 2016 happened to be the year when it went down, but the planning to do that would have been happening over the last couple of years, which would mean that they were already worried about the unit, irrespective of the presence of this person (Letby),” he added.

Photographs of Letby's bedroom shown at the trial
Photographs of Letby's bedroom shown at the trial - Cheshire Constabulary/CPS

In court, Letby’s barrister, Ben Myers KC, argued that “suboptimal care” played a part in the deaths of seven of the infants and that Letby had been scapegoated to cover up failings on the unit.

Mr Myers said that the Countess of Chester Hospital “failed” in 2015 and 2016 and “blame for absolutely everything” had been “heaped” on Ms Letby.

The unit was “understaffed”, treatment was “hurried”, “mistakes made” and records “not kept”, Mr Myers claimed.

He told the jury there had been a “marked increase” in the number of babies on the unit during 2015 and 2016, “too many” of which had had “additional requirements”.

Referring to Baby J he said the hospital was “well out of its depth” while he claimed that  With another – Baby K “should not have been in the Countess of Chester Hospital in the first place”.

Mr Myers said “one way or another”, the unit had “failed” and the case against Letby was a “prime opportunity” to “hide” poor outcomes.

There were also concerning examples of poor care before the rise in deaths in 2015 and 2016.

An inquest for a newborn who died in 2014 found that doctors had inserted a breathing tube into the baby’s oesophagus rather than his trachea, ignoring several indications that the tube was misplaced.

Medics ignored five warning signs – assuming the equipment was faulty. The baby’s mother said staff shortages meant blood tests and X-rays were not assessed for seven hours and there was only one consultant on duty when her son began to deteriorate.

Beyond the staffing issues, there were also problems with the plumbing on both the neonatal and maternity wards. The pipes often leaked or were blocked, causing sewage to rise up into the toilets and sinks.

Giving evidence at Letby’s original trial, Lorenzo Mansutti, the hospital’s plumber, said that on one occasion he recalled human waste or sewage entering the neonatal intensive care room from the drains of a ward above.

Asked by Mr Myers how often he was called out for “these sort of issues” in 2015 and 2016, Mr Mansutti replied: “For the whole building it was maybe weekly.”

Even years after the death spikes, there have continued to be problems in the hospital.

In a 2022 CQC inspection report, the trust was served with two warning notices and told it had to make significant improvements in its maternity care.

Dr Dewi Evans was the prosecution’s key expert witness, whose evidence in the trial was central to the case against Letby.

But questions have been raised about Dr Evans’ expertise and impartiality, with a judge in another case in which he was involved describing his evidence as “worthless” and accusing him of making “no effort to provide a balanced opinion”.

A former paediatrician, he retired in 2009, setting up a company called Dewi Evans Paediatric Consulting providing expert medical advice in legal matters such as child abuse cases and criminal trials.

In 2017, just a month after the first report had been made to Cheshire Police, Dr Evans read about the investigation in the local paper and sent an email to the National Crime Agency, which read: “If the Chester Police have no one in mind I’d be happy to help. Sounds like my kind of case.”

After being invited to review some of the deaths on the neonatal ward, it was Dr Evans who suggested the possibility of air embolus as a cause of death. This was despite acknowledging he had never actually seen a case himself.

In other deaths, it was Dr Evans who determined likely methods of murder, including deliberate over feeding and smothering.

Prosecutors relied on him to explain the complex medical evidence to the jury and ensure they did not become bamboozled by the science.

But Letby’s lawyers became concerned that rather than serving as an expert witness, Dr Evans had been acting as an investigator, putting forward theories rather than expressing a view on them.

At one stage in the trial, Ben Myers KC, Letby’s barrister, asked for his evidence to be struck out, but the judge, Mr Justice Goss, dismissed the application.

Further doubts over the credibility of Dr Evans’ testimony were raised when it emerged that his expert evidence in another case had been lambasted by a Court of Appeal judge.

In February 2020, a newborn baby sustained nine fractures to his arms and legs, and a court ruled that the injuries had been inflicted by the parents.

