This study examined the effectiveness of three different failure to rescue definitions in evaluating the quality of care among general surgery patients. The authors conclude that the original definition, which uses all deaths rather than specific complications or deaths, provides the best reliability and validity.