Background Significant resources have been allocated to decreasing the number of preventable deaths in hospitals, but identifying preventable factors and then leveraging them to effect system-wide change remains challenging. Objective To determine the ability of a novel in-person, multidisciplinary “rapid mortality review” process to identify deaths that are preventable and action items that lead to improvements in care. Methods Rapid mortality review sessions were conducted weekly for patients who died in the medical intensive care unit. Patient data and clinician opinions regarding preventable deaths were discussed and recorded. Bivariate analyses were done to detect associations between case variables and the formation of an action item. Results From 2013 to 2018, 542 patient deaths were reviewed; of those, 36 deaths (7%) were deemed potentially preventable. Facilitators identified issues in 294 cases (54%). A total of 253 action items were identified for 175 cases (32%); 60% of those action items were subsequently completed and led to tangible systemic change in 29 instances (11%). Action items were more likely to be identified for patients who had not been receiving comfort care (P < .001), for patients who had received cardiopulmonary resuscitation (P < .001), when the treatment team (P < .001) or the rapid mortality review facilitator (P < .001) had care-related concerns, and when the patient’s death had been preventable (P < .001). Conclusions Even in settings with low reported rates of preventable deaths, an in-person multidisciplinary mortality review can successfully identify areas where care can be improved, leading to systemic change.