Fresh Expert Psychiatric Evidence Is Not “Fresh” If It Re-runs Trial Material: No Second-Opinion Appeals on Diminished Responsibility

1. Introduction

Thomas (Previously Edgington) v R [2026] EWCA Crim 353 is a Criminal Division decision arising from a Criminal Cases Review Commission (CCRC) reference concerning the safety of a murder conviction where diminished responsibility was the central trial issue.

The appellant had been convicted in 2013 of (i) the murder of Sally Hodkin and (ii) the attempted murder of Kerry Clark. The appellant admitted the acts but disputed the requisite mental states: she contended that the killing of Mrs Hodkin amounted only to manslaughter by diminished responsibility, and that she lacked intent to kill Ms Clark.

The appeal—brought after a CCRC reference—focused on an alleged unreliability and selectivity in the trial evidence of the prosecution psychiatrist, Dr Philip Joseph, and sought to adduce fresh psychiatric evidence (Professor Rix and Dr Farnham, supported by post-conviction clinical material) to show that the verdict of murder was unsafe and should be substituted with manslaughter.

Key issues

  • Whether additional psychiatric opinions criticising a trial expert—based largely on material available at trial—constitute admissible fresh evidence under section 23 Criminal Appeal Act 1968.
  • Whether, even if admissible, the new psychiatric material might reasonably have affected the jury’s verdict (the Pendleton safety approach).
  • The proper boundary between legitimate fresh-evidence review and impermissible relitigation of an expert dispute already tried before a jury.

2. Summary of the Judgment

The Court of Appeal (Criminal Division) refused to admit the proposed fresh evidence and dismissed the appeal.

Although the new reports were credible and would have been admissible in principle, the Court held that they did not provide a proper basis for allowing the appeal because:

  • The core criticisms of Dr Joseph were based on material available at trial, and were in substance put to him in a robust cross-examination.
  • The appeal would, in reality, amount to re-running the same expert controversy with different experts—precisely what R v Kai-Whitewind [2005] EWCA Crim 1092 and R v Hunnisett [2021] EWCA Crim 265 warn against.
  • Post-conviction clinical material was not uniformly supportive of the appellant; it contained mixed indications (including support for personality disorder features).
  • Crucially, the jury’s decision was not dependent on psychiatric opinion alone: there was substantial non-medical evidence (notably CCTV and contemporaneous conduct) from which the jury could rationally reject diminished responsibility.

The Court therefore rejected all three grounds and concluded the murder conviction was safe.

3. Analysis

3.1 Precedents Cited

R v Edgington [2013] EWCA Crim 2185

This earlier refusal of leave to appeal is significant procedurally: it established that the original conviction had already passed appellate scrutiny, and framed the present appeal as one dependent on new argument or new evidence (as is typical in a CCRC reference context). The 2026 Court noted that the 2013 appeal did not criticise Dr Joseph; this reinforced the conclusion that the present complaints were not genuinely “newly discovered” issues but rather a different forensic focus years later.

R v Pendleton [2001] UKHL 66, [2002] 1 WLR 72

The Court treated Pendleton as setting the enduring test for fresh-evidence appeals: save in a clear case, the appellate court should ask whether the new evidence, if heard at trial, might reasonably have affected the jury’s decision. This is not a pure merits re-trial; it is a safety inquiry sensitive to the appellate court’s disadvantage in reconstructing the full trial dynamic.

Here, even having considered the new psychiatric evidence de bene esse, the Court’s reasoning shows that the “might reasonably have affected” question was answered negatively because the expert dispute would remain materially the same, and the jury had strong independent evidence of deliberation and control (CCTV, acquisition and concealment of knives, targeting of victims, admissions shortly after).

R v Kai-Whitewind [2005] EWCA Crim 1092

Kai-Whitewind supplied the controlling principle for fresh expert evidence: where expert evidence was given and apparently rejected by the jury, it will be only in the “rarest of circumstances” that a near repetition by another expert can found a successful appeal. Otherwise, the trial becomes a “dry run” for replacing experts after an adverse verdict.

The Court applied that rationale directly: admitting Professor Rix/Dr Farnham would require admitting Professor Blackwood in response, reproducing a two-versus-one clash akin to the trial’s structure—so the difference would be identity of experts, not the evidential problem the jury had to decide.

R v Hunnisett [2021] EWCA Crim 265

Hunnisett was cited as reiterating and reinforcing Kai-Whitewind. The Court used it to underline that the appellate fresh-evidence jurisdiction is not designed to re-stage expert contests where the underlying materials and issues were already available and ventilated at trial.

3.2 Legal Reasoning

(a) The statutory framework: diminished responsibility, CCRC references, and fresh evidence

  • Homicide Act 1957, section 2 (as set out in the judgment) supplies the elements of diminished responsibility: (i) abnormality of mental functioning from a recognised medical condition; (ii) substantial impairment of one or more specified abilities; and (iii) causal explanation (causes or significantly contributes). The defence bears the burden on the balance of probabilities.
  • Criminal Appeal Act 1995, section 13 governs CCRC references: the Commission refers where there is a “real possibility” the conviction would not be upheld, typically because of new evidence or a new argument.
  • Criminal Appeal Act 1968, section 23 governs whether the Court of Appeal will receive evidence not adduced at trial, focusing in particular on: credibility, admissibility, explanation for non-use at trial, and (critically) whether the evidence would afford a ground to allow the appeal.

