Lucy Letby - have you changed your mind thread 4

[MistressoftheDarkSide]

With thanks to the original poster @kittybythelighthouse and @Tidalwave for continuing the discussion.

Yes, but as far as I can tell the pharmacist didn’t describe a screw cap and there’s no screw cap in the linked photo. I’m just curious where you got that from, so I can go and read it.

The bag would have been more like this one. They come empty for filling via a wide bore tube (the one that has a screw cap but is then clamped), and have two more ports

• one for additives, that has a tamper evident cap that, if used, has a new and different cap added afterwards and
• one for the administration tubing that also has a tamper evident cap that is removed before the administration set can be attached (with a spike).

https://www.integratedmedsys.com/h938738

Anyway, it’s probably not important overall, but there’s so much supposition with this case it’s difficult to know what’s relevant and what isn’t.

I think this question was aimed at me, although I see that Kitty appears to have answered it (incorrectly) on my behalf.

For the purposes of this conversation I don’t care about it because my main point was to highlight something being stated as fact by another poster that wasn’t actually true.

However, more broadly I don’t really care about it as in my view it only exists to demonstrate that Letby had the opportunity to do what the prosecution were alleging she did, rather than being used to say because she was there she must have done it. I also have never understood why people feel it is so important that other deaths on the unit were left off the chart when no one was accusing her of having anything to do with these. What difference would that have made?

But I know many others disagree, I am not a statistician, and no doubt other posters are poised and ready to jump in and tell me how misguided and entirely representative of the wider uneducated, unintelligent masses I am.

Very true. Speaking as someone who falls into that list (present tense, thus up to date experience) we can absolutely have a problem understanding that being very clever in one area doesn't mean you have a good understanding of others.

I’ll see if I can find the previous conversation and share it with you. As I say, I’m not a nurse so I’m not 100% sure either way.

We all have blind spots somewhere. I think statisticians in particular have a heavy load to bear in this respect. Doomed to be Cassandras, constantly witnessing life-destroying miscarriages of justice happen in slow motion but being totally unable to get anyone to listen to them.

Being in yhd right place at the wrong time does not amount to guilt.

And given that Dewi Evans had identified many other ‘suspicious’ events I think it was (and very much still us) in yhd publuc interest to know whether herby was there or not.

After all, she wasn’t there for Baby C whrn Dewi famously claimed, in court, that the baby collapsed and died Becsuse of air in the stomach that was noted to be there in an X-ray taken before Letby had ever met baby C (and then on the stand changed his mind and decided for the first time ever - after no less than eight previous reports about baby C where he had never ever previously mentioned it - abx said baby C had died of an air embolism) Interestingly in his first ever report he had said that we may never know why baby C died and that he was at GREAT RISK OF COLLAPSE.

She also was there for the unplanned and unexpected second hanging of the generic TPN bag selected at random from the store of 5 in the fridge.

I'd agree with you that it was soley to demonstrate that she was there for the incidents that she was accused of (and only them) if it didn't include other nurses' data. The fact that it shows when other nurses weren't there means it becomes statistical data. You're showing a correlation and implying a causation. It implies that all these incidents are suspicious and no others and because there's only one person who was there for all of them then it must have been down to her.

The fact that it doesn't include all the deaths/collapses during the time period (but doesn't explicitly state this) could imply that she was being accused of all of the incidents that happened. There's no indication of the total number AFAIK.

If you really wanted to present a fuller picture you could include all possible incidents (remembering that none of them were considered suspicious at the time) and you could include at least the total number of shifts done by each nurse you're considering, given that there's going to be a wide range. You could include how many shifts worked by each person where nothing happened too. You could also include which doctors/consultants were on duty to see if there was a correlation there and any other staff who were present.

Such dishonesty from a hall monitor! Tut tut.

Let’s be specific. I said that the shift chart was shown “every day”. Clearly none of us know whether it was literally shown every day or not, so clearly I was not being literal. I did not state it “as fact”. It was also completely beside the actual point, as you well know.

“Every day” is, however, much closer to the truth than your claim of “a couple of times” given that the rota was at the very least referenced in opening statements, closing statements, and at several points during the presentation of evidence for each charge, which amounts to A LOT and not a “couple of times” no matter how you slice it.

I’m still waiting to hear how presenting grossly misleading evidence to a jury “a couple of times” or even once is acceptable.

It’s not and that’s the actual point.

