Neil Wilby Media

Cheshire Constabulary’s response to Parliamentary criticism of the Lucy Letby investigation has prompted renewed scrutiny — not only of what was said, but of how the case itself has been constructed, described, and presented.

At the centre of that scrutiny lies a deceptively simple question:

What role did statistical or pattern-based reasoning play in the investigation and prosecution of Lucy Letby?

The answer, on the available public record, is not straightforward. It appears to depend —sometimes markedly — on who is describing the case, and at what stage of the process.

During an adjournment debate speech in March 2026, Sir David Davis MP raised concerns about the potential role of statistical reasoning in the Letby case, drawing comparisons with a historic miscarriage of justice.

In its response, Cheshire Constabulary’s chief constable, Mark Roberts, rejected that proposition in clear terms.

His letter stated that:

• no statistical evidence was relied upon in the prosecution case
• no expert statistical report was obtained
• conclusions were based on clinical and forensic analysis

That formulation is precise, and, in one sense, narrow: it speaks only to the formal evidential framework presented at trial.

However, it does not necessarily resolve how the case was analysed, framed, or understood at earlier stages or even during the trial itself.

Insight into the investigative process comes from comments made by the Senior Investigating Officer, Paul Hughes (pictured above), in a recorded interview broadcast in 2023.

Reflecting on the point at which suspicion crystallised, Hughes stated:

“Our evidence in statistical analysis showed Lucy Letby being present for everything… she’s a suspect…”

This is not a formal evidential statement and was not presented in court. But it is nonetheless revealing.

It indicates that, at the investigative stage, some form of analytical work — described by the SIO himself as “statistical analysis” — played a role in shaping the direction of the inquiry and the identification of a suspect.

That observation does not directly contradict the Chief Constable’s formal position. It does, however, suggest that the term “statistical” is being used differently depending on context.

Further complexity arises from the way in which the prosecutors have described the case.

The Crown Prosecution Service has characterised the evidence as comprising multiple strands, including what has been described as statistical material — a formulation articulated in part by senior prosecutor, Pascale Jones.

If accurately reflecting the CPS’s position (it appeared in a post-trial BBC News article), this introduces an additional layer of interpretation. It suggests that statistical considerations may have been understood, within the prosecutorial framework, as forming part of the overall evidential structure, even if not presented as formal expert statistical testimony at trial.

This distinction — between evidential form and analytical substance — is central to the present debate.

A similar line of reasoning was evident in the clinical evidence presented at trial. Expert witnesses, notably including Dr Dewi Evans, addressed the recurrence and timing of unusual clinical events across multiple cases, offering interpretations that engaged questions of coincidence and likelihood.

While grounded in medical expertise rather than formal statistical modelling, such evidence nonetheless invited the jury to consider whether repeated incidents of this nature could reasonably be attributed to chance.

The way in which the case was quite forcefully presented to the jury by the CPS provides further insight into how such material was framed.

At the very outset of the trial, prosecuting counsel, Nick Johnson KC, introduced a chart, mapping neonatal deaths and collapses against Lucy Letby’s presence on duty.

This document — widely referred to as the “shift chart” — was used to illustrate an apparent association between Letby’s shifts and a series of adverse incidents.

While not presented as statistical evidence in a formal sense, the chart invited the jury to consider patterns across multiple events.

That approach was reinforced in closing submissions, where the jury was repeatedly asked to consider questions framed in terms of likelihood and coincidence:

• “What are the chances of that?”
• “Do you think it’s a matter of coincidence?”
• “That cannot be just coincidence or bad luck”

Such language does not constitute statistical methodology. But it does engage probabilistic reasoning in substance, inviting the jury to draw conclusions based on perceived patterns and their apparent improbability.

A further dimension is provided by the early involvement of Jane Hutton, a leading medical statistician who was approached during the investigative phase.

Professor Hutton is understood to have raised concerns about the proposed analytical approach, including:

• the need to account for base rates
• the risks of selection effects
• the importance of appropriate statistical modelling

A consultancy arrangement was reportedly put in place, but her involvement was not ultimately developed into formal expert evidence.

