Clark, R v

Where admissible fresh expert evidence presents a respectable alternative explanation for death and might reasonably have affected the trial jury’s verdict, the conviction is unsafe. An appellate court should not resolve a difficult conflict between reputable experts by substituting its own view for that of the jury.

In suspected infant-death cases, a pathologist must disclose abnormal ancillary findings to those reviewing the case. Statistical recurrence figures cannot determine whether an individual death was natural or unlawful, and evidence which encourages a jury to decide separate counts as a single statistical package should ordinarily be excluded.

Sally Clark had been convicted at Chester Crown Court, by majority verdicts, of murdering her sons Christopher and Harry. Her original appeal was dismissed. The Criminal Cases Review Commission later referred the case under section 9 of the Criminal Appeal Act 1995 after undisclosed microbiological reports concerning Harry were discovered.

The reports supported a respectable medical opinion that Harry may have died from overwhelming staphylococcal infection. The court considered whether that fresh evidence rendered the convictions unsafe. It also addressed the non-disclosure by the pathologist and the statistical evidence that two natural infant deaths in the family had a probability of one in 73 million.

1. Appeal allowed; convictions quashed. The fresh microbiological evidence concerning Harry was admissible fresh evidence for the purposes of section 23(2) of the Criminal Appeal Act 1968. Applying Pendleton [2002] 1 Cr. App. R. 441, the court was required to assess the evidence but, in a difficult case, to ask whether it might reasonably have affected the trial jury’s decision.

2. Professor Morris’s opinion that infection was the most likely cause of Harry’s death was a respectable medical opinion. Dr Klein’s contrary view was also respectable, but it did not eliminate the realistic prospect that a jury could have accepted Professor Morris’s opinion. The court was not entitled to resolve that conflict in place of the jury. Harry’s conviction was therefore unsafe.

3. The conviction for Christopher depended materially on the inference drawn from Harry’s death. The medical evidence relating to Christopher alone could not safely establish an unnatural death. Once Harry might have died naturally, no safe conclusion remained that Christopher had been murdered. His conviction was also unsafe.

4. Although unnecessary to the disposition, the court held that Dr Williams’s failure to disclose the abnormal microbiological results fell seriously below the required standard. A pathologist conducting the original post-mortem must make available material findings and test results, including abnormal results he considers capable of being discounted, so that other experts can reach their own informed conclusions. The court found no bad faith, but described the explanation for withholding the material as wholly unacceptable.

5. The statistical evidence should not have been placed before the jury in the manner used. Squaring an individual SIDS-risk figure did not assist determination of the causes of these particular deaths and encouraged the jury to consider separate counts together. The court considered that the evidence would probably have been excluded if its admissibility had been challenged.

6. The Crown properly did not seek a retrial. Further contemporaneous testing was no longer possible, publicity created additional difficulty, and the divided medical evidence could not safely support a finding of murder.

• Court of Appeal (Criminal Division): on a reference by the Criminal Cases Review Commission, allowed the appeal and quashed both convictions: [2003] EWCA Crim 1020.
• Court of Appeal (Criminal Division): dismissed the original appeal on 2 October 2000.
• Crown Court at Chester: convicted Sally Clark, by majority verdicts of 10 to 2, of the murders of Christopher and Harry on 9 November 1999.