top of page

Lucy Letby Case: Dewi Evans and the "Doctored" Roster Chart

1 day ago
11 min read

Updated: 10 hours ago

by Peter Elston


Dewi Evans, the lead police and prosecution medical expert in the Letby case, has boasted how easy it was for him to diagnose inflicted harm. And yet, leaked documents show he appears to have changed his diagnosis in relation to 30 of 49 clinical events, often because he diagnosed inflicted harm when Letby was not on duty. This is the "drawing around the arrow" that might prove at the Court of Appeal that the roster chart misled jurors in a material way.




Remember the interviews Dewi Evans gave to Dr Raj Persaud and Tortoise Media in the wake of the verdicts against Lucy Letby in August 2023? In them, Evans boasted how easy it had been for him to diagnose inflicted harm in the case of Baby O, whose medical records he was shown when he first met Cheshire Police officers in July 2017.


Evans told Persaud:


One of the medics had written down that he'd noticed what he thought was bruising over the area of the liver. Well, that doesn't happen. And I got the postmortem results as well. And saw some pretty horrid photographs that showed several areas of bleeding over the liver. In other words, trauma to the liver. And I thought, oh, my God, this is dreadful. This is inflicted injury. So right from the beginning. I said, look, this is inflicted injury.


In relation to the same meeting with the police, he told Tortoise Media:


Immediately, I think within 10 minutes or so of arriving and having a look at these notes and over the coffee, I felt, oh, my God, this baby is the victim of inflicted injury.


Evans wrote three substantive medical expert reports on Baby O. In the first he wrote that the bruise was evidence of harm inflicted "from around" 01.30hr on 23 June. In the second, the time became "from around" 07.00hr.


The change in time is bad enough. That Letby wasn't on duty until the day shift is worse still.


Moreover, the bruise wasn't a bruise as Evans was told it had disappeared quickly. So he instead said it was evidence of intravenous air, while still maintaining the allegation of physical attack despite no mark being left on the baby's skin (remember prosecution pathologist Marnerides' trampoline and car crash comments?) You'd be forgiven for thinking he was making it up as he went along.


All the above is already in the public domain. What follows isn't. For the avoidance of doubt, this blog piece is intended for court consumption (not just media), subject to legal opinion as to the strength of what I have suggested might be challenges to the roster chart.


When Dewi Evans joined the Cheshire Police investigation into Lucy Letby he was given the medical records of 28 babies to appraise. The 28 babies comprised 13 who died and 15 who suffered non-fatal collapses who were selected by The Countess paediatricians (Evans confirmed this to me by email in answer to a question about whether he'd requested medical records in relation to all babies admitted to the NNU during the index period as an epidemiologist would have. Moreover, that events for Evans to appraise were selected by the treating paediatricians who were not only Letby's accusers but also should have been treated by the police as suspects represents a serious conflict of interest).


You might have thought Evans only needed 280 minutes to study the 28 sets of medical records. Instead, he was given three months. On 11 and 12 October he sat down with Cheshire Police to tell them his diagnoses (suspicious, suspicious?, or not suspicious). I have both DC Janet Moore's handwritten and typed up notes of the meeting. Indeed, these are what I used to help David Rose with his important February 2025 article in UnHerd Magazine (a neonatal consultant was able to confirm whether Evans was or was not invoking a suspicion in each instance).


In the case of 11 of the babies, Evans had no suspicions and none of the babies ended up on the indictment (his opinion did not change).


For the other 17 babies, he identified 28 clinical events that he said were either suspicious (21) or suspicious? (7). Of these 28, Letby was not on duty for 10 of them so they never made it onto the indictment.


There was a further event with respect to Baby I at 22:00 on 23 August 2015 (for which Letby was not on duty) that Evans did not cite as suspicious in the October 2017 meeting but he did in his written witness statements. In other words, there were 11 events that Evans said were suspicious but where Letby was not on duty.


Evans said he had "no suspicions" about Baby K, who did end up on the indictment.


With respect to six clinical events that also ended up on the indictment he did not make any comment at all at the October meeting (though later did in written statements).


You may already be thinking that Evans may not have been one of the world's greatest diagnosers of inflicted harm. However, it gets worse.


In relation to the 17 clinical events that Evans said were suspicious or suspicious? and for which Letby was on duty, two of them never made it to the indictment. Of course, whether this was because the CPS felt there was insufficient evidence should be investigated.


It gets worse, still.


Of the other 15, Evans cited a method of inflicted harm that bore no or little resemblance to his later opinion in 7 cases. This may of course be because Evans was provided with new information, something that, again, should be investigated.


In sum, Evans appears to have changed his opinion in 30 of 49 cases, a whopping 61%.


The 30 challenges to the roster chart fall into six categories as below.


