Lucy Letby thread

[Words]

Starting this as don’t think we have a new one.

Did you not know this Vivienne?

She was there for 10 out of 13 deaths, and had finished her shift an hour before for 2 others. The only other one was a baby that never made it onto her unit as they died before, possibly on the maternity unit (not sure why they're even counted actually) so she was there for a hell of a lot of deaths and it's possible she was there at the point of collapse for the two who died after she'd left.

Then there were 4 who died at other hospitals (and we know she was there at the point of collapse for at least baby K)

The unit was dirty, chronically understaffed and taking on babies who were far sicker / more vulnerable than they were rated to be caring for. The consultants didn’t come out of the trial looking at all good. Rounds just twice a week, and reportedly dismissive and bullying the nurses on duty (not just Letby either).

This is all just general stuff that doesn't in any way explain what happened to the babies.

OK? So what is their job after they've said no amount of "coincidental" patterns is suspicious? Seems like the statistician had her say, they took it on board and she was no longer needed. Why all the outrage over letting her go? What more help could she be to the case?

You are the one claiming statistical significance (“she was always there”, “what are the chances”) without data. You are making this about statistics without understanding what you are doing.

Well if she's guilty the statisticians are wrong and it probably IS statistically significant!

The outside agency was never told a nurse was suspected of harming patients.

The consultants opted to ignore these assessments and continued to blame Letby.

What else do you expect them to do when they were sure she was responsible?

The statistician was not allowed to do her job. The full data was not provided to her, she was sacked and the discussion of other deaths and collapses was prohibited by the judge. This is the whole point.

Some one-sided cherry-picked “stats” were presented in court, people like you think they were meaningful or proved something, but they proved nothing.

Your last comment once again shows your total lack of understanding about statistics. You are obviously determined not to educate yourself on the subject so there’s no point in bothering to try to explain again.

The outside agency was told of these suspicions, though not early in planning. They were told consultants suspected a connection on arrival. They met with the consultants, and the RCPCH notes of that meeting published by Thirlwall show that the consultants had undertaken a "war gaming" exercise to consider how babies might have been murdered - insulin, air embolism, injection of chemicals.

Jane Hawdon was not told about the consultants' suspicions beforehand, but she was told after producing her report, and her response was to advise the hospital executive on how to handle the consultants' egos.

But anyway, it doesn't matter. The prosecution case was that no natural causes could explain the deaths. But the various autopsies and reviews found natural causes. It was always open to the pathologists and reviewers to state that they couldn't find a cause of death (as most did for baby A). Unascertained cause of death in a baby is a known possibility.

To add to that, air in stomach, air embolism, overfeeding are all "methods" that could be accidental. The first two can also be progression of natural disease. So nobody would have to be told murder was suspected to find them, especially with the established concerns about the hospital's standard of care.

Some one-sided cherry-picked “stats” were presented in court, people like you think they were meaningful or proved something, but they proved nothing.

Nope it was all the other evidence. Can you actually say how long they dedicated at trial to the shift chart? Anything being "cherry-picked" is very much in question and not a fact.

Your last comment once again shows your total lack of understanding about statistics. You are obviously determined not to educate yourself on the subject so there’s no point in bothering to try to explain again.

All they're going to say every time is "that's not statistically relevant" even IF there are multiple murders. So the police, rightfully imo decided the statistician wasn't relevant either. If you've got all this evidence proving she's most likely a murderer and then a bloody mathematician comes along and says "no she's not because you're using stats wrong" how long do you think it'd take before you got fed up with that?

I think the fact you are determined to say that even IF she's a murderer, the stats aren't significant is exactly why we don't need any statisticians involved in the first place holding up the investigation.

If they'd dropped the statistician and then not made statistical claims in the prosecution, that would be fine. The problem is that they did make many statistical claims, which were never vetted by a statistician. Every time they said that finding a given pattern of data in multiple cases is improbable without murder, that was a statistical claim. There are well-developed ways of testing such claims, as explained in the Royal Statistical Society report. The report is written very accessibly for non-statisticians. It explains why simple "gut feeling" is not a reliable guide in these cases.

