Lucy Letby: Have you changed your mind?

[Kittybythelighthouse]

The other thread has had a lot of really interesting discussion but we are running out of pages so here’s a new one for those who are interested in continuing the conversation.

Whether you’re sure she’s guilty, sure she isn’t, or are somewhere in between, I’m interested in hearing how your opinion has evolved (or hasn’t!) since you first heard about the case,

Please try to be respectful - this is a heated topic. Its a matter of huge public interest with a lot of strong opinions, but we are all adults and can disagree with each other in a respectful manner.

Old thread is here (the poll still has a few days left):
https://www.mumsnet.com/talk/_chat/5388914-lucy-letby-have-you-changed-your-mind?page=38&reply=146359313

Thanks for explaining. Here's what you'd need to consider.

The children were already known to be hypoglycemic. This is straightforward testing done on the ward, for a common condition. There is nothing suspicious in children with sepsis being hypoglycemic.

The test results don't indicate the babies would crash "that way". The hypoglycemia and sepsis account for that without the insulin results. The sepsis was treated and hypoglycemia were treated, and the children returned to normal once this was managed effectively.

The insulin test results indicate the poor children would be dead - unless one of two things apply: either hugely elevated results are a testing error, as the hospital assumed and as the lab certainly produced at least once in 2016; or they are accurate and the parameters were physiologically tolerable for these children. If it's the latter, that would suggest as Letby's expert panel are proposing, that particular sub-cohorts of preterm neonates, very early in life, do indeed metabolise insulin differently.

TLDR hypoglycemia is too common to be treated as suspicious in itself, regardless of the results of the tests.

The medical records accidentally uploaded by Thirlwall and then removed suggested that the child's blood sample was taken before Lucy Letby ever met him, and that his hypoglycemia continued after her first shift with him and through several changes of fluids and treatments. So, hard to pin that one on her!

While that child's medical records were online, one could see that Panorama (last year's episode) had literally blurred out the data showing his blood sugar crash again after Letby left the ward, and through two changes of treatments. They claimed he "recovered" once she was off shift.

Look, I don't fancy picking over this all day so this is my last reply on this. I said you were patronising, which you were, not combative. But you are combative now. You're doing something that men often do when they wade into threads on mumsnet: assuming superiority and then getting upset when they aren't given way to. I don't love that, I have to admit. That said, I've no interest in bogging the thread down in back and forths.

"My workings are perfect and, if you had actually bothered to read my post, I mentioned that correlations would change this. But then you actually need to start attaching numbers to the correlation to see how much."

Obviously I read your post, if I hadn't read your post how did I quote and respond to every sentence? You're showing a sliver of anger here that you're pretending you haven't got. That aside, you originally said “unless the results were somehow correlated,” thus framing correlation as an unlikely exception. In reality, given both tests came from the same hospital, lab, equipment, and process, correlation isn’t a remote possibility, it’s the most plausible scenario. Once correlation is likely, the independence assumption behind your 0.0004 figure collapses.

You can’t call your workings “perfect” if they rest entirely on an assumption that doesn’t hold in the real-world context. And the burden isn’t on me to produce an exact correlation coefficient, it’s on you to justify assuming zero correlation before multiplying probabilities. Without that, your number is meaningless here and it's the exact same mistake Roy Meadow made in the Sally Clark case which is quite something in this context.

"This is where I (respectfully) disagree with you. If each single thing has some value in shifting the probability towards guilt, even if not hugely, a whole pile of these things will make a strong case."

That simply isn't how it works. Your logic is flawed because if the individual pieces of evidence are each too weak, unreliable, or flawed to withstand scrutiny on their own, piling them together doesn’t create strength, it just creates a larger pile of weak material. In probabilistic terms, you can’t simply add up “maybes” without accounting for their quality and independence. Weak circumstantial points often share the same underlying assumptions, so the errors and uncertainties compound rather than cancel out.