Challenge category

Number of instances

Evans deemed suspicious but Letby not on duty

11

Evans did not identify as suspicious but on indictment

6

Apparent change in opinion by Evans

7

Questionable remark about original pathologist comment

1

Timing inconsistencies

1

Evans deemed suspicious but not on indictment

4

Total

30



The detailed table of of information relating to all 49 cases is appended below (click it and "save as image", though there is also a link here - https://t.co/ue2Ddlyyix). Note that Babies F, L and N were not among the 28 babies that Evans was given to appraise at first as they entered the picture only later so they are not among the 49 cases.


Table of of information relating to all 49 cases:


Each of the 30 cases where Evans appears to have changed his opinion, explicitly or otherwise, may well represent a serious challenge to the scientific integrity of the roster chart shown to jurors. And therefore to the safety of Letby's 15 convictions.


Also note that I have not referred in this blog piece to other challenges, for example,


Baby C Letby could not have injected air seen on 12 June x-ray

Insulin babies Farfetched prosecution bag spiking theories

Baby N Letby not on night shift on 14 June 2016 (per roster chart)


These are also I believe important with respect to Letby's appeal process.


Nor have I referred to what prosecutor Nick Johnson told the court in relation to there being no relevant information with respect to the matrix or to Letby's non-presence but these (particularly the latter) are also important. There were two elements to the submission on 31 January 2023 by Myers challenging the roster chart.



To summarise, Myers is making two requests in his submission, based on undisclosed Case Review Flow Charts in respect of two children not on the indictment:


1) To see the matrix criteria in order to determine the extent to which Letby's presence was one of them, and

2) To know, with respect to the incidents appraised, how many Letby was not present for.


After Myers' submissions came those from lead prosecuting barrister Nick Johnson's with respect to Myers' two requests:


1) Matrix criteria


MR JOHNSON: Can I make it entirely clear: this matrix has not been applied to this case.


MR MYERS: Well, I follow that. I suppose if it is of no part in this case, there's little that we can point to where the matrix can have played a part in determining what's on this indictment...


MR JUSTICE GOSS: In which case, arguably, it shouldn't be on the disclosure schedule at all.


MR JOHNSON: I agree.


2) The number of appraised incidents at which Letby not present


MR MYERS: The only matter that remains, insofar as the court sees merit in it, is the point we raised under paragraph 7, which is the extent to which, given the total number of cases, Ms Letby is not present in any of those. If there is material that can be provided setting that out, we would wish to see it because then at least we can gauge the extent to which this investigation has looked beyond just her identity. That's the only matter we are left with in the light of how things have developed here.


MR JUSTICE GOSS: All right. Mr Johnson has said, if I've understood him correctly, that there is no relevant material to disclose in relation to that aspect.


MR JOHNSON: Correct.


MR JUSTICE GOSS: All right. So there we are.


I'm not a lawyer but I would imagine that Nick Johnson must be wondering whether what he said to the judge was correct, namely:


a) The matrix was not used with respect to the [Phase 1] indictment cases, and

b) There was no material indicating whether Letby was not present at other incidents appraised


b) now seems particularly pertinent given the 11 incidents that Evans deemed as suspicious when Letby was not no duty. Although the notes of the meetings between Cheshire Police and Evans in October 2017 and March 2018 which I drew on were in a bundle of unused evidence disclosed to the defence, I think a court would understand why they didn't do the analysis I did. It is hard to imagine just how much documentation the three defence barristers had in front of them to deal with. And how little time they had to digest it all (I understand the only thing that Ben Myers told Mark McDonald when he took over was that he would drown in information). It only occurred to me after having had the two sets of meeting notes for some time that they contained information that spoke to Evans' having made things up as he went along.


Other important points to mention:


  1. In July 2017, Evans was sent Cheshire Police's summary of the case which referred to the allegation against a nurse, as evidenced below. I would be grateful for a legal opinion as to whether he should have been sent this.


"Sonya [NID] has recommended various experts (as per below). We were hoping to meet with Dr Dewi Evans in advance of deciding on the other experts. Is this possible and should we go through yourselves to organise this? I can attach a summary of the enquiry so far if he is agreeable?"


"Dear Janet, In relation to your e-mail below and confirmation of one of our services i.e. instruction of Dr Dewi Evans, I will now formalise his services and e-mail him as well as try and call him (though I have left several messages including to Paul). Once I have made contact I will get his availability to meet and discuss the case further with yourselves and what his costings will be. In the meantime, a copy of the current situation report would be useful please and confirmation that a copy of the case summary can be provided to Dr Evans."


"Hello Sonya, I attach a summary of Op Hummingbird if you can please pass onto Dr Evans? Many thanks, Janet"


From Evans: "Thanks for the message and for the Summary, which I have read."