The family appealed against the ruling and instructed Dr Evans, who submitted a report claiming the injuries could have occurred accidentally because of the baby’s low vitamin D levels.

He suggested a fracture to the baby’s arm could have been caused when it moved during breastfeeding, while a broken ankle could have been the result of a parent gripping too tightly during a nappy change.

But in a damning ruling, Mr Justice Jackson wrote: “It is of great concern that the parents and the wider family might have been encouraged by this opinion… The report is, I regret to say, worthless.”

The judge criticised Dr Evans for a “lack of transparency”, and accused him of being ill-informed about the case.

He concluded: “Finally, and of greatest concern, Dr Evans makes no effort to provide a balanced opinion. He either knows what his professional colleagues have concluded and disregards it, or he has not taken steps to inform himself of their views.

“Either approach amounts to a breach of proper professional conduct. No attempt has been made to engage with the full range of medical information or the powerful contradictory indicators. Instead, the report has the hallmarks of an exercise in ‘working out an explanation’ that exculpates the applicants.

“It ends with tendentious and partisan expressions of opinion that are outside Dr Evans’ professional competence and have no place in a reputable expert report.”

When this was raised during the Letby trial, Dr Evans said: “This is the first judgment that has gone against me in 30 years. I have prepared dozens and dozens of reports for the family court. I’m in huge demand for opinions in the family court because of my track record as a witness. This is a one-off for me.”

In Letby’s appeal, the defence reiterated its concerns that Dr Evans did not have the relevant expertise to give evidence in the trial.

However, Court of Appeal judges ruled: “He was a highly experienced paediatric consultant with decades of clinical hands-on experience with neonates. He certainly had sufficient knowledge to render his opinion of value.”

When jurors are asked to determine if a medical professional has killed their patients, they are not only deciding if a clinician is a killer, but whether a murder has taken place at all.

According to statisticians it is extremely difficult to distinguish if clusters of deaths arise from criminal acts, or coincidentally from other causes.

While a group of deaths may seem extraordinary, the chance that a serial killer is roaming a ward is also extraordinary.

In the 2022 Royal Statistical Society (RSS) report, the authors warned that even if there was a one in ten million chance that a cluster of deaths would implicate a health professional, there are so many doctors and nurses that “it is inevitable we will falsely incriminate innocent people”.

“The prosecution’s theory is typically that a medical professional, previously trusted to perform critical life-saving functions has unexpectedly, and sometimes inexplicably, chosen to murder patients in his or her care,” the authors wrote.

“While history has shown that humans are capable of such behaviour… proven instances are extremely rare.

“In the absence of any other evidence of misconduct the prior odds of any other medical practitioner engaging in such conduct must be considered extremely low, in the order of one chance in millions.”

Letby's parents attend the trial
Letby's parents attend the trial - Peter Byrne

The failure to understand coincidence and the correct interpretation of statistics and the science has led to several serious miscarriages of justice.

The Dutch nurse Lucia de Berk was convicted of seven murders and three attempted murders in 2004, after a criminologist claimed the chance of so many deaths occurring when she was on duty was one in 342 million.

Prominent statisticians came forward to state that the modelling was over-simplified and unrealistic, calculating the real probability of such deaths occurring was as high as one in 25.

The case was retried in 2010 and, in light of the revised statistical evidence alongside new medical evidence, de Berk was acquitted.

Criminologist Professor Wilson said one of the problems with the Letby investigation was that it was never obvious from the outset whether any crime had actually taken place.

“Rather than say that the deaths and incidents were a consequence of the infrastructure; of the staffing levels; or the experience of the consultants; or the lack of finance to upgrade the facilities… they alighted on the idea that the incidents and deaths were the result of someone deliberately taking a child’s life,” he said.

“It seems to me that should be the last thing that one would consider.

“I have studied nurses who kill in a hospital setting academically and produced peer reviewed research about that phenomenon and it seems incredibly rare, so the idea that this must be the cause of what was happening in Chester was always a very weak basis on which to begin.”

