@davidallengreen I read the Phil Hammond piece in Private Eye, and the most staggering point he raised (quoted in the Adam King blog you linked to) was about a recent review of neonatal deaths that found no identifiable cause of death for around 50% of cases. Astonishing, and really relevant, when much of the evidence againsi Letby was of a statistical nature. That kind of figure really needs a Bayesian approach to any kind of inference drawn.