Main Edition
US
FR
May 14, 2024 - 1:30pm
Conservative MP David Davis today used Parliamentary privilege to raise an urgent question about yesterday’s New Yorker essay challenging the Lucy Letby murder verdict. The 13,000-word piece is available to read on the magazine’s US site but has not been made available on the UK version for legal reasons. It is visible for British users of the New Yorker app and subscribers to the print magazine.
Davis was speaking this afternoon during a session of Justice questions in the House of Commons. Addressing Justice Secretary Alex Chalk, the backbencher stated that the New Yorker article “raised enormous concerns about both the logic and competence of the statistical evidence that was a central part of that trial”. Davis went on to say that the piece “was blocked from publication on the UK internet, I understand, because of a court order”. While the MP acknowledged that the order was “well-intended”, he suggested that it was “in defiance of open justice”, and asked Chalk whether he would “look into the matter and report back to the House [of Commons]”.
The Justice Secretary responded that “court orders must be obeyed” and can be “displaced by someone applying to the court for them to be removed”. He added that the process would “need to take place within the normal course of events”, and that “juries’ verdicts must be respected.” Chalk also said: “if there are grounds for an appeal, that should take place in the normal way.”
Letby, a British nurse, was sentenced to life imprisonment last year for the murder of seven babies in her care. She pled not guilty to the total of 18 charges against her, and now faces a retrial next month on a single charge on which the jury did not reach a verdict in 2023.
Join the discussion
Join like minded readers that support our journalism by becoming a paid subscriber
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.
SubscribeHere’s the article you can read in the UK on the flawed use of statistics in the Letby trial, published last September:
https://www.scienceontrial.com/post/shifting-the-data
This doesn’t mean that Letby is innocent, but she should, if guilty, have been convicted on the other evidence presented at her trial, not this.
Very interesting article, thank you. The analysis of the statistics presented at trial makes a compelling argument that at least that specific part of the evidence was flawed and should at least have been challenged.
I can offer no real insight into this case, only a vague sense of unease I’ve had all along that things don’t quite add up.
This is a most important article and highlights the abuse of statistics or ignorance of them by medical experts in criminal trials. The case of Sally Clark condemned by the faulty statistics of Sir Roy Meadow should have provided a red flag against the introduction of any statistical evidence in trials that has not been throughly reviewed by an expert statistician. This is an area that common sense can be lead astray.
I add a useful article on the problem of the failure of those involved in Criminal trials to understand the limitation of statistics here:
https://understandinguncertainty.org/node/545
Unfortunately there is great emotional pressure on those conducting investigations into the death of infants to find and convict a culprit rather than admit that there is insufficient evidence to be sure. The manner in which the distinguished Dr Wayne Squire was treated for challenging the dubious certainties of the triad of symptoms supposedly associated with shaken baby deaths is a stark reminder of how theories can be policed for the purpose of bolstering convictions. True science can be abandoned for the purpose of supporting an unreal certainty.
It appears that very few of our institutions understand the distinction to be made between correlation and causation. Moreover, it seems that some might well be accused of being bad actors after the manner in which data may well have been cherry-picked and presented as randomly selected data. I can’t be the only person, either, who thinks there may be some uncomfortable truths lurking behind the just published report by the Health and Social Care Committee on the parlous state of maternity care in our country; the facts of which have been plain to many grieving parents for a decade or more. One suspects that the upcoming trial of Lucy Letby on the one unresolved death of a baby may well turn out to be rather more than Health Care professionals are expecting.
Thanks, very interesting. A nice example of the ‘prosecutors fallacy’ was found in the trial of OJ Simpson – only it was applied by the defence. There was evidence that OJ had beaten his wife, but the defence produced evidence that it was very rare for a man who beat his wife to go on to kill her. The jury proceeded to acquit.
The best way to deal with it that I have seen is to translate all probabilities to numbers : ‘out of so many X you find so many Y’. In the OJ Simpson case what is being said is that most women whose husband beat them are not murdered. Instead they die from cancer or heart disease or traffic accidents. But since Mrs Simpson did *not* die from cancer or heart disease, it is obvious why this is not relevant.
Roy Meadows statistical ‘evidence’ was so obviously idiotic, and yet crucial to the conviction. IIRC at the time it was not possible to challenge an expert witnesses statements. Perhaps that is still the case now? If so, that might explain why the defence didn’t challenge the statistics in Lucy Letby’s case.
Having said that, and again IIRC, there was additional evidence in her trial, unlike the case of Sally Clarke.
The other issue is the use of using what would individually be weak charges to reinforce each other to secure a conviction when had they been prosecuted individually they would hve been thrown out at an early stage.
I do not know if she did it or not but in reading accounts of the trail at the time y impression was that the evidence was flimsy and too circumstantial.
On a related point, why have we not seen any prosecutions in relation to South Staffs NHS Trust. Did not the enquiry come up with a figure of more than 800 unnecessary deaths
There are all sorts of ‘matters arising’ from the Letby trial, not least the actions of the senior management at the Countess of Chester Hospital.
It’s all gone very quiet, but I hope that no one is allowed to quietly wriggle off the hook.
We haven’t heard from the management yet. I feel sure they have different evidence from that of their accusers. They have a right to defend their actions which will be heard at the Public Inquiry.
OK, this is interesting. Thank God for parliamentary immunity.
It awful to think that we could be facing another miscarriage of justice through bad statistics.