The RSS report warned that investigators “must always bear in mind” that there may be innocent explanations for apparent and even striking correlations between a medical professional’s presence and deaths or other incidents.

In the early 80s, a cluster of 43 deaths at a neonatal unit at Toronto’s Hospital for Sick Children led many to suspect a nurse was poisoning the babies with digoxin.

It was later discovered that new artificial latex products in feeding tubes and bottles could have been responsible.

Likewise an increase in neonatal deaths in England raised suspicions until a statistician linked the timing of the deaths to a change in the supplier of milk formula.

In 1999, English solicitor Sally Clark was found guilty of murdering her two infant sons after  paediatrician Roy Meadow testified that the chance of two children from an affluent family suffering unexplained cot death was one in 73 million.

Meadow had based his calculations on the children’s deaths being independent of each other, but the Royal Statistical Society issued a statement stating a second case in a family is far more likely because of shared genetic and environmental factors.

After serving more than three years in prison, Clark was released, although developed severe psychiatric problems and died of alcohol poisoning four years later.

In 1988, Jane Bolding, an American nurse who worked at Prince George’s Medical Centre in Maryland, was charged with murdering patients, allegedly with potassium chloride.

The key evidence was that she had been on duty when 57 cardiac arrests had taken place, and the US Centres for Disease Control concluded that her patients were 47.5 times more likely to die on her watch.

However a judge threw out the case saying that the statistical evidence had “failed to supply the missing link that would connect the defendant with the alleged criminal act” and warned that “the state’s reach hopelessly exceeded its grasp”.

The judge came to the conclusion even though Bolding had originally confessed to the killings, an admission which she later retracted.

Letby also appeared to confess on a Post-It note found at her home which read: ‘I am evil, I did this.”

But Prof Wilson pointed to the case of Kathleen Folbigg, an Australian mother who was convicted of murdering her four children, despite the fact there was no direct evidence.

The prosecution presented her diary entries, written at a time when she was clearly in trauma in which she made statements such as: “All I wanted was to shut her up.”

She was pardoned in 2023 after it was shown that her children all had a genetic mutation that had caused their deaths.

“If you have been accused of something of this magnitude there are all kinds of ways you express how you are feeling because you are experiencing trauma, have depression, are reacting to stress and the (Letby) notes seem to me to reflect that as much as guilt. In fact I didn’t think they did reflect guilt,” added Prof Wilson.

“Letby didn’t share the other common features of healthcare serial killers that we have uncovered. She did not make her colleagues feel nervous. Healthcare serial killers often come with an institutional history, a kind of institutional culture where the other nurses would go ‘Oh I don’t like being on shift with him or her’, they feel uneasy being around them.”

Letby will spend her life in prison
Letby will spend her life in prison - Cheshire Constabulary

More recently, in 2016 the Italian nurse Daniela Poggiali was found guilty of killing two of her patients, with the prosecution claiming the death rate soared on her watch.

After a lengthy appeals process, in which statisticians showed the flaws in the shift rota evidence, both convictions were eventually overturned with the judges ruling that no murders had taken place.

Prof Green, co-author of the RSS paper said: “As the RSS report makes clear, correlation is not causation, and it is essential to consider alternative explanations for the observed outcomes.

“(In the case of Letby) there may have been no crime in the sense of any malice by any one clinician or nurse.

“There may have been incompetence, you can’t actually rule that out – but it’s not a crime to be bad at your job. And remember that these were desperately ill babies in a really sick, badly managed, hospital.”

He added: “Misuse of statistics in court can be very damaging to British justice.”

Letby’s defence team did not call any scientific, medical or data experts. Experts said they were concerned that there may have been a miscarriage of justice.

The RSS has repeatedly warned there is little training in matters of statistics for legal professionals and said it would like to see further interaction between legal and statistical communities.

Defence expert Dr Hall added: “I don’t know whether Lucy Letby is guilty or innocent, but what I do feel is she did not have a fair trial. And the reason is that some of the information given to the jury was flawed, and some was misinterpretation.

“My concern is that the legal system allowed it to happen.”