Authored By: Bhumika Sham Jagtiani
Middlesex University Dubai
Introduction
Circle back to August 2023, Lucy Letby became the most prolific child killer to be convicted in modern British legal history.[1] The jury at Chester Crown Court found her guilty of murdering seven babies and attempting to murder seven more at the Countess of Chester Hospital’s neonatal unit. The verdict rested substantially on expert medical opinion. Within two years, however, the heavily relied-upon expert opinion was found to be flawed; an independent international panel of senior neonatologists concluded that the collapses attributed to Letby were explicable by natural causes and that the medical analysis central to two of the convictions was scientifically flawed.[2] Ever since, the Letby case has sparked intense public debate. However, the more fundamental and lasting question is not about Letby’s guilt, but about whether the English criminal justice system has sufficient mechanisms to manage convictions based primarily on expert evidence, especially when that evidence is later challenged. Academic commentators and the Law Commission have questioned whether the existing framework provides adequate safeguards where convictions depend heavily upon expert evidence.[3]
This article contends that the Letby case should be seen not simply as an isolated controversy, but as a symptom of a recurring structural issue within modern English criminal justice. Specifically, the system lacks a clear, principled procedure for resolving genuinely disputed scientific evidence, whether at trial or on appeal. The ‘real possibility’ standard applied to the Criminal Cases Review Commission (the ‘CCRC’) under section 13 of the Criminal Appeal Act 1995 exacerbates this issue by causing the Commission to be overly cautious, even when decisive action is most needed.[4]
This article continues in Section II by outlining the history of wrongful convictions in England and Wales arising from flawed expert evidence, demonstrating the pattern that Letby represents the latest example of. Section III examines the statutory framework governing criminal appeals and CCRC referrals, including a structural critique of the real possibility test. Section IV evaluates the fresh evidence now before the CCRC in the Letby case, applying the legal framework to its principal features. Section V identifies the deeper systemic failures the case exposes and offers proposals for reform.
A Recurring Crisis: Expert Evidence and Wrongful Convictions in England and Wales
The issue of convictions based on expert medical evidence that later proves unreliable is not new to the English courts. In its pivotal 2011 report on expert evidence in criminal cases, the Law Commission highlighted several miscarriages of justice caused by this problem and criticised the then-current permissive approach to admitting expert evidence as ‘laissez-faire’.[5] The Commission observed that an excess of expert opinion evidence was presented in criminal cases without proper examination, and that juries tended to rely heavily on expert witnesses, which could lead to unsafe verdicts.[6]
Two specific cases significantly influenced this critique and served as direct predecessors to the Letby controversy. In R v Clark, Sally Clark was convicted of murdering her two infant sons on the basis of expert medical evidence that included a notorious ‘one in seventy-three million’ statistical claim.[7] That figure, which claimed to show the probability of two sudden infant deaths in the same family, was later proven to be based on a major statistical mistake. New expert evidence demonstrated that the infants’ deaths aligned with natural causes, leading to the overturning of the conviction.[8] In R v Cannings, the Court of Appeal quashed the convictions of Angela Cannings for the murder of her three infant sons, holding that where ‘the medical evidence is genuinely uncertain and expert opinion is divided, it is not safe to convict’.[9] The Court articulated what has become a foundational principle: ‘if the outcome of the trial depends exclusively or almost exclusively on a serious disagreement between distinguished and reputable experts, it will often be unwise, and potentially unsafe, to proceed.[10]
Both cases share features with the Letby prosecution: infant deaths in medical settings, convictions based on expert interpretation rather than direct evidence, and expert opinions later challenged by new science. The 2011 Law Commission report aimed to address such systemic issues, but the failure to implement the proposed reliability test left vulnerabilities unaddressed.[11]
III. The Statutory Framework: New Evidence, the CCRC, and the Challenge of the Real Possibility Test
Fresh Evidence on Appeal: Section 23 of the Criminal Appeal Act 1968
The admissibility of fresh evidence in criminal appeals is governed by section 23 of the Criminal Appeal Act 1968[12]. Ladd v Marshall outlined a test for the Court of Appeal to assess whether new evidence is credible, whether it could justify allowing the appeal, whether it would have been admissible at trial, and whether there is a justification for its absence at the original trial.[13] The three limbs of the Ladd v Marshall test specify that: first, the evidence could not have been obtained with reasonable effort for trial purposes; second, the evidence would likely have significantly influenced the outcome; and third, the evidence appears credible, even if not undeniably so.[14] Each part serves a specific function: the first prevents strategic withholding of evidence; the second prevents the waste of appellate resources on minor matters; and the third filters out unreliable opinions, regardless of their potential impact. Collectively, these standards are strict but manageable, especially when the new evidence represents genuine scientific progress made after the trial.
