Same patient, same evidence, different treatment:

There’s a saying in anaesthesia: ask three anaesthetists how to put a patient to sleep, and you’ll get five different answers. This talk explores when variation in practice fits care to the patient, and when it reflects genuine uncertainty (equipoise) about which option is best.

Drawing on the learning health system, it introduces decision architecture randomised trials (DART), which use nudges in the electronic health record to generate evidence from routine care, and sets out a PhD roadmap: find variation with true equipoise, prioritise the questions that matter, and use DART to answer them.