BK-S (Children) (Expert Evidence and Probability)

The court upheld factual findings in care proceedings where the timing and likely effects of drug ingestion were established by unchallenged expert evidence and supported by the surrounding facts. A judge may draw a proper inference from scientific evidence without making an impermissible calculation or speculation. The inherent improbability of a parent harming a child does not displace reliable evidence. In cases involving harm where the perpetrator is unknown, a split hearing confined to perpetration will generally be inappropriate because the welfare context and assessments of those in the pool may assist in determining the facts.

Hampshire County Council brought public law children proceedings concerning four children. The appeal concerned findings that the appellant mother was the sole perpetrator of the administration of Olanzapine to Z, a six-month-old child, on several occasions. The Family Court at Portsmouth relied on toxicological evidence about the drug’s half-life, concentration and likely symptoms, together with the factual evidence concerning the adults who had contact with Z.

The mother appealed the findings of fact, principally arguing that the judge had attributed an unjustifiably fixed period for excretion and had wrongly calculated the likely time of administration. The central issues were whether the judge’s inference was properly supported by the evidence and whether the proceedings had appropriately been conducted as a split hearing.

1. Appeal dismissed. Lord Justice Ryder gave the reasons, with Lord Justice Briggs and Lord Justice Longmore agreeing.
2. The judge was entitled to rely on uncontradicted expert evidence that a cautious half-life assumption for Olanzapine in a child was 18 hours. The expert evidence permitted interpolation from the later test result to estimate the concentration during the father’s and paternal grandmother’s last contact with Z.
3. That evidence, together with the absence of the expected severe symptoms during that contact and on Z’s return to his mother, supported the inference that the dose identified by the later test had been administered after the contact. The conclusion was a proper inference from the evidence, not the judge’s own speculation or an unwarranted calculation.
4. The judge was also entitled to consider the other episodes, medical evidence and surrounding circumstances in concluding that the mother was the perpetrator. The inherent improbability of a parent poisoning a child did not displace the actual evidence.
5. The court repeated the guidance in In the matter of S (A Child) [2014] EWCA Civ 25 that a split hearing is generally inappropriate where harm has occurred but the perpetrator is unknown. The issue is perpetration in the context of the harm and welfare proceedings, and social work assessments of persons in the pool may assist in assessing the allegations, witnesses and other material.
6. The court noted that the split-hearing decision in this case had not been right, but that procedural criticism did not undermine the findings. The welfare hearing remained undisturbed.

• Court of Appeal (Civil Division): appeal against findings of fact made by His Honour Judge Levey in the Family Court at Portsmouth on 11 June 2014. The appeal was dismissed.