The CCRC and the Real Possibility Test
A series of notable wrongful convictions, including the ‘Birmingham Six’ and ‘Guildford Four’ cases, revealed the lack of a fair and effective appeal system.[15] This issue was further emphasised by the Royal Commission on Criminal Justice (the Runciman Commission), which highlighted the necessity for an independent body with investigative authority to review cases after appeals.[16] This led to the establishment of the Criminal Cases Review Commission (CCRC). The CCRC was established under section 13(1)(a) of the Criminal Appeal Act 1995, outlining that the CCRC may refer a conviction to the Court of Appeal only where there is a ‘real possibility’ that the conviction would not be upheld.[17]
The test has faced criticism from academics and institutions, notably after the Westminster Commission on Miscarriages of Justice review of the CCRC.[18] The review concluded that the test’s ‘predictive nature encouraged the CCRC to be too deferential to the Court of Appeal’, essentially prioritising the Court of Appeal’s predictions over the Commission’s independent judgment.[19] Similarly, scholars Fayanju and James submitted that the CCRC’s operations under section 13 are ‘ at times unduly constrained,’ with undue deference given to the Court of Appeal’s likely approach.[20] Alternatively, the Westminster Commission recommended redrafting the real possibility test to allow references where the CCRC determines the conviction ‘may be unsafe’ or that it is in the interests of justice to refer.[21] The original test remains; the recommendation was not enacted. [22]
Consequently, the CCRC continues to make predictive judgments about judicial conservatism rather than a primary assessment of the evidence. High-profile cases with convictions that attract strong public interest exert particular pressure on the CCRC’s assessment, mirroring the exact position the Letby case places the CCRC in.
The Structural Problem: Courts and Contested Science
English criminal courts currently have no rigid doctorine for resolving contested scientific evidence. Ward identifies the current approach instead as a standard of ‘sufficient reliability’, requiring only an arguable scientific basis for the new evidence to be admitted, leaving the competing experts’ opinions to the jury to resolve.[23] Conversely, the Law Commission proposed a statutory reliability test that would require judges to assess admissibility before presenting expert opinions to the jury.[24] The government rejected the test, citing the fact that additional admissibility hearings would increase costs more than the expected benefits would justify. Now, juries continue to resolve disputes between competing experts in domains like neonatology, forensic pathology, and psychiatric evidence, where they have no independent means of evaluation. The Court of Appeal in Cannings acknowledged this structural vulnerability and addressed it by quashing the conviction. However, this retrospective action does not prevent similar operational vulnerabilities from arising in future cases.[25]
The Letby Case: Applying the Framework to the Fresh Evidence
The leave to appeal in Letby’s case was refused by the Court of Appeal in 2024. Following that refusal, Dr Shoo Lee coordinated an independent international panel of neonatologists, drawing on over 250 peer-reviewed references, many of which were published after the trial. The panel concluded that there was no medical evidence supporting malfeasance and identified multiple natural alternatives for the infants’ deterioration, including sepsis, infection, and pre-existing vulnerabilities. The insulin analysis concerning Babies F and L, which had formed a significant plank of the prosecution’s case on two counts, was found to reflect typical C-peptide levels in preterm infants, levels that the original trial evidence had mischaracterised as indicative of exogenous insulin administration.[26]
The admissibility of fresh evidence on appeal is governed by section 23 of the Criminal Appeal Act 1968 and the common law framework of Ladd v Marshall, which requires that the evidence could not have been obtained with reasonable diligence, would have likely influenced the result, and is credible.[27] In this case, these conditions are met: the evidence is recent scientific literature and challenging expert interpretation, which could have affected the verdict. The credibility is supported by the panel’s methodology and expertise, including their focus on methodological rigour and empirical findings about courts’ receptiveness to such evidence. The evidence was not previously raised in the proceedings, and the findings differ from the Dr Shoo Lee report that was rejected on leave to appeal in 2024 . The case remains pending a comprehensive review by the CCRC. How this new finding will influence the ‘real possibility’ and whether the new discovery will even be regarded as fresh evidence is yet to be seen. However, the Letby controversy is not principally about Letby. It pertains to a system that has not sufficiently addressed the issues identified by Clark and Cannings more than two decades ago. Whether the International Expert Panel’s conclusions ultimately prove persuasive is a matter for the CCRC and, potentially, the Court of Appeal. The significance of the Letby case lies not solely in the strength of the fresh evidence itself, but in the institutional difficulties that arise when convictions depend heavily on specialised scientific opinion that later becomes contested. It is these difficulties that reveal broader weaknesses in the existing framework.
