Five years ago, a team at Banner Health, a hospital system in Phoenix, AZ, tried to find a way to identify and monitor patients with sepsis, which causes between 1/3 and 1/2 of all hospital deaths. They tried to develop an algorithm that would pick up sepsis' symptoms at the infection's early stages but ultimately failed. What they instead created, was an algorithm that successfully identified patients that were much much sicker than average, therefore stirring clinicians to them.
Article source: https://www.buzzfeed.com/stephaniemlee/how-a-failed-hospital-algorithm-could-save-lives?utm_term=.xd5j20R5B#.th5WG2bz3
Photo source: By Norbert Kaiser (Self-photographed) [CC BY-SA 2.5 (http://creativecommons.org/licenses/by-sa/2.5)], via Wikimedia Commons