Enthusiasm for rapid-learning health systems exceeds the current standards for conducting it

Critics of the comparative effectiveness research (CER) enterprise believe that it will proceed too slowly and at too high a cost to transform healthcare practice in the short term. The proposed alternative is a ‘rapid-learning (RL) health system’ (RLHS), a concept whose genesis was continuous quality improvement [1]. RL would not replace traditional CER studies, such as pragmatic randomized trials or observational cohort studies, but would aim for quick learning using the growing data infrastructure to impact either clinical care or coverage/payment decisions. RL is iterative, conducted in real time and employs a rapid ‘plan, do, check and act’ cycle to address real-world knowledge gaps [2].