Structural Failure and the Case for Reform
Three interlocking structural failures are visible. The first failure is at trial. The Law Commission’s proposed statutory reliability test, which would have required judges to screen expert evidence for scientific adequacy before it reached the jury, was not enacted. [28]As a result of this lack of change, juries continue to adjudicate between competing scientific theories in technically complex fields, with no independent means of evaluating the methodological foundations of the opinions they hear. Whether or not the expert evidence at the Letby trial was reliable, the system offered no mechanism to assess that question independently before the verdict was returned.
The second failure occurs at the appellate level. The Court of Appeal relies on general admissibility rules such as section 23 of the Criminal Appeal Act 1968 and Ladd v Marshall, which are not tailored to address challenges involving convictions based on scientific evidence that has become outdated due to new scientific developments. [29]There is no specialised scientific advisory body to help the Court determine whether contested expert opinions truly reflect a significant shift in the field or merely a minority viewpoint. Judges are tasked with evaluating the epistemological validity of conflicting neonatal or forensic pathological theories without specific training or institutional support to guide this assessment.
The third failure is at the CCRC level. The real possibility test, as currently applied, requires the Commission to predict the Court of Appeal’s response to the fresh evidence rather than to make its own primary assessment of whether the conviction is safe. The Westminster Commission’s 2021 recommendation, that the test be redrafted to permit referral where the CCRC finds that a conviction may be unsafe, was directed precisely at this problem.[30] Its non-implementation leaves the CCRC institutionally inhibited from performing its core function in exactly the cases where that function matters most.
This article submits that three reforms are required. First, Parliament should implement the Law Commission’s 2011 recommendation for a statutory reliability test for expert evidence in criminal proceedings. Second, the Court of Appeal should be empowered, in cases where the safety of a conviction turns on contested scientific evidence, to appoint an independent expert panel to assess the current state of the relevant discipline. Third, the real possibility test in section 13 of the Criminal Appeal Act 1995 should be redrafted in line with the Westminster Commission’s recommendation, enabling the CCRC to refer where it determines that a conviction may be unsafe without requiring it to second-guess the Court of Appeal’s likely disposition.[31]
Conclusion
The Letby case is part of a pattern that should now be familiar. Convictions heavily reliant on expert medical testimony, evidence that juries cannot fully assess scientifically, have led to some of the most serious miscarriages of justice in modern England. Sally Clark spent over three years in prison for murders she did not commit and died before the injustice could be corrected. Angela Cannings was convicted based on a statistical claim that was fundamentally flawed according to probability theory. The Law Commission identified systemic issues that made both cases possible, yet successive governments refused to act on its recommendations.
In the Letby case, this article argues that the international expert panel’s findings meet the legal criteria for admissibility under section 23 of the Criminal Appeal Act 1968 and the common law requirements of Ladd v Marshall. These findings indicate a real possibility, per section 13 of the Criminal Appeal Act 1995, that the Court of Appeal might overturn the convictions. But regardless of whether the CCRC refers Letby’s case, and whatever the outcome, the key lesson remains: a criminal justice system that cannot properly evaluate contested scientific evidence or correct unsafe verdicts based on scientific advances will face ongoing crises. The core issue is not whether Lucy Letby is guilty, but whether the law is capable of knowing.
