Lucy Letby - bad inference

Discussions about serious topics, for serious people
IvanV
Stummy Beige
Posts: 3151
Joined: Mon May 17, 2021 11:12 am

Re: Lucy Letby - bad inference

Post by IvanV » Thu Jul 18, 2024 12:17 pm

dyqik wrote:
Thu Jul 18, 2024 10:51 am
It's also plausible that she was fully or partially responsible* for many or all of the deaths, but that the evidence presented was massively flawed, and insufficient to actually show that beyond reasonable doubt.

*Responsibility doesn't have to be complete, and can be shared. It also doesn't always rise to the level of criminal culpability.
This is consistent with what was being argued by Rebecca Watson in the link above. There seems to be a clear indication that the evidence was substantially flawed. But what you might conclude from better evidence is unclear.

To invent my own conspiracy theory, maybe the defence's failure to respond adequately to the flawed evidence was deliberate. If you are convicted on flawed evidence, you may have a chance of a determination of an unsafe conviction later. If you force out the best evidence, and are convicted on that, you are stuffed.

IvanV
Stummy Beige
Posts: 3151
Joined: Mon May 17, 2021 11:12 am

Re: Lucy Letby - bad inference

Post by IvanV » Thu Jul 18, 2024 12:35 pm

And while we are at it, there are calls today for the sacking of the chair of the Criminal Cases Review Commission, BBC, following a review of the case of Andy Malkinson, who spent 17 years in prison despite very dubious evidence against him, and the police finding new evidence that excluded him 3 years after the incident.

Even when there is proof of innocence, as in this case, still it took many years and several adverse determinations to get the case overturned. The lawyer who carried out the review, Chris Henley KC, said:
“This case demonstrates a deep-seated, system-wide cultural reluctance, which starts right at the top in the Court of Appeal, to acknowledge our criminal justice system will on occasion make mistakes.

“It is not by any standard a success, or a demonstration that things are working properly, that Mr Malkinson had to wait 20 years to be exonerated.”
Most cases will not have that proof of innocence, but are still terrible convictions. And it is not just attitude, it is also funding. The CRCC has only a small fraction of the funding needed to investigate the many deeply worrying convictions.

And then there's the issue of compensation for the wrongly convicted. Our laws now make that almost impossible to obtain.

User avatar
snoozeofreason
Snowbonk
Posts: 542
Joined: Fri Nov 15, 2019 1:22 pm

Re: Lucy Letby - bad inference

Post by snoozeofreason » Thu Jul 18, 2024 1:04 pm

bob sterman wrote:
Thu Jul 18, 2024 6:14 am
One possibility worth considering - it could be the case that Letby was in fact directly responsible for 1-2 deaths (either through negligence or deliberate acts). However, the prosecution may have then used weak evidence to blame many more deaths on her - deaths that were really due to poor staffing and failings in the unit.

This would be be consistent with some of the diary entries - but the generally weak evidence presented in relation to many of the deaths.
A lot was made of Letby's writings, but the conclusions that can be drawn are not particularly clear. Supposedly incriminating entries that are often quoted come from a post-it note (below) on which she had jotted all sorts of random things, in random directions. It's not clear whether she was writing down things that she thought were true, or things that had been said about her. It does include the sentence "I killed them on purpose because I am not good enough," which you could take as evidence of guilt if you were predisposed to think her guilty. But it also includes the sentence "I haven't done anything wrong," which you could take as evidence of innocence if you were predisposed to think of her as innocent. A more reasonable conclusion, IMO, is that it's not really evidence of anything, just stream of consciousness jottings. There were apparently some other jottings, but they sound equally inconclusive. More details in BBC article here.


notefromrvlucyletby.jpg
notefromrvlucyletby.jpg (196.77 KiB) Viewed 2848 times
In six days the Lord made the heavens and the earth, the sea, and all that is in them. The human body was knocked up pretty late on the Friday afternoon, with a deadline looming. How well do you expect it to work?

User avatar
Brightonian
Dorkwood
Posts: 1561
Joined: Mon Nov 11, 2019 3:16 pm
Location: Usually UK, often France and Ireland

Re: Lucy Letby - bad inference

Post by Brightonian » Mon Jul 29, 2024 2:51 pm

Phil Hammond from Private Eye's seeking experts re embolisms and insulun: https://x.com/drphilhammond/status/1817844095416447317

Allo V Psycho
Catbabel
Posts: 796
Joined: Sat Nov 16, 2019 8:18 am

Re: Lucy Letby - bad inference

Post by Allo V Psycho » Thu Jan 16, 2025 5:09 pm

David Speigelhalter on the statistics
https://www.bbc.com/news/articles/cpdxn4x5477o
Death rate not a marked outlier.
Honestly, I'm not posted this as evidence of her innocence. It's a small piece of evidence in an impossibly obscure situation. I just don't know on this one, and it worries me quite a lot.

