Henderson v R.

Where guilt depends wholly on expert medical evidence, a conviction is safe only if there is a logically justifiable basis for accepting the prosecution evidence and rejecting competing evidence. The exclusion of all known natural causes does not itself prove inflicted injury. A realistic possibility of an unknown cause must be excluded beyond reasonable doubt.

Previous judicial findings cannot establish medical propositions. Their evidential force depends on the expert evidence in the particular case. Fresh expert evidence will not ordinarily permit an appeal to become a second trial or a substitute for expert evidence that could properly have been called at trial.

The appellant, a childminder, was convicted at Reading Crown Court of the manslaughter of an 11-month-old child who collapsed and died while in her sole care.

The prosecution case depended on medical evidence. It relied on the triad of encephalopathy, retinal haemorrhages and subdural haemorrhages, together with retinal folds and traumatic axonal injury. The appellant denied any trauma and contended that the death was unexplained.

On appeal, she sought to rely on fresh ophthalmological and neuropathological evidence under Criminal Appeal Act 1968, s 23. The central question was whether that evidence undermined the safety of the jury’s verdict.

1. Appeal dismissed. The fresh evidence did not render the conviction unsafe. The jury had been properly directed that medical knowledge was developing, that the defence bore no burden to provide an alternative explanation, and that they could convict only if they excluded an unexplained cause.

2. The court reaffirmed that, even where all known causes have been excluded, an unknown cause may remain. That possibility must be confronted where it is realistically raised. However, the question on appeal was whether the new evidence undermined the verdict, not whether the court would itself reach the same factual conclusion.

3. The ophthalmological evidence did not do so. Although the mechanism by which retinal folds form was uncertain, the fresh expert could identify no non-traumatic explanation for the severe haemorrhages and folds. His evidence amounted to greater caution, rather than a material challenge to the treating clinician’s evidence that the folds were an additional feature pointing towards trauma.

4. The neuropathological evidence likewise did not assist the appellant. The court rejected the fresh expert’s challenge to the significance of beta-APP staining in the cortico-spinal tracts. His limited and outdated experience, lack of familiarity with important material, and concessions after reviewing the slides meant that his evidence did not undermine the evidence of traumatic axonal injury. That injury was powerful evidence additional to the triad.

5. Under Criminal Appeal Act 1968, s 23, the court heard the evidence de bene esse because of the case’s complexity. In substance, however, it was evidence from the same disciplines as experts whom the defence had obtained but elected not to call at trial. It was not genuinely fresh in a way that could displace the verdict.

The court also gave non-binding guidance for expert-evidence trials. Expert issues should be identified and managed before trial, and the summing-up should address them issue by issue. A jury should be directed to assess the basis, expertise, current clinical experience and independence of conflicting expert opinions.

• Court of Appeal (Criminal Division): The appeal against conviction was dismissed in [2010] EWCA Crim 1269.
• Reading Crown Court: The appellant was convicted of manslaughter by a majority of 10 to 2 on 13 November 2007.