Could statistical science have caught Harold Shipman earlier?

with David Spiegelhalter

Get Media:

View Online

The Shipman Inquiry has concluded that there were 215 "confirmed" and 45 "probable" victims of Harold Shipman, and it is natural to ask if he could have been caught earlier if some sort of statistical monitoring procedure had been in place. We show that an adapted version of an industrial quality-control technique could in theory have led to earlier detection and the saving of many lives. In practice, however, the design and introduction of such a scheme would not be straightforward.

David Spiegelhalter FRS is a Senior Scientist in the MRC Biostatistics Unit, where he has worked for over 20 years. His main area of research has been in Bayesian statistics, covering methodology, software and applications. His current interests include the monitoring of clinical performance, and he is a statistical consultant to the Healthcare Commission. He led the statistical team in the Bristol Royal Infirmary Inquiry, and contributed to the Shipman Inquiry.

This lecture was broadcast live on Tuesday, 6th September 2005