I'd agree with you that it was soley to demonstrate that she was there for the incidents that she was accused of (and only them) if it didn't include other nurses' data. The fact that it shows when other nurses weren't there means it becomes statistical data. You're showing a correlation and implying a causation. It implies that all these incidents are suspicious and no others and because there's only one person who was there for all of them then it must have been down to her.

Fine, but it’s not like the prosecution presented it and then just sat back and twiddled their thumbs for ten months. They went through each case in meticulous detail, setting out why they believed she did it.

The fact that it doesn't include all the deaths/collapses during the time period (but doesn't explicitly state this) could imply that she was being accused of all of the incidents that happened. There's no indication of the total number AFAIK.

She was accused of harming the babies she was accused of harming. Surely everyone realises that babies also die on NNUs for reasons other than murder? I still don’t get why it would make any difference to show that she was or wasn’t there when this baby died of congenital defects, or this baby died of a known infection.

Whatever type of nag it was there is no proof that it happened that way.
It was a theory put forward with no evidence, and somebody saying that she must have done it that way.

It is a fucking farce.

The way the figures were presented suggested that she was there for every unexpected death or collapse.

Originally the deaths she is associated with were not identified as the result of murder at post mortem.

The case was built retrospectively on the premise that there were too many unexplained or unexpected deaths - the spike if you will.

Dr Brearey developed a narrative that Lucy Letby was a common denominator. He is known to have sown those seeds before he took any official action - there is a reference to canteen gossip.

Again in retrospect, Ravi Jayaram re-interpreted past events through the lens of these suspicions. I think he's a prime contender for accusations of "attention seeking".

I really don't think it's a solid foundation for justice to assume "unexpected" death must mean murder, advance hypotheses on how it could have been done from an expert who offered his services to the police, and boasted about "never losing a case" which is not his professional remit, and pin it on the nearest person.

Also, given the narrative built, if a child had died on her watch from clearly defined reasons, that wouldn't necessarily exclude her from suspicion, because of the prurient speculation rampant in this case.

They are allegedly examining the records of all babies cared for by Lucy Letby during her entire career, on the assumption she must have been doing "bad things" throughout, despite no concerns or evidence of that.

It's all farcical. And they will use the bare minimum to build further cases if this isn't properly handled.

I still think the implication is there from the way the data is presented. Also, can we be sure all the other deaths/collapses are definitely not suspicious? Presumably some are unknown causes. Presumably also there were some she couldn't have been involved with because she wasn't there. Including them gives a very different picture (I believe that the defense weren't allowed to mention the fact that there were others).

Just as an example, if you have 2 unknown reason collapses and one is included (LL present) and one not (LL not present), would you agree that that would be misleading?

I doubt that the prosecution would regularly remind the jury that babies in NNUs are ill and sometimes they die despite best efforts of medical staff.

When you're including the insulin incident which required something to happen when LL wasn't there and the baby C? incident where the xray presented as evidence was taken before she'd met the baby then it begins to feel like the evidence is being bent to fit the narrative.

Tbh the fact that there are so many methods of murder/attempted murder, which is unusual for a serial killer, it does make me wonder whether they picked any and all incidents where she was there and worked backwards, although they claim it was done independently. AFAIK the police didn't interrogate anyone else about other suspicious incidents where it couldn't have been LL but I could be wrong on that.

Dewi himself said he identified more ‘suspicious’ cases after the trial.

However you were the one who asked about the shift chart a page or so ago.

Of course the implication is there. That's why it was presented in the first place. And it won't do to say there was other prosecution and the defence had time to rebut it, because we don't know (and nor should we) how persuaded the jury were by misleading use of statistics.

Which isn't to blame the prosecution barrister for the shonky stats either. The system allows it and he was acting on his client's instructions. We need reform to be systemic.

That document from the royal college of statistics really needs to be taken seriously by the police, cps and judges. It is a long but interesting read. The investigation into Lucy seems like a case study on how to make many of the errors they highlight.

Some of the people who think she is guilty beyond all reasonable doubt really ought to read it.

rss.org.uk/RSS/media/File-library/News/2022/Report_Healthcare_serial_killer_or_coincidence_statistical_issues_in_investigation_of_suspected_medical_misconduct_Sept_2022_FINAL.pdf

“if a child had died on her watch from clearly defined reasons, that wouldn't necessarily exclude her from suspicion”

That’s exactly what happened with every one of the murders she is currently serving prison time for!

Actually, she didn’t even have to be the assigned nurse. She just had to be “on shift” which became an elastic measure extending beyond the time she was actually on shift anyway.