"Statistically significant" means highly unlikely in a specific scenario. You have to say what the scenario is, then compute the chance of various outcomes in that scenario (e.g. chance of 2 deaths in a year, chance of 3 deaths, etc). Then you can compute the chance of the outcomes that you saw, and see if it's very low. If it is, that's strong evidence against the specific scenario you were testing, though not against other scenarios. But you have to do each of these steps. E.g. to test whether a nurse being at 10 of 13 deaths is statistically significant, you need to specify (and justify) your assumptions about the risk of death on a shift, whether different shifts have different risks, whether some nurses get assigned more risky shifts, etc. It's a detailed analysis, none of which was done for the trial.

No. Let’s do this one last time.

The prosecution claimed LL’s presence correlated with more deaths and collapses than could be explained by chance. This is presented as circumstantial evidence that there were murders, and that LL committed those murders.

Moreover you, and other commentators have repeated this concept over and over with your “always there” narrative.

So the question is, can this claim - that there is statistical significance - be verified?

• If it can, then that’s circumstantial evidence to add to the rest.
• If it can’t, then the claim is untrue and misled the jury

At the moment we do not have an answer to that question because the data is unavailable. So it is impossible to say whether or not there is a statistically significant correlation.

That is all. I am only talking about statistics at this point, not any other evidence.

If you were genuinely interested you would look to see what the RSS has to say on the matter. But you have discounted them as biased too, haven’t you?

Now she's been found guilty all the so called expert witnesses have come out.

Why are you referring to them as 'so called' expert witnesses? Are you aware of their credentials?

And the fact that they have come out after she's been found guilty says a lot,. Letby is in jail convicted of the worst crimes imaginable, yet so many experts are willing to work pro bono and put their names out there publicly because they are so concerned that the verdicts are unsafe, that's huge.

Look, Shoo Lee isn't a real expert because reasons, and he should've come forward before he even found out about any of it.

They only had the weakest form of evidence, circumstantial and even that was corrupted with bad data. There is absolutely no forensic evidence of infanticide.

That is circumstantial and not evidence of murder. In an NICU, babies die especially if raw sewage is dripping from the ceiling and the tap water is contaminated with a killer bacteria. Especially if they are so short staffed and one nurse volunteers for most of the overtime.

In addition, she requested additional funding for expert witnesses beyond a plumber and it was refused. It’s not like her defence team did not try.

😂This seems to be what people expect.

I believe that they said that none of them followed the trial, I know earlier someone said it was reported on worldwide but the case wasn't even that widely reported on here in Ireland, the day she was convicted I do remember saying it to some people and no one had heard of it.

And if experts had been following the case during, they would not have known that no defence witnesses were going to be called anyway.

And if it got to the point of the trial where Letbys defence only called the plumber do people think that experts would have just been able to insert themselves into the trial at that point? and that the judge would have given them time to prepare reports and then become a witness?

I think people must be basing their understanding of court cases off TV and movies, and think that they are full of surprise last minute witnesses and dramatic ambushes, but in reality both sides know who the witnesses are going to be and they know broadly what they are going to say so that they are prepared in advance.

The only witness I know who gave evidence after coming forward during the trial was the nurse who heard the prosecutions opening statement saying that Letby had turned the monitor off, this nurse came forward to say that that wasn't true, that it was 2 doctors who had turned it off and they had apologised for it, and she had reported it. This was allowed but this was a different type of witness, a fact witness, not expert, AND the police should have definitely interviewed this nurse before as she was the first to respond when Letby called for help, for a collapse that Letby was accused of harming.

So this was a nurse who was actually present, involved in the care and had first hand knowledge of one of the collapses which is a completely different category from an outside expert trying to come forward at an advanced stage of the trial just because of what they're reading in the news

Well circumstantial evidence isn't inherently weak, it can be extremely strong, fingerprints on a gun is circumstantial evidence, but it can be strong or weak depending on the circumstances.