Statistics is a valuable tool when we use it to direct our efforts but it rarely offers proof.
I hear that many statisticians think that more statistics should be taught at schools. I’ve started to feel sceptical about this (though it wouldn’t harm my career). A little knowledge of statistics seems to encourage people to believe that they understand the world.
I think the opposite. In the Sally Clarke case the ‘statistics’ used were idiotic. In fact, I wouldn’t even call them statistics. (In this case it seems less clear cut, and the stats were probably less significant in getting the conviction.)
Either way, a better understanding of stats in the general population would help people to spot their misuse and make such things less likely to happen.
Not so: the statistical diagram was central to the conviction, given that there was absolutely nothing else that connected Letby to any of the deaths – no forensic evidence, witness evidence or confession.
As the New Yorker article puts it: ‘It gave an impression of mathematical clarity and coherence, distracting from another possibility: that there had never been any crimes at all.’
If I had been on the jury at the Letby trial, I would probably have accepted the stats without any comment as statistics – along with all Maths beyond primary school level – is a closed book to me. It would simply not have crossed my mind to challenge the interpretation of a medic/scientist whom I would assume was providing correct information. I doubt if I’m on my own in that respect.
In something of such critical importance as a murder trial, stats should come with a ‘health’ warning attached.
Yes, a lot of very clever people haven’t a clue in regard to evaluating the validity of statistical evidence. Unfortunately this often includes many of the lawyers involved in such trials.
I certainly wasn’t privy to all the information that convicted Letby so I’m not going to pretend I know more than anybody else, but it wasn’t only these stats that convicted her.
There was also the fact that she was the only one on shift every time one if these unlikely deaths/emergencies occurred, as well as numerous rambling confessional diary entries that corresponded with the days of the deaths. Other nurses also raised concerns around her behaviour with the parents of those babies killed, and senior doctors had raised concerns as well
Billy, “the fact that she was the only one on shift every time one if these unlikely deaths/emergencies occurred” are the statistics that are in fact dubious because the babies who died in the same period when she wasn’t on duty weren’t included:
https://www.scienceontrial.com/post/shifting-the-data
As above, the point is not that she is innocent (of all or any of the deaths) but that this part of the evidence was flawed.
Maybe the other deaths were expected or more easily explained, whereas the ones under Letbys watch were much more out of the blue?
The figures used in your link are just as cherry picked as those they complain about in my opinion in an attempt to cast doubt.
Maybe they do have grounds to appeal, but this wasn’t the only evidence used to convict her but one of many. It’s a pet hate of mine with these supposed real crime articles where they’ll ignore the 9 pieces of solid evidence used to convict a person on trial while relentlessly highlighting the 1 part that in isolation wouldn’t be enough to secure a conviction.
Read this & the follow up articles on the subjecft:
https://lawhealthandtech.substack.com/p/ll-part-1-hospital-wastewater
No – the infamous diagram only included the deaths when she was present and then magically showed that she was the only one present at all of them. Bingo! The worst miscarriage of justice in English history turned out to be really easy to manufacture.
No! You include ALL the infant deaths! You are simply reiterating the terribly flawed reasoning of the trial. There was no “solid evidence” that Lucy Letby killed any babies.
You seem to be almost implying that abysmal miscarriages of justice have not in fact occurred, on the grounds that there obviously must be loads of such “solid” evidence. But a prosecution case in a very emotive case combined with strongly put but bogus reasoning can easily convince a jury, and has done in many occasions.
Read the article…….
Well done to David Davis – just about the only MP in the Commons with anP osted ounce in integrity.
Anyone interested in this case really must read Dr Scott McLachlan’s substack articles on it. He’s an ex-nurse, now medical risk analyst, who took a deep dive into the case & especially the court reports regarding the deaths Letby was associated with. It’s horrifying. It includes dues caveats for UK readers:
https://lawhealthandtech.substack.com/p/ll-part-1-hospital-wastewater
https://archive.is/AWpyz
Everyone should read this article.
I cannot believe the grade A lunacy I am reading in these comments. Even the Defence didn’t dispute that at least two of the babies were deliberately poisoned and that Letby was on shift for ALL of these cases. The Police investigation was measured in years before going to trial, and said trial lasted 10 months with the jury deliberating for 1 month. If there was anything to demonstrate that Letby was innocent, does anybody really think the Defence would not have raised it in that time? No, Letby is an evil, murdering POS and deserves the fate she has imposed on herself.
Summary_Healthcare_serial_killer_or_coincidence_statistical_issues_in_investigation_of_suspected_medical_misconduct_September_2022_FINAL.pdf (rss.org.uk) – This is a summary report from the Royal Statistical Society – it’s worth a read.
Just because some of us are critical of the way in which the statistical data was mishandled doesn’t necessarily mean we think Letby is innocent. But had there not been other evidence to suggest her guilt, it could well have provided the defence lawyer with grounds for challenging the verdict – and might ultimately have allowed a criminal to go free.
This is not the first time that statistical evidence has been completely wrong in major trials. There was also the infamous Sally Clark cot death case.
Unfortunately most people are statistically illiterate; indeed that includes some people who have studied statistics and gained qualifications in it. Verbal reasoning can be very misleading, and common sense can be anything but a good guide. Cherry picking data, which occurred here, is a terrible statistical sin, and how a court could allow this to go unchallenged is deeply disturbing.