Bibliography:
Primary sources:
Cases:
– R v Letby [2024] EWCA Crim 748
– Ladd v Marshal [1954] 1 W.L.R. 1489
– R v Clark [2003] EWCA Crim 1020
– R v Campbell [2024] EWCA Crim 103
– R v Cannings [2004] EWCA Crim
– R v McIlkenny [1991] 2 Cr App R 287
– R v Richardson; R v Conlon; R v Armstrong; R v Hill (Court of Appeal Criminal Division, 19 October 1989) The Times (20 October 1989)
Statutes:
– Criminal Appeal Act 1968
– Criminal Appeal Act 1995
Secondary sources:
Journals:
– Ayokunle Fayanju and Annabelle James, ‘The Criminal Cases Review Commission — A Quarter of a Century On’ (2025) 89(3) Journal of Criminal Law
– Tony Ward, ‘Explaining and Trusting Expert Evidence: What is a “Sufficiently Reliable Scientific Basis”?’ (2020) 24(3) International Journal of Evidence & Proof 219
Reports/ Official Publications:
– Law Commission, Expert Evidence in Criminal Proceedings in England and Wales (Law Com No 325, 2011) https://lawcom.gov.uk/project/expert-evidence-in-criminal-proceedings/#4-Updates accessed 23 June 2026
– Royal Commission on Criminal Justice, Report (Cm 2263, 1993) (the Runciman Commission)
– Westminster Commission on Miscarriages of Justice, In the Interests of Justice: An Inquiry into the Criminal Cases Review Commission (All-Party Parliamentary Group on Miscarriages of Justice, March 2021) (the Westminster Commission Report)
Other:
– Dr Neil Aiton and 6 others, ‘Summary of Joint Expert Witness Report on baby F and L, 2nd April 2025, < https://lucyletbyinnocence.com/shoolee/Summary%20of%20Joint%20Expert%20Witness%20Insulin%20Report%20on%20Babies%20F%20and%20L.pdf> accessed 12 December 2025
– Criminal Cases Review Commission, ‘CCRC refers R v Sally Clark to the Court of Appeal’ (March 2002) < https://ccrc.gov.uk/decision/clark-sally/accessed 12 December 2025
[1] R v Letby [2024] EWCA Crim 748.
[2] Dr Neil Aiton and others, ‘Summary of Joint Expert Witness Report on Babies F and L’ (2 April 2025) < https://lucyletbyinnocence.com/shoo-lee/Summary%20of%20Joint%20Expert%20Witness%20Insulin%20Report%20on%20Babies%20F%20and%20L.pdf > accessed 23rd June 2026.
[3] Law Commission, Expert Evidence in Criminal Proceedings in England and Wales (Law Com No 325, 2011)
[4] Criminal Appeal Act 1968, s 13.
[5] Law Commission (n3), paras 1.6–1.10.
[6] ibid, para 1.7.
[7] R v Clark [2003] EWCA Crim 1020; Law Commission (n 4) para 1.5.
[8] Criminal Cases Review Commission, ‘CCRC refers R v Sally Clark to the Court of Appeal’ (March 2002) < https://ccrc.gov.uk/decision/clark-sally/> accessed 23 June 2026.
[9] R v Cannings [2004] EWCA Crim 1.
[10] ibid [178] (Judge LJ).
[11] Law Commission (n 3).
[12] CAA 1968, s 23.
[13] Ladd v Marshall [1954] 1 WLR 1489.
[14] Ibid.
[15] R v McIlkenny [1991] 2 Cr App R 287 (CA); R v Richardson; R v Conlon; R v Armstrong; R v Hill (Court of Appeal Criminal Division, 19 October 1989) The Times (20 October 1989).
[16] Royal Commission on Criminal Justice, Report (Cm 2263, 1993) (the ‘Runciman Commission’).
[17] CAA 1995, s 13(1)(a).
[18] Westminster Commission on Miscarriages of Justice, ‘In the Interests of Justice: An Inquiry into the Criminal Cases Review Commission’ (All-Party Parliamentary Group on Miscarriages of Justice, March 2021) (the ‘Westminster Commission Report’).
[19] Ibid, page 7.
[20]Ayokunle Fayanju and Annabelle James, ‘The Criminal Cases Review Commission — A Quarter of a Century On’ (2025) 89(3) J Crim L, citing C Hoyle, ‘The Challenges for England’s Post-Conviction Review Body’ (2020) 13(4) Erasmus Law Rev 33.
[21] Westminster Commission Report (n 18), Recommendation 5.
[22] Law Commission, Expert Evidence in Criminal Proceedings in England and Wales (Law Com No 325, 2011) https://lawcom.gov.uk/project/expert-evidence-in-criminal-proceedings/#4-Updates accessed 23 June 2026.
[23] Tony Ward, ‘Explaining and Trusting Expert Evidence: What is a ‘Sufficiently Reliable Scientific Basis’?’ (2020) 24(3) International Journal of Evidence & Proof 219.
[24]Law Commission (n 3) para 1.20.
[25] R v Cannings (n 9) [178].
[26] Aiton and others (n 2).
[27] Ladd (n 13).
[28] Law Commission (n 3).
[29] CAA (n4); Ladd (n13.
[30] Westminster Commission Report (n21).
[31] CAA (n17); Westminster Commission Report (n 18).