Strong would be if if say it was hidden at the suspects residence and there's no innocent reason for them handling it and there's other evidence linking them to the crime and ballistics testing shows that the bullet came from that gun.

Weak would be if there was a gun found at someones residence with their fingerprints on it but they had valid reasons to always touch the gun, such as being able to verify they were always at a shooting range, and the ballistics being inconclusive and there not being other strong evidence which would point to them being a murderer.

All of the circumstantial evidence in this case was extremely weak because none of it is really out of the ordinary or points to anything like murder, and as you said in this case there is no forensic evidence of this either, so without the forensic evidence showing murder the circumstantial evidence really would need to be extremely strong to actually prove murder.

And that's why this case invokes a visceral bemused Scooby response in so many. The circumstantial evidence preceded any forensic evidence. Babies died - she was there alot morphed into "what could she have done" as opposed to "what did she do" and then became "she must have done x,y,z because she was there" even in the absence of direct forensic evidence.

Our resident "Yeah but" poster will cling to the insulin issue, but even that is based entirely on theory and speculation. The statistics thing does my head in, because what are the odds both babies attacked in this way, allegedly, didn't die? Seems a pretty poor outcome for someone allegedly so invested in killing. And it was introduced in retrospect as the most bizarre "smoking gun" with no means of testing the theory, although I still don't understand why it wasn't, as no babies needed to be involved to demonstrate the tampering, and the likely effects of introducing various amounts of insulin to a TPN bag, and how long it would take to diffuse into a subject.

I cannot fathom how this case came to court. And if the original pathologists were allegedly so incompetent, why haven't we heard about investigations into that? Why have all their cases not been scrutinised, as potentially, they have missed dozens of murders by these methods alone. Obviously I'm being a bit sarcastic here, but honestly, the Swiss cheese nature of the whole thing points to huge systemic failures throughout.

Yep people make a big deal out of the fact there was a 10 month trial, how much of that was devoted to asking whether the mechanisms were actually plausible?

And then of course no mechanism at all was put forward for the liver injury, I have seen in other cases that an expert will say the injury was the 'force of a car crash' but they will often say it's consistent with say stamping for example, these are in cases of child abuse and death in the childs own home, there would have been no witnesses and there would have been opportunity to do it. But no theory at all put forward in court for what Letby could have possibly done to inflict that much force on a baby in a ward where there was other people around.

And the gotchas regarding her answers were ridiculous, it's a 'gotcha' that she asked if they still kept the bag even though that's a normal thing to ask, it's a 'gotcha' that she said she was arrested in her pyjamas or that she said she knew what she was looking 'for', and then corrected it to 'at'. And Dewi says it's the doubters who are the poundshop Poirots??

I always say it but it's just insane that this court case happened in 2022/23.

I keep coming back to the lack of serial killers in the other hospitals.

The graph is the mortality (death) rate increase in neonatal deaths in 2015 compared to 2014.

It shows the top 10 hospitals with the highest increase in deaths in 2015 compared to deaths on their neonatal units in 2014.

And the Countess of Chester isn’t in the top 10. Spoiler alert - it was number 12.

So either there were 11 other serial killers in the UK killing babies in neonatal units, or the rise in deaths was comparable to other hospitals, probably because of poor care, sanitation, lack of qualified nurses and lack of specialist doctors.

Or there were serial killers we haven’t found yet.

There is one strange asymmetry in how statistical vs medical evidence gets handled by the courts. Jurors are told that what counsel say is not evidence; only what witnesses say counts as evidence. But then, the prosecutor says that something is highly unlikely by chance, without citing any statistical expert witness, and that's allowed. If he had said something medical without citing a medical expert witness, that wouldn't be evidence.

We need to see whether nurses in any other hospitals have been searching patients’ families on Facebook 🧐

I can guarentee at least 90% of them have, at least once. The other 10% will be like me and don't have social media and are not nosy enough to set it up, so they will hang around in the staff room when others are searching for families of patients/residents and casually mention, "I wonder how Betty's family is doing these days?" until someone searches for them.

If they’ve taken handover sheets home too we’ve got a murder epidemic on our hands.