Some try to brush the chart off as “just showing opportunity.” This isn’t the case.

It’s not what it does. It’s not how it was used. It’s not how it was discussed by anybody, including those who say it now, until the statistics were demolished post reporting ban.

First, all but one of the babies on the chart had post-mortems confirming natural causes. The only one who didn’t (Baby E) didn’t have a post-mortem because the consultant on duty thought the cause was so obviously necrotising enterocolitis (a fairly common cause of death in premature babies). Nobody besides the original pathologists examined the actual bodies, including the prosecution witnesses. The original pathologists were not called to trial.

In other words, the chart was actually populated with cases where the medical evidence pointed to natural explanations.

Second, there were events initially treated as suspicious which later dropped off the list once it was found that Lucy Letby wasn’t present. This shows that suspicion was tied to her presence. Cases were selected into the chart because she was there, and excluded when she wasn’t. That makes the chart circular: suspicious if she’s present, ignored if she isn’t.

Third, the whole process began with a misunderstanding of statistics. Dr Brearey looked at what was actually a random statistical cluster of deaths, noticed that the number of deaths was higher than usual, though actually it was statistically unremarkable, and concluded that someone must be harming babies. He then mistakenly thought “nice Lucy” was the common factor on duty.

That early wrong-headed supposition drove the entire investigation. The chart is just a visualisation of that original error.

Fourth, the idea that leaving out other deaths “doesn’t matter” misses the point. If other babies died or collapsed in similar ways when Letby wasn’t there, including them would have broken the illusion that she was uniquely present every time something bad happened. By only plotting the hand-picked newly “suspicious” cases, the chart created the appearance of a pattern that wasn’t there in the wider dataset.

Fifth, the chart also only plotted nurses, not doctors or consultants or anybody else who might have been present. Then, even worse, the police asked the actual doctors making the accusations against Letby to help them investigate the case. Are nurses uniquely capable of murder?

Finally, the prosecution did not use the chart neutrally. In his opening, Johnson KC told the jury that “by a process of elimination” it showed Letby was the killer. This is turning correlation into causation and it was a key part of the prosecution’s case at trial.

The chart was built on natural deaths, which were reframed as suspicious, curated to include only cases where she was present, and presented as if that proved her guilt. In reality, it just recycles Dr Brearey’s original mistake in confusing unremarkable small number stats with evidence and adding to that his mistaken assumption that LL’s presence was therefore suspicious.

Wasn’t she there for 12 out of the 13 deaths on the unit during that period anyway, so would it really have helped her if they’d shown that on the chart too?

It might have got people thinking about why some deaths were suspicious because she was there but others weren't?

This case is apparently a terrible indictment of the skill of presumably reputable pathologists. I'm still baffled how they could allegedly get it wrong in so many cases without an investigation etc. But there's been little about that in the press, at least not that I've seen. The breakdown of protocol in this case is astonishing all round.

It would be completely unacceptable for the private medical histories of babies that didn’t feature on the indictment to be thrashed out in court.

I'm not suggesting that though, I'm suggesting the numbers should have been fairly presented.

It might have got people thinking about why some deaths were suspicious because she was there but others weren't?

It wasn’t the prosecution’s case though that some deaths were suspicious because she was there.

It was their case that she had committed deliberate harm to each of the babies referred to on the indictment, and that this was supported by the large amounts of evidence that they put before the jury as they worked through each one.

If the prosecution had presented the rota purely as a neutral diary of who was on duty, then perhaps you could argue that. But the chart wasn’t used neutrally. It was presented as if it proved a pattern. That makes what was left out highly relevant.

If other babies had similar collapses or deaths when she wasn’t on duty, which we now know to be true, then the chart’s entire rhetorical force would have evaporated. Leaving them out gave the jury a misleading impression that she was uniquely linked to harm.

It’s not about “thrashing out” anything (do you use that framing for everything mentioned in court btw?) it’s about the prosecution choosing to show only part of the dataset while asking the jury to draw broad conclusions from it. In statistics or science, that’s called selection bias.

By the way, where did you get your data that there were 13 deaths and that she was there for 12?

In your view how should they have done that in a way that was fair and without going into any detail about how the other babies died? Particularly given that she was actually there for 12 out of the 13 deaths anyway? Wouldn’t that also create a potential risk that jurors might start to wonder if she had also harmed the other babies but that there just hadn’t been enough evidence to charge her?