Key extract from the summary sent to Evans:


"Concerns were raised by the clinical team of the hospital to the Executive regarding a higher than usual number of neonatal deaths from January 2015 (8 in 2015 and 5 in the first six months of 2016 compared with an average of 2.4 per annum in the previous 5 years). This resulted in a meeting between various staff members in May 2016. It was highlighted at this meeting that there was one member of the nursing staff who had been present at more of the cases than any other member of staff. However, there was no evidence, other than coincidence. The nurse was noted to work full time and have the Qualification in Speciality (QIS). She was therefore more likely to be looking after the sickest infant on the unit. She also regularly worked overtime when the acuity was high or unit was over capacity. There were no performance management issues, and there are no members of staff that had complained regarding her performance. The nurse had been moved on to days to ensure that she was supported."


  1. Given he read the police case summary that mentioned the allegation against a nurse, Evans would have been able to work out it was Letby since the medical notes given to him were not redacted.


  1. Evans was sent the treating paediatricians' contemporaneous notes and the nursing notes. These are highly subjective. I understand it would have been better if Evans, at least at first, had been sent only the objective physiological measurements. Evans having sight of the subjective medical notes induces confounding pathways (as well as overdispersion) depending on who's on duty, since different doctors and nurses will have different inclinations of what to report and what not to report. This holds even if the writers of the notes had been successfully anonymised.


  1. Cheshire Police should have at first engaged an epidemiologist/medical statistician to determine the cause of the spike in mortality. They didn't. They believed the paediatricians that it was Letby. They didn't even conduct a formal tender for a neonatologist.They went straight to Evans, who touted for the business, who wasn't a neonatologist, who was long retired. Who simply was not credible.


  1. The below is from an email from Evans to DC Janet Moore on 20 Sep 2017. It suggests that before the October 2017 meeting, which was before he wrote his first set of written statements, Evans was indeed being shown granular shift information, including Letby's. Which he shouldn't have been.


The common theme I have noted so far is that several babies seem to collapse wholly unexpectedly. Some go on to develop additional complications. We'll need to discuss independent neonatal pathology opinion, but that can wait until I've sorted all 28 cases, and you have clarified who was there [doctors and nurses] at the time, and also in which nursery. [If the baby was in one of the small nurseries it's likely that there would be just the one nurse in charge].


It seems that Myers referred to the above email in an application he submitted on 9 January 2023 to have Evans' evidence thrown out (prompted by Myers having received Judge Peter Jackson's written judgment in relation to another trial in which he dismissed a medical report by Evans as "worthless" and accused him of breaching proper professional conduct). In oral argument, Myers noted:


There's an email chain from the 15th to 20 September. It's at page 148 and I will just summarise it, my Lord. You'll see when you look at that, at page 149, Dewi Evans is talking to the police about the plans of the unit, shift arrangements and asking for photographs of various parts of the unit.


What was meant by "who was there [doctors and nurses] at the time, and also in which nursery" and "shift arrangements" should be investigated.


  1. According to the notes of the March 2018 meeting, Evans said, "Consider doing gene sequencing for [Babies OPR] as per Dr Kokai suggestion". In my email correspondence with Evans in late 2024, I asked him, "Did you request full genome sequencing on Guthrie cards to test for genetic factors? If not, why not?" He responded, "I did not request genome sequencing. Why was that necessary? Are you aware of any genetic disorder / inborn error of metabolism that presents the same way as any of the 7 murdered babies? [Two of the dead babies have a surviving twin, who have not shown any unusual disorders. The triplets were ‘identical’. The survivor is fine]."


I welcome comments with respect to any mistakes (whether in the text of this blog or the table); further observations that could/should be made; or legal and other expert opinion about the merit of my inferences.

2 Comments


dcm539
3 minutes ago

It is evident that Lucy Letby is innocent yet the wheels of justice(?) have to turn and due process has to ocurr irrespective of all else. Doctor whose evidence helped to convict Lucy Letby was barred from treating patients during trial | Lucy Letby | The Guardian.

Like

andycannon66
59 minutes ago

Peter, this is excellent work. Thank you. It is clear how the police and Dewi Evans combined to manufacture evidence that did not exist. All of Evans evidence is contaminated by his knowledge of information that is outside of his role as an expert witness. He literally gave different evidence based on whether Letby was in the frame for a case or not. He should have been thrown out of court.

Like
Receive posts direct from Mephitis by email

Note: subscription to emails is not the same as membership of the site. In addition to emails, membership allows the posting of comments etc.  The membership area is here.

Thank you for subscribing 

SOCIAL

  • X

© 2024 by Mephitis

CONTACT

​

​Mephitis is based in London and will see any comments you append to posts, be thankful for them, and may very well respond to them, but if you’d like to contact them directly click here.

Mephitis_transparent.png
bottom of page