Lucy Letby bombshell as new memo from sole medical witness threatens to blow prosecution's case wide open.

[MissJoGrant]

This seems like a pretty major development.

www.lbc.co.uk/crime/lucy-letby-bombshell-new-memo-sole-medical-witness-nurse-conviction/

I bet you most people who have served on juries couldn’t tell you the threshold for a statistically valid finding compared to one which cannot be relied upon and could have occurred by random chance. So in the absence of training and a test to ensure jury members understand what a statistically valid finding is, how can they possibly be in a position to judge what constitutes “beyond reasonable doubt” (a well-established statistical threshold to prove anything).

Moreover, as I’ve said, unless there are checks and balances on the evidence presented and how it is presented to the jury, even if they had an appropriate mathematical understanding of statistical validity vs coincidence they couldn’t make a valid judgement if the evidence presented is itself misleading, hence the need for far more scrutiny of what is admitted as evidence and juries are presented with.

At the moment you have a double whammy in some cases of misleading evidence presented to a jury of people who aren’t equipped to assess the implications of evidence even if it was valid, and certainly aren’t in a position to be likely to identify when it isn’t.

But she wasn’t accused of harming the other babies, so why on earth would they be included? The prosecution’s case was never that it had to be her because she was on shift, instead they used the chart to show that she had the opportunity to commit the crimes that they were accusing her of.

Notably, not one of whom was an expert in neonatology, and the evidence that a crime had actually taken place at all almost entirely resting on the interpretation of an “expert witness” who knew nothing about this area of medicine and grossly misinterpreted described symptoms, misunderstanding the physical effects described in scientific studies he tried to reference to support his assertions which have been shown to have no merit and no evidence at all to support them.

Complex cases involving medicine or finance in particular pose higher risks when you have a jury who is highly unlikely to genuinely understand a lot of the evidence so it is imperative that a proper panel of experts in highly specialised areas make sure what the jury is told is at least scientifically and statistically valid. I can’t see how anybody who wants justice (in terms of convicting the guilty as well as not convicting the innocent!) could argue otherwise.

Because you can’t select a chart of only the cases where somebody was present, and then say “look, here we’ve ticked which people were present for each case and as you can see this particular person was present for all of them…” 🤦🏻‍♀️

Of course they were, because you have restricted the population data to only the ones that occurred when they were there, so whether that had been 3 or 5 or 50 she’d obviously be shown to be there for 100% of them if you only select those ones to show.

It therefore gives the impression that every time a baby dies she is on shift, because they deliberately excluded all of the others. It’s incredibly misleading and will make people think “what other reason cpuld there be?”.

If, instead, they’d shown the full list of deaths - many of which she was not present for, in fact her being on shift for less than half of the deaths in the relevant period - clearly that gives the jury an extremely different impression when they have access to all of the information, as they should. Instead, they might think “what is going on at this hospital for there to be so many deaths no matter who is on shift?”.

Given that it’s now been shown there is no evidence of these particular babies having been killed deliberately, that argument makes no sense at all. These cases were no more suspicious than the others in terms of any evidence of deliberate harm. They were, it seems, cases where with proper care these babies should have had a higher chance of survival than for all of them to die, which indicates poor care in the hospital in general rather than deliberate malpractice.

Why were these specific cases chosen to be the ones they’d accuse her of? Because she was on shift. Therefore showing a graph that shows she was on shift when they took place as though it constitutes some kind of revelation or evidence of anything is clearly nonsensical because by definition it would be the case she was on shift for all of them because that’s how they selected the data. If it didn’t constitute evidence then why was it allowed to be presented at all? Obviously they wouldn’t have been accusing her of crimes when she wasn’t even at the hospital. 🤣

I don't think you're really that naive.

I suppose it could be an honest mistake from Dr Jayaram. Though it doesn't look too good.

I don’t know if Lucy Letby did it, but she will be released at some point, I guarantee it. There is just too much doubt for this to be a reliable conviction.

I could keep track for a year of all the times I am sad and all the times it rains. Then, if I filter that data for all the times I am sad and it’s raining, guess what? It’ll show me being sad 100% of the times that it rains within that data set made up only of times when I am sad and it’s raining.

It won’t show any of the times when I was sad and it was sunny (or, by analogy, a baby died but Letby wasn’t on shift) or any of the times when I was happy and it rained (or, by analogy, Letby was on shift but no baby died).

Someone who hadn’t been told how this data about my emotions/ rains had been selected and with minimal understanding of maths and statistics might look at that chart and see it says that I’m sad 100% of the time when it’s raining (of the selected data in the chart which deliberately has been designed only to include these cases of correlation). They might also, therefore, falsely infer a causal relationship between the two, thinking it “too much of a coincidence” for there not to be (the correlation/ causation fallacy, which happens frequently even when people are presented with a full data set). And in this case, they were not.

That is why it matters. It’s incredibly misleading.

The very fact that poster asked that question (if we are to believe it is genuine) demonstrates exactly what I’ve been saying about the importance of having statisticians checking how evidence is presented, and ensuring that jurors have sufficient ability and can pass a basic test in mathematical and statistical competence so that they can demonstrate they are able to appropriately judge the evidence presented and reach valid conclusions on whether the threshold of “beyond reasonable doubt” has been proved.

it is ridiculous to say that Dewi Evans knew nothing about neonatology. He set up the first neonatal service in Wales and was a clinical director of paediatrics and neonatology. Some may take issue with the depth of his experience, or how current it is, but it’s just plain incorrect (not to mention hyperbolic) to say he knew nothing about the subject matter. Also, his findings were peer reviewed by Sandie Bohin who is a practicing neonatologist.