In the Letby case, where each “shard” is disputed or explainable by non-criminal causes, stacking them doesn’t produce a reliable inference of guilt, it creates a rubbish in, rubbish out problem, where multiplying bad inputs leads to a bad conclusion dressed up as a strong case. This is why courts require each link in a circumstantial chain to be individually sound, not just numerous.

And again, we need to know whether there were any murders in the first place.

"I think you are misunderstanding my analogy here. The similarity is that both defences tried to pull apart the evidence strand by strand, as opposed to challenging the major thrust of the prosecution case. Each tried to introduce shreds of doubt into scientific tests, procedures etc."

I am perfectly capable of understanding you. I know what you're trying to say, it's just not a valid point. The thrust of the prosecution case here was to throw a ton of bad evidence, fake science, bad data, confusion, and irrelevant nothings at a jury over ten months hoping the general miasma of guilt would overwhelm them, and it did. The OJ case was very different. I don't care to pick over the previous defence performance at the trial. It's not relevant now, but it isn't relevant to compare it to the OJ case either.

Again, all that matters is whether or not there were murders to begin with.

Yes. Well put.

Also the court was told that these blood tests must mean insulin was given.

The fact that another baby had similar test results and no such conclusion was drawn means that the jury was misled, probably deliberately given that this case was known about at the time.

I don't understand why this alone hasn't led to the case being retried.

Unless there was a second murderer on the ward that attacked this baby. Though funnily enough no one seems to be looking for one....

I really must say how much I admire how @Kittybythelighthouse responds with grace and integrity to each and every post that tries to undermine her intellect and question her motives for starting a LL thread; all whilst convalescing.

I have very much valued reading this thread and all the contributions from all sides. I’m in the camp that feels a MOJ has occurred and it’s in the best interests of the families, LL, Government agencies and the population as a whole to have a review of the case and the judicial process.

PS I wish Kitty a speedy recovery ❤️‍🩹

This is quite a significant incident, because the prosecution had the files, with police interviews and all, but they were not disclosed to the defence.

Non-disclosure is a major issue and taken pretty seriously by the CCRC.

It's easy to forget with all the attention on them that the medical reports are only three of McDonald's fourteen files for the CCRC - at least it was fourteen already back in March! This was another on the list.

One of them was diagnosed with hyperinsulinemia at Alder Hey or Arrowe Park.

To test for that you need a specialist unit and the insulin test needs to be done fasting. I’m not sure if CoCH covers that.

The expert panel (summary report linked) composed of serious heavyweight experts who hold senior positions in e.g The Karolinska Institute, are "making a big thing" about antibodies binding to insulin (are you a neonatologist?)

You really are being insufferable. If you're going to hang around would you mind reigning in the pompous 'who gave you ladies the right to discuss important things' attitude? You don't have any grounds for assuming superiority to anyone, so I'd appreciate it very much if you stopped doing that.

https://drive.google.com/file/d/1aV4zwwdBYw8Z_E-Tpe9_-iPR7n8cZdFk/view?pli=1

Thank you! I really appreciate that. I'll mend up alright in time🙂

Chiming in to applaud Kitty. Showing us well thought out analysis instead of self-pronouncing perfection.

Here for anyone interested are Mark McDonald's 11 grounds for reviewing the case, back in February. (Apologies, I have 14 in an earlier post and now can't edit). We are aware of more since!

https://thirlwall.public-inquiry.uk/document/letter-from-bhandal-law-to-lady-justice-thirlwall-dated-17-march-2025/

i. Nineteen detailed reports prepared by sixteen experts from seven different countries who are leading clinical specialists in the fields of neonatology, paediatric pathology and paediatric surgery. These international experts have had access to all the medical records, prosecution expert reports, statements from nurses and clinicians on the Neonatal Unit (‘NNU’) between 2015 and 2016 at the Countess of Chester Hospital.

In addition, where relevant, they have also seen summaries of evidence presented to the jury.

