withdrawal. Our study has two unique features: first, we observe each patient?s reported satisfaction after consuming a drug … pharmaceutical advertising, allows us to model individual prescription decisions. Second, we distinguish across-patient learning of a … drug?s general efficacy from the within-patient learning of the match between a drug and a patient. Our results suggest …

forward-looking algorithms to fully exploit learning from multiple patients simultaneously. Compared to the best existing …Clinical trials have traditionally followed a fixed design, in which randomization probabilities of patients to various … about the efficacy of treatments. Response-adaptive designs, on the other hand, allow clinicians to use the learning about …

approval in relation to the perceived need for earlier access to new therapies for critically ill patients. Much less in the … news, but lurking behind both problems, is the need for better information for physicians and patients to use in making … information doctors and patients need to make decisions about drugs and the information the government and private insurers need …

patients, but complicate resource allocation. We study the patient enrollment and trial termination decision in a Phase 3 group …