User avatar
sTeamTraen
Stummy Beige
Posts: 2589
Joined: Mon Nov 11, 2019 4:24 pm
Location: Palma de Mallorca, Spain

Re: Lucy Letby - bad inference

Post by sTeamTraen » Thu Jan 16, 2025 8:10 pm

Allo V Psycho wrote:
Thu Jan 16, 2025 5:09 pm
It's a small piece of evidence in an impossibly obscure situation.
I humbly submit that we should not be putting people in prison for their guilt "beyond a reasonable doubt" when the situation is impossibly obscure (which I agree it is).
Something something hammer something something nail

User avatar
TopBadger
Catbabel
Posts: 887
Joined: Mon Nov 11, 2019 6:33 pm
Location: Halfway up

Re: Lucy Letby - bad inference

Post by TopBadger » Thu Jan 16, 2025 8:26 pm

Didn't some book called Bad Science (don't know where I remember that from) have a chapter on a similar conviction in similar circumstances that was statistically unsafe and later overturned?
You can't polish a turd...
unless its Lion or Osterich poo... http://dsc.discovery.com/videos/mythbus ... -turd.html

User avatar
Sciolus
Dorkwood
Posts: 1424
Joined: Mon Nov 11, 2019 6:42 pm

Re: Lucy Letby - bad inference

Post by Sciolus » Thu Jan 16, 2025 10:38 pm

Dewi Evans, the lead expert witness for the prosecution, has been giving a number of press interviews recently. Presumably he thinks they show how terribly clever he is, but to me, every quote of his I read makes me shout "Charlatan!" Exasperatingly, I've just thrown out my Private Eyes which carried a load of jaw-dropping quotes, but this article in the Graun gives you a flavour of his lack of scientific insight. Perhaps my favourite of his claims is that he was convinced that deliberate harm had occurred within 10 minutes of seeing the notes of one of the children, something missed by all the medics, pathologists and coroners who actually examined the body, as well as another neonatal pathologist who looked at the notes blind.

User avatar
Woodchopper
Princess POW
Posts: 7419
Joined: Sat Oct 12, 2019 9:05 am

Re: Lucy Letby - bad inference

Post by Woodchopper » Fri Jan 17, 2025 6:11 am

Allo V Psycho wrote:
Thu Jan 16, 2025 5:09 pm
David Speigelhalter on the statistics
https://www.bbc.com/news/articles/cpdxn4x5477o
Death rate not a marked outlier.
Honestly, I'm not posted this as evidence of her innocence. It's a small piece of evidence in an impossibly obscure situation. I just don't know on this one, and it worries me quite a lot.
Sir David said its mortality rate in 2015 and 2016 was only 10% higher than the average for neonatal units with similar birth rates.
Yes, that doesn’t give support to either guilt or innocence.

User avatar
Woodchopper
Princess POW
Posts: 7419
Joined: Sat Oct 12, 2019 9:05 am

Re: Lucy Letby - bad inference

Post by Woodchopper » Fri Jan 17, 2025 6:41 am

TopBadger wrote:
Thu Jan 16, 2025 8:26 pm
Didn't some book called Bad Science (don't know where I remember that from) have a chapter on a similar conviction in similar circumstances that was statistically unsafe and later overturned?
Pretty sure that the book Bad Science didn’t cover that (it was homeopaths etc). Ben Goldacre might have written about the de Berk case elsewhere.

User avatar
TopBadger
Catbabel
Posts: 887
Joined: Mon Nov 11, 2019 6:33 pm
Location: Halfway up

Re: Lucy Letby - bad inference

Post by TopBadger » Fri Jan 17, 2025 10:49 am

Woodchopper wrote:
Fri Jan 17, 2025 6:41 am
TopBadger wrote:
Thu Jan 16, 2025 8:26 pm
Didn't some book called Bad Science (don't know where I remember that from) have a chapter on a similar conviction in similar circumstances that was statistically unsafe and later overturned?
Pretty sure that the book Bad Science didn’t cover that (it was homeopaths etc). Ben Goldacre might have written about the de Berk case elsewhere.
It's chapter 14 in my copy...
You can't polish a turd...
unless its Lion or Osterich poo... http://dsc.discovery.com/videos/mythbus ... -turd.html