In short, they have received and read all and more of the evidence put before the prosecution experts when they drafted their reports. Many of the experts have given testimony before their own domestic courts as expert witnesses and have a full understanding of what is required of
a witness in criminal proceedings.
The expert panel was headed by Professor Shoo Lee, the leading expert in neonatology in Canada, who has published over 400 papers. He has reviewed every new expert report.

This is by far the largest forensic expert neonatal review ever undertaken. The results of which say that there is no evidence of harmful acts committed by Lucy Letby and in fact highlight a litany of errors by the treating clinicians.

ii. Two reports from the United Kingdom’s leading statisticians refuting the premise of
the prosecution case of an unexplained spike in deaths and a coincidence of Letby
being present when babies are said to have died or collapsed. This has been done not
by relying on unproven anecdotal evidence but with the use of extensive research and
data analysis. The experts conclude that the jury were misled as to accurate status of
the data. In short, there was no unusual spike, it was not an outlier, and the staff rota
presented to the jury was incomplete, selective and, therefore, meaningless.

iii. Three reports on the issue of insulin rejecting the hypothesis that exogenous insulin was given to any baby by Lucy Letby. The reports are written by nine internationally
renowned experts in epidemiology, toxicology, biochemistry, biomechanics, statistics, neonatology and engineering. The authors include a Professor of Chemistry and Forensic Science, a Professor in Forensic Toxicology (retired), a Consultant in
Clinical Biochemistry & Chemical Endocrinology, an Associate Professor of the
Department of Statistical Science, a Consultant Paediatric Endocrinologist, an
Emeritus Professor of Paediatrics, and two highly experienced neonatologists, one
based in the UK the other in Canada (in addition to Dr Shoo Lee).

iv. The failure of the prosecution to disclose to the defence that the police had instructed
an expert, met with the expert, taken guidance and advice from the expert and then not proceeded on that advice. This arguably led to the jury being misled on the central thesis of the prosecution case, that there had been a spike in deaths on the NNU and staff rota showed Ms Letby being on duty of the time of each incident.

v. The failure of the prosecution to disclose a medical statement from a treating clinician
which could have had a bearing on the defence approach at trial. [third insulin case]

vi. Evidence from numerus interviews, podcasts and articles from Dr Dewi Evans since the trial, where he arguably undermines his independence as an expert witness.

vii. The failure to disclose a new report drafted by Dewi Evans twelve months after Ms Letby was convicted of murder. This report addresses the cause of death of one of the babies for which Ms Letby was convicted of murder. Not only is this fresh evidence but this again goes to the inconsistency and reliability of Dr Evans.
viii. The failure of the prosecution to adduce before the jury the report from the
Royal College of Paediatrics and Child Health (RCPCH) which, following a full review of the NNU, raised several issues concerning suboptimal care. Importantly,
having met with and interviewed several members of staff (including Lucy Letby),
they raised no concerns against any one member of staff. This report was consistent
with the internal review set up by the Trust. This again was not put before the jury.
This failure led the jury to not have the complete picture in relation to a unit which
seemed to be in crisis and poorly managed by inexperienced and overworked
clinicians.

ix. The failure of the prosecution to disclose to the defence the involvement of a senior
coroner’s officer into the investigation of Lucy Letby and the results of her
investigation.

x. The failure of the prosecution to disclose that the coroner investigating the death of
one baby (for which Ms Letby has been convicted of murder) did not see important
evidence of a hospital procedure which we say ultimately may have led to the child’s
death.

xi. The change of position by Dr Evans on a key element of the case against Ms Letby.
This, we say, not only may have misled the jury but also the Court of Appeal.

Those who find the "confession" note compelling might be interested to read about the psychological experiment that demonstrated that people can frequently be persuaded that they previously committed a violent crime: https://www.psychologicalscience.org/news/releases/people-can-be-convinced-they-committed-a-crime-they-dont-remember.html

Much appreciated! ❤

The OP's post and later posts are admirable.