(b) The decisive move: “fresh” means more than a new expert voice

The Court accepted that the proposed evidence was credible and, in principle, admissible. The appeal turned on section 23(2)(iv): whether the evidence would afford a ground for allowing the appeal.

The Court’s core reasoning was that the appellant’s case depended on portraying Professor Rix’s criticisms as discrediting Dr Joseph. But Professor Rix expressly acknowledged that most of his opinion was based on information already available at trial; and the alleged selectivity/partisanship points were in substance explored through lengthy cross-examination.

Accordingly, the Court treated the application as an attempt to repackage trial arguments with a new expert and a decade of hindsight, rather than as the discovery of material previously unavailable. This is the key doctrinal message of the case: fresh evidence is not created by commissioning a new expert to re-interpret existing trial materials.

(c) Post-conviction clinical history did not transform the case

While some later treating clinicians maintained a schizoaffective diagnosis and the appellant experienced serious relapse post-sentence, the Court stressed that post-trial evidence did not point in one direction; one treating psychiatrist (Dr Inga) supported a personality disorder view. The Court also held that any admission of the new defence evidence would fairly entail admission of the Crown’s responsive expert (Professor Blackwood), who supported Dr Joseph’s essential conclusions.

(d) The “whole evidence” context: the jury were not deciding on psychiatry alone

A substantial part of the Court’s reasoning is trial-structural: the jury were directed that they could assess expert evidence in light of all the evidence. The Court emphasised the non-medical evidence available to the jury, including CCTV showing apparently calm and purposeful behaviour (travel, phone use, knife purchase and concealment, acquisition of a second knife, selection of lone victims), and the timing of the appellant’s later claims of psychosis.

This supports the conclusion that even if the psychiatric evidence shifted at the margins, the jury could still rationally reject substantial impairment and causal explanation based on the overall picture of deliberation and control.

(e) Disposition of the grounds

  • Ground 1 failed because Dr Joseph’s view about the 2005 plea and diagnosis was a genuine professional opinion necessary to explain his stance; excluding it would have distorted the jury’s understanding of why he opposed diminished responsibility in 2011.
  • Ground 2 failed because the alleged selectivity/incompleteness was explored at trial in the presence of extensive medical documentation and witness evidence, and the new material did not materially change the strength of those contentions.
  • Ground 3 was treated as a reformulation of Ground 2 and failed for the same reasons; the Court found no basis to say the jury were misled.

3.3 Impact

(1) Appeals based on “new” psychiatric opinions after a contested trial

The judgment reinforces a strict approach to section 23 applications where the trial featured a contested expert issue and the appellant later seeks to introduce further expert commentary that largely re-analyses the same underlying records. The decision underscores that:

  • Absent genuinely new material (or an inability to deploy material at trial), a later expert critique will often be characterised as impermissible relitigation.
  • Where admission would simply recreate the trial’s expert polarity, the Court is likely to treat the application as falling within the “dry run” prohibition.

(2) CCRC references do not lower the appellate threshold for “freshness”

Although the CCRC referred the case on a “real possibility” basis, the Court signalled that the appellate court’s section 23 and safety analyses remain exacting: a reference does not itself supply “fresh evidence”, nor does it convert trial-disputed expert matters into appealable errors absent something materially new.

(3) Diminished responsibility litigation strategy

The case highlights the practical forensic reality that diminished responsibility trials often turn on the interaction of expert opinion and objective conduct evidence (CCTV, planning, concealment, post-event admissions). Future cases may cite this decision when resisting appeals that focus narrowly on diagnostic labels (schizophrenia vs personality disorder) without showing a realistic effect on the jury’s assessment of “substantial impairment” and causation in light of the whole record.

4. Complex Concepts Simplified

  • Diminished responsibility: a partial defence reducing murder to manslaughter if a recognised medical condition caused an abnormal mental functioning that substantially impaired relevant abilities (understanding conduct, rational judgment, self-control) and explained the killing.
  • Balance of probabilities: the defence must prove diminished responsibility is more likely than not (a lower standard than “beyond reasonable doubt”).
  • Substantial impairment: more than trivial or insignificant; it does not require total loss of control or understanding.
  • Fresh evidence (section 23): evidence not used at trial that the Court may receive if it is credible, would have been admissible, there is an explanation for its absence (where relevant), and—most importantly—it could provide a ground to allow the appeal.
  • De bene esse: the Court considered the new evidence provisionally “for what it is worth”, while still deciding it should not be formally admitted.
  • “Dry run” principle: appellate courts generally will not allow parties to re-run a failed expert case by presenting another expert saying similar things, unless truly exceptional circumstances exist.

5. Conclusion

Thomas (Previously Edgington) v R [2026] EWCA Crim 353 consolidates an important boundary in criminal appeals: fresh expert psychiatric evidence will not be admitted where it substantially re-argues issues already tried on the basis of material available at trial. The Court applied R v Pendleton [2001] UKHL 66, [2002] 1 WLR 72, R v Kai-Whitewind [2005] EWCA Crim 1092, and R v Hunnisett [2021] EWCA Crim 265 to refuse what would have been, in effect, a second attempt to persuade a tribunal on the same expert dispute.

The judgment’s broader significance lies in its insistence on viewing diminished responsibility through the lens of the whole evidence, not diagnostic debates alone, and in reaffirming that the appellate fresh-evidence jurisdiction is not a mechanism for expert “replacement” after conviction.