Well maybe if the entire prosecution case during the 9 month trial was ‘she was there when these babies died, therefore she had to have killed them’. But obviously that wasn’t the entire (or in fact any of the) prosecution case was it? Detailed evidence was provided for each of the babies on the indictment, over many weeks, to make a case for how the prosecution believed she had harmed them. They did not include other babies who had died as the prosecution were satisfied that they had died of natural causes and therefore were not accusing LL of killing them.

I wouldn't say he knew nothing about neonatology, but he is not a specialist in neonatology, and he never practised using the equipment in use on the ward in Chester. Understanding this equipment would be necessary to understand many of the errors and misinterpretations identified in the international expert panel's report. Evans hasn't responded to their report, except to say he has no respect for them and they are wrong, but it will be hard to believe that he has the expertise to mount an informed objection.

this is like how people used to trust priests in the Catholic Church! Yet they were abusing heaps of kids and abusing their position of tower

I notice how these posts attempt to sidetrack the current discussion into already well-rehearsed re-hashes of disagreements over the competence and expertise of various medical professionals yet - conspicuously - none of these latest responses to me mention anything about the statistical and mathematical problems with the evidence presented in this trial and others; or the consequences of the provision of data to juries in misleading formats inviting misinterpetation; or the competence of jurors selected at random to draw meaningful and valid conclusions - even if the evidence shown to them was properly validated and not misleadingly presented - regarding whether the probabalistic threshold of “beyond meaningful doubt” has been proved.

For juries’ conclusions to be valid their conclusions must be determined, as for all proof, by a mathematic threshold, yet it is clear that often those providing the evidence either don’t understand statistics and probability or try to manipulate it, and most juries don’t understand how to assess what they are looking at. They simply can’t because the majority of the population don’t have the statistical skills to make such a judgement. This is a HUGE problem given the entire purpose of a trial is to establish whether a probabalistic threshold has been met.

How strange that nobody responding wishes to engage with this point, which is the reason I posted on the thread and it is essential to rectify if any of us are to feel confident of a fair trial if ever accused of anything.

So if the prosecution had shown the jury the full list of deaths together with who was on shift (and worth noting LL was actually on shift for some of them anyway) but in doing so said we’re not suggesting she killed these other babies, then presumably that would have been okay with you? I honestly can’t see how that would have made any meaningful difference 🤷‍♀️ As per my previous post, the chart was used to show opportunity.

Then they should have showed all of the data with extra columns containing that information, not kept the jury in the dark about the facts.

Because it is predictated on the assumptions that 1) the “expert witnesses” - such as they were - were correct in deciding that these deaths were more suspicious than the others. The jury had not been presented with any evidence to support this and, indeed, it seems that it does not exist; 2) it is unclear if the decision to prosecute Letby with these cases was made because these ones correlated to her shift times or because a) doctors had already made complaints about her/ were trying to cover their backs, or b) other cases would have been put forward that were equally as “unexpected” if she’d happened to be on shift for those. I.e. the jury would need to see the whole data set and hear evidence about how each baby died, how likely it should have been to survive, in order to agree with the prosecution’s case that these ones selected because they correlated to her shifts were on average more likely to die than the others and therefore her happening to be on shift for these cases was relevant. This is part of determining if there was a crime at all and if so whether she is guilty of it. It’s not a fair trial if the evidence presented already assumes that only the cases when she was present are suspicious and only evidence on those is presented implying that all questionable deaths involved her being at work, when there is in fact no evidence to support this. This made the prosecution’s case not something that was tested in court but a pre-assumed assumption of the data presented to the jury without them being given the full data so that they could assess other possibilities e.g. general failure of care resulting in higher deaths due to poor general standards in the hospital (and, it seems, reliance on poor “expert witnesses” leading to partial data showing only part of the picture being provided to the jurors).

It is quite obvious that presenting data in such a way as to keep the jury in the dark about a large proportion of pertinent data, show them a 100% correlation because all other data is excluded, and therefore make their perception be far more likely to be that the prosecution’s case is plausible and there’s no other plausible explanation when in fact there most certainly are other possible and plausible explanations, is clearly a misrepresentation of evidence that is likely to influence a jury when determining whether the threshold of “beyond reasonable doubt” has been met.

In short, the evidence you present cannot be collated on the presumption that the case you are trying to prove is valid and exclude all evidence that contradicts it or may indicate otherwise.

That’s not testing a case in court, that’s a show trial.

The facts of the prosecution’s case against LL were set out in very great detail to the jury over many months. And the defence had the opportunity to challenge/cross-examine these facts throughout.

Her defence was clearly very poor or they’d have challenged this. Prosecutions get away with gross misrepresentations by incompetence or simply because they can. See my first post on the thread. My entire point is that the system needs reforming to ensure that statistically misleading or scientifically dubious evidence is not presented in court and that jurors are properly screened to ensure they are competent to assess it.

All I can really say is that - if I am ever accused of anything - I really hope you are not part of my jury @rubbishatballet

She was represented by one of the top criminal barristers in the country, who is still instructed to represent her.

The defence wasn't allowed to engage with the construction of the case as part of its line of questioning - the judge prohibited this, as he did references to deaths not part of the prosecution case. So it would have been extremely difficult for the defence to counteract the prosecution's statistical inferences.

Unfortunately, we have since found that quite a lot of “facts” were merely assumptions and inexpert opinions.

What are you talking about? The prosecution has to disclose to the defence any evidence that might support the defence and/or undermine the prosecution, but it can present in court the evidence that best supports its case (and the defence can challenge this).