User avatar
snoozeofreason
Snowbonk
Posts: 542
Joined: Fri Nov 15, 2019 1:22 pm

Re: Lucy Letby - bad inference

Post by snoozeofreason » Fri Jan 17, 2025 5:08 pm

TopBadger wrote:
Fri Jan 17, 2025 10:49 am
Woodchopper wrote:
Fri Jan 17, 2025 6:41 am
TopBadger wrote:
Thu Jan 16, 2025 8:26 pm
Didn't some book called Bad Science (don't know where I remember that from) have a chapter on a similar conviction in similar circumstances that was statistically unsafe and later overturned?
Pretty sure that the book Bad Science didn’t cover that (it was homeopaths etc). Ben Goldacre might have written about the de Berk case elsewhere.
It's chapter 14 in my copy...
He also commented about it online
https://www.badscience.net/2010/04/lucia-de-berk-a-martyr-to-stupidity/
In six days the Lord made the heavens and the earth, the sea, and all that is in them. The human body was knocked up pretty late on the Friday afternoon, with a deadline looming. How well do you expect it to work?

User avatar
Gfamily
Light of Blast
Posts: 5683
Joined: Mon Nov 11, 2019 1:00 pm
Location: NW England

Re: Lucy Letby - bad inference

Post by Gfamily » Sun Jan 19, 2025 11:18 am

Sciolus wrote:
Thu Jan 16, 2025 10:38 pm
!Exasperatingly, I've just thrown out my Private Eyes which carried a load of jaw-dropping quotes...
All MD's columns on the Letby case are available here:
https://www.private-eye.co.uk/special-r ... lucy-letby
My avatar was a scientific result that was later found to be 'mistaken' - I rarely claim to be 100% correct
ETA 5/8/20: I've been advised that the result was correct, it was the initial interpretation that needed to be withdrawn
Meta? I'd say so!

User avatar
Woodchopper
Princess POW
Posts: 7419
Joined: Sat Oct 12, 2019 9:05 am

Re: Lucy Letby - bad inference

Post by Woodchopper » Sun Jan 19, 2025 12:53 pm

TopBadger wrote:
Fri Jan 17, 2025 10:49 am
Woodchopper wrote:
Fri Jan 17, 2025 6:41 am
TopBadger wrote:
Thu Jan 16, 2025 8:26 pm
Didn't some book called Bad Science (don't know where I remember that from) have a chapter on a similar conviction in similar circumstances that was statistically unsafe and later overturned?
Pretty sure that the book Bad Science didn’t cover that (it was homeopaths etc). Ben Goldacre might have written about the de Berk case elsewhere.
It's chapter 14 in my copy...
Of course, my memory fails me again.

User avatar
Sciolus
Dorkwood
Posts: 1424
Joined: Mon Nov 11, 2019 6:42 pm

Re: Lucy Letby - bad inference

Post by Sciolus » Mon Jan 20, 2025 9:40 pm

Gfamily wrote:
Sun Jan 19, 2025 11:18 am
All MD's columns on the Letby case are available here:
https://www.private-eye.co.uk/special-r ... lucy-letby
Thanks. Having reread them, there are fewer direct quotes of Evans than I remember, but they're still pretty astonishing from an expert scientist. Block quotes are from the article, so "I" is MD aka Hammond, unless there are quote marks.
MD asked Dr Evans two questions: 1. What other causes of death or deterioration did you consider alongside deliberate harm? 2. How did
you exclude them?
His responses were:
1. “In relation to the seven deaths, it was possible firstly to exclude natural causes such as haemorrhage, infection or some congenital
problem. The unexpected collapses were very unusual – and consistent with air embolus, air injected into the bloodstream. This was the most
likely cause before the radiology evidence was flagged up by Owen Arthurs (air in the great vessels) and the peculiar skin discolouration
noted by the local medics. These findings were not essential to the diagnosis but added to the clinical presentation already noted.
“Having made the diagnosis (the injection of air), it could have happened accidentally or intentionally. If it was accidental, the cause would have been easily spotted, as it’s normal for two nurses (or a doctor and nurse) to be present when babies are given fluids or drugs intravenously. Any deterioration would have occurred there and then. There were no reported events of this nature. The collapses occurred when the infants were in Letby’s sole care.”
2. “I was able to exclude other causes, such as the ones noted above, because there were no other causes. Sorry if that sounds rather odd. But
that’s clinical practice for you.”
Dr Evans did not think there needed to be statistical experts at the trial because “stats weren’t relevant” to the case, and said: “The
Guardian letter from ‘24 experts’ – statisticians and neonatologists who have had nothing to do with the case – asking that the Thirlwall inquiry
is postponed is professional hubris of the worst order.”
I have sent Dr Jayaram a list of plausible alternative causes for all 17 cases, as requested, from tertiary centre specialists who have knowledge of the case notes. He has not responded to it.
In contrast, Dr Evans responded very promptly: “I hope this doesn’t sound partisan, but the more information I receive challenging the prosecution case, the greater the evidence in their favour… Your documents are the first I’ve received that contain ‘alternative’ explanations. I’ve had very little difficulty in challenging them so far.”
As he wrote to me: “In 35 years (1988–2023), I have never lost a murder, manslaughter or serious abuse case other than one when acting for the defence… Losing my one case still rankles.”
The articles raise a lot more questions that the legal system doesn't seem interested in answering.