That sequence is significant—not because it establishes error, but because it demonstrates that statistical considerations were actively engaged and examined within the investigation before being set aside.

Taken together, these accounts present a layered and, at times, divergent picture:

• Police leadership emphasises the absence of formal statistical evidence
• Investigators refer to statistical analysis informing suspicion
• The key trial witness collates events when Lucy Letby was alleged to be present
• The prosecution describes the case as including a statistical strand
• Courtroom advocacy relies on patterns, coincidence, and probability
• Early expert input highlights methodological considerations that were not pursued

These accounts are not necessarily contradictory. On a generous view, they may instead reflect different ways of categorising the same underlying material.

However, the variation in language and emphasis raises an important point: The role of statistical or probabilistic reasoning in the Letby case is not a settled description, as the Cheshire police might wish the public to believe, but an evolving and hotly contested one.

This is not an abstract or academic distinction.

The way in which evidence is:

• analysed
• categorised
• and presented

goes directly to how complex cases are understood by investigators, prosecutors, and ultimately, and most crucially, juries.

In past cases — most notably that of Sally Clark — the misuse or misunderstanding of statistical reasoning has had profound consequences. It is for that reason that questions of methodology, interpretation, and evidential framing continue to attract scrutiny, particularly in cases of exceptional public importance.

None of this neutrally-formed, journalistic analysis determines whether the conclusions reached in the Letby case were right or wrong.

Nor does it suggest that statistical evidence was improperly used: Those are, plainly, matters for experts in this specialist field.

What it does show, however, is that the way in which the investigation and prosecution of Lucy Letby is now being described by Cheshire Constabulary — particularly in response to both public and Parliamentary scrutiny — does not even begin to capture the range of analytical approaches that appear to have been engaged at different stages.

A matter that calls for the transparency and clarity of the police so clearly articulated by David Davis MP. Sooner rather than later.

Reflecting on the Chief Constable’s intervention, noted legal commentator Joshua Rozenberg KC (Hon) has questioned the basis on which the MP’s remarks could be said to undermine the administration of justice, noting that there are no ongoing criminal proceedings relating to the case.

He has also observed that where questions raised in Parliament are not fully addressed by the police, it becomes more difficult to criticise the accuracy of the answers offered in their place.

That observation highlights a broader issue: Where public institutions respond selectively to detailed criticism, the line between rebuttal and avoidance can become increasingly difficult to define. The Mark Roberts letter is a classic case in point.

As debate around the Letby case continues — across Parliament, the media, and ongoing review processes — the focus is increasingly shifting from outcomes to process.

Questions of:

• methodology
• evidential construction
• and transparency

are unlikely to dissipate. If anything, they are becoming much more sharply defined:

In that context, the question is no longer simply what happened in the Lucy Letby case, but how it has been explained to a Member of Parliament raising valid concerns in an appropriate forum. 

Neil Wilby concludes: “The response of David Davis to the chief constable’s letter will, doubtless, open up this dispute to a wider and increasingly perplexed audience. A growing majority of whom simply cannot comprehend how a popular, dedicated, caring, effective nurse is serving 15 whole life orders on the basis of the case presented at court and what has been uncovered by her campaigners since”.

________________________________________________________________________________________________

Neil Wilby is a journalist, court reporter and transparency campaigner who has reported on police misconduct, regulatory failures, and criminal and civil justice since 2009. He is the founder and editor of Neil Wilby Media, launched in 2015.

Page last updated: Tuesday 14th April 2026 at 08h55

Corrections: Please let me know if there is a mistake in this article. I will endeavour to correct it as soon as possible.

Right of reply: If you are mentioned in this article and disagree with it, please let me have your comments. Provided your response is not defamatory, it will be added to the article.

Image credits: Cheshire Constabulary

Contact: Email us via this weblink

© Neil Wilby 2015–2026. Unauthorised use, or reproduction, of the material contained in this article, without permission from the author, is strictly prohibited. Extracts from, and links to the article may be used, provided that credit is given to Neil Wilby, with appropriate and specific direction to the original content.