This piece makes total sense:

"Your logic is flawed because if the individual pieces of evidence are each too weak, unreliable, or flawed to withstand scrutiny on their own, piling them together doesn’t create strength, it just creates a larger pile of weak material. In probabilistic terms, you can’t simply add up “maybes” without accounting for their quality and independence. Weak circumstantial points often share the same underlying assumptions, so the errors and uncertainties compound rather than cancel out".

I have always maintained that LL was not guilty, no concrete evidence provided, just a mishmash of circumstances.

I am convinced there were sinister actions going on in the background with a particular Consultant opting to appear on a morning TV show rather that treat very sick preterm babies.

Heads are going to roll.

Another perspective - if she claimed to be a miracle baby which was the reason she wanted to be a neonatal nurse to save other babies - was she tampering with them in order to save them? Not that it’s any less awful, but could her intent not be murder but to be held a hero by pulling them back from the brink of death? She seemed very keen to work with the most vulnerable babies and had said she found less intensive wards boring. She even stated in a message that in order to get over a death she needed to get back to another poorly baby - which her colleague responded to that mindset as being ‘odd’.

People are generally fine with circumstantial evidence.....as long as there's proof that a murder/crime has been committed.

A woman stabbed to death for example, someone has to have murdered her, and then it can be very easy to see how the circumstantial evidence showed that X person did it.

A woman dying though, and they say she was murdered, but yet it hasn't been proven that she actually was and it could have been due to other medical issues, well circumstantial evidence is not going to be enough to convince a lot of people in that case.

Most of the guilty verdicts were a unanimous 10-1.
I think it's safe to assume that it was the same hold out juror for all of those.

The 3 unanimous guilty verdicts were for the insulin cases and Child O who had the liver injury, so even though that juror did think she was guilty of and capable of killing and attempting to kill babies, they still didn't find her guilty on most charges which shows they really needed to be convinced in order to find her guilty.

Of course we'll never know but I would not be surprised at all if that juror in particular had serious doubts now.

It would have only taken 1 more juror who wasn't sure enough to convict for it to have been a mistrial.

There's just no evidence anything went wrong with nursing care on her shifts, though, and the consultant who ended up having her removed from the ward and led in accusing her to that police confirmed that.

She wasn't even the nurse actually caring for 9 of the 17 babies she was accused of harming or killing.

I don't know about sinister, but imagine the amateur psychological profiling that would be going on if Jayaram was the one accused! Something tells me the word narcissist would feature heavily.

I second this. 🎉

I followed the trial at the time and my thoughts were that she was guilty but would probably be found not guilty as the evidence wasn't strong. It was her behaviour and lies she told that made me feel she's guilty of something. There was a particular incident with a mother finding her baby covered in blood and letby lied about it and was caught out. She also wrote false times on records too. It's possible she was just bad her job and injured some babies by accident. Add that to babies dying for other reasons and it looks like she's a mass killer. I honestly don't know. I don't think we'll ever know the whole story.

Not quite sure why he’s linking a study on hyperglycaemia & insulin resistance to a discussion of hypoglycaemia & immunoassay tests but hey ho.

This is a sad story that has been exaggerated in the press.

Lucy Letby wasn't actually caught out lying about this child. He wasn't covered in blood. He had a little blood, like a goatee his mum said, on his mouth and chin.

Letby told his mother it was probably the tube rubbing the child's throat and that she had called a doctor. The child's doctor's records show she did call him, and he examined the child later. The time this happened at is where Letby and the mother disagree, and that's why the prosecution lawyer called her a liar. He did this often, and the press quoted him, but when you are cross-examining you can say what you like. It doesn't have to be proved and it doesn't have to count as proved in court.

The expert panel found that the child most likely died of a GI tract haemorrhage, because his bowel had been damaged before birth. That bleeding was the first sign of it. The doctor saw the child repeatedly that evening but he was junior and he didn't order a blood transfusion until it was too late.

The child had no post-mortem. The body was never examined because the consultant in charge told parents and the coroner there was no need. That's how sure they were it was a natural, unsuspicious death until it was pinned on Letby years later, just because she was there.