User avatar
El Pollo Diablo
Stummy Beige
Posts: 3644
Joined: Wed Sep 25, 2019 4:41 pm
Location: Your face

Re: Lucy Letby - bad inference

Post by El Pollo Diablo » Tue Jan 21, 2025 9:24 am

Yes, thanks for that link. I've now read seven of them and I must say it's done a very good job of explaining some of the statistical problems with the prosecution, to the extent that I can absolutely now see that Letby might be innocent.

I think, though, this in turn goes back to a central problem for Letby - the case against her really does look to the average person like an open and shut one. The issues with the case are subtle and difficult to understand. It's going to be very, very hard for her to come back from where she is now, if she ever does at all.
If truth is many-sided, mendacity is many-tongued

User avatar
El Pollo Diablo
Stummy Beige
Posts: 3644
Joined: Wed Sep 25, 2019 4:41 pm
Location: Your face

Re: Lucy Letby - bad inference

Post by El Pollo Diablo » Tue Jan 21, 2025 9:40 am

David Davis put forward an adjournment debate a couple of weeks ago on Letby, and this is part of his speech, and I think is very pertinent:
There is a case in justice for a retrial in my view, but there is a problem—I would say it is a technical problem. One of the problems we face is that much of the evidence was available at the time. What I have described is an expert analysis of the case notes that were there at the time, but it was simply not presented to the jury. That means that the Court of Appeal can dismiss it, basically saying that the defence should have presented it at the initial trial. It is in essence saying, “If your defence team weren’t good enough to present this evidence, hard luck, you stay banged up for life.” That may be judicially convenient, but it is not justice. This has been a historic problem in Britain, delaying the resolution of a number of miscarriage of justice cases.

Secondly, the so-called expert evidence in this case largely amounted to putting together theories suiting the prosecution case, from the bogus statistical arguments through to the wrong diagnoses I have just detailed. Again, this problem is not new. In 2011 the Law Commission made a number of recommendations precisely on the handling of expert evidence. There is even a section in it entitled, “Inferring murder from unexplained infant deaths.” Those recommendations have not made it into law nearly 15 years later, and it is past time that was put right. Perhaps most worrying of all in this case is that it is just the latest example, more than a decade after the Mid Staffs scandal, of the national health service’s inability to analyse its own failings. This matters because until we learn these lessons, there will continue to be unnecessary deaths in hospitals up and down the country.

My central argument today, which comes back to what my right hon. Friend asked me, is about what to do about a trial which, in my view, is a clear miscarriage of justice by a judicial system that could not manage admittedly difficult statistical and medical scientific evidence. The only body available to correct this today is the Criminal Cases Review Commission. In my view, it should look at all the new diagnoses when they come out, and if necessary consult the leading neonatal and statistical authorities in the land—the most expert people, who are much more equipped to give proper assessment than the experts who were employed by the police at the time. As a result, in my view it should order a retrial, and it should do it quickly.

This is significant because the CCRC has recently been criticised—Members may have read about it in the Andrew Malkinson case, where there was effectively a 17-year delay in releasing him from prison. DNA evidence proving him innocent of the rape he was convicted of was known four years after his conviction, yet it took a further 13 years to correct it. That cannot happen again. We cannot repeat that. If, as I believe it will, a retrial clears Lucy Letby, she should be released in her thirties, not in her fifties.
If truth is many-sided, mendacity is many-tongued

IvanV
Stummy Beige
Posts: 3151
Joined: Mon May 17, 2021 11:12 am

Re: Lucy Letby - bad inference

Post by IvanV » Tue Jan 21, 2025 3:36 pm

El Pollo Diablo wrote:
Tue Jan 21, 2025 9:24 am
I think, though, this in turn goes back to a central problem for Letby - the case against her really does look to the average person like an open and shut one. The issues with the case are subtle and difficult to understand. It's going to be very, very hard for her to come back from where she is now, if she ever does at all.
One of the big issues, at least as MD in Private Eye sees it, is that some doctor repeatedly diagnosed the cause of death as air embolism, but with few or none of the usual evidential requirements for that being present. That doesn't seem to be a difficult point for the general public to grasp. Statistics only play a minor part in this problem, what it is mainly about is what facts you need to make a particular cause of death diagnosis.

Meanwhile we are having a huge judicial enquiry on how to make sure we detect NHS serial killers quicker, when the real issue may well be something quite different. Indeed regardless of Letby's guilt, the following are longstanding issues that have been repeatedly identified: the NHS covering up its incompetence, shooting whistle-blowers, using their own version of the bad apple argument to explain poor outcomes, and the use of rent-a-convenient-diagnosis forensic "experts".

User avatar
snoozeofreason
Snowbonk
Posts: 542
Joined: Fri Nov 15, 2019 1:22 pm

Re: Lucy Letby - bad inference

Post by snoozeofreason » Tue Jan 21, 2025 8:59 pm

The case presented at Letby's original trial obviously was convincing to the average people selected for the jury, and to lots of others besides. But it's worth noting that it would be hard to present the same evidence at a retrial, if she ever got one. It would be difficult to replicate the famous spreadsheet, because Dewi Evans has said that he won't be testifying again, and no one else seems able to explain why the deaths and adverse events mentioned in the spreadsheet were included, when numerous other deaths and adverse events were not.

An expert witness would require considerable courage to recreate the arguments about injection of air, given the scepticism expressed by other experts.

It might be possible to rerun the insulin evidence. The prosecution could argue that, although the tests referred to in the original trial were flawed, they would still be accurate most of the time. But even if a jury accepted that, there would, in the absence of the other evidence, be a problem. The insulin tests, even if reliable, could only prove that someone administered insulin. In the absence of any other evidence pointing her way, there would be a good reason to think the someone in question could not have been Letby, because she was off shift when one of the supposedly contaminated bags was attached. That point was made to the original jury, but they might well have already been convinced that Letby was guilty on the other counts, and would then have felt it unlikely that there could be two serial killers on the loose. If the evidence on the other cases is less solid, that line of reasoning is less compelling.
In six days the Lord made the heavens and the earth, the sea, and all that is in them. The human body was knocked up pretty late on the Friday afternoon, with a deadline looming. How well do you expect it to work?

User avatar
JQH
After Pie
Posts: 2208
Joined: Mon Nov 11, 2019 3:30 pm
Location: Sar Flandan

Re: Lucy Letby - bad inference

Post by JQH » Wed Jan 22, 2025 3:43 pm

But it's worth noting that it would be hard to present the same evidence at a retrial, if she ever got one. It would be difficult to replicate the famous spreadsheet, because Dewi Evans has said that he won't be testifying again, and no one else seems able to explain why the deaths and adverse events mentioned in the spreadsheet were included, when numerous other deaths and adverse events were not.
He could be compelled by court order to testify.
And remember that if you botch the exit, the carnival of reaction may be coming to a town near you.

Fintan O'Toole

User avatar
snoozeofreason
Snowbonk
Posts: 542
Joined: Fri Nov 15, 2019 1:22 pm

Re: Lucy Letby - bad inference

Post by snoozeofreason » Thu Jan 23, 2025 8:03 am

JQH wrote:
Wed Jan 22, 2025 3:43 pm
But it's worth noting that it would be hard to present the same evidence at a retrial, if she ever got one. It would be difficult to replicate the famous spreadsheet, because Dewi Evans has said that he won't be testifying again, and no one else seems able to explain why the deaths and adverse events mentioned in the spreadsheet were included, when numerous other deaths and adverse events were not.
He could be compelled by court order to testify.
IANAL, but my understanding is that no one can be compelled to give expert testimony at a trial. Evans could, in theory, be compelled to give evidence on matters of fact, but it's unlikely that this would happen in practice. Any factual evidence that Evans could give could equally well be given by someone else, and the facts that he relied on aren't really in dispute. What is disputed is the inferences that he drew from them.

The situation is summed up in the Courts and Tribunal Service's advice to potential witnesses.
Can I refuse to be a witness?

Yes, if you are asked to be an expert witness. You must decide whether you can spare the
time from your work or business to prepare a report and, perhaps, go to a court hearing.

If you are asked to be a witness of fact, you can also refuse. But the party who asks you can
take steps to make you come to court to act as their witness.
...
In six days the Lord made the heavens and the earth, the sea, and all that is in them. The human body was knocked up pretty late on the Friday afternoon, with a deadline looming. How well do you expect it to work?

Post Reply