Background: Fox Chase Cancer Center set about to identify opportunities to improve care (OIC) for cancer patients dying in the hospital. The team developed and tested a mortality review decision tree and paradigm that helped to identify OICs. Volunteer members of hospital staff representing most domains of patient care comprised the committee. Monthly meetings defined the review process, developed the form, initiated the program and refined the process as experience accumulated. Methods: Forms incorporated information from Temple University Hospital, Comprehensive Cancer Center Consortium for Quality Improvement(C4QI), and measures to be monitored by CMS. Deaths were considered “anticipated” or “unanticipated,” and with or without OIC. Issues in cases with OIC were recorded for each death. Level 1 review was completed by a research student or a member of the risk management team to organize clinical data. A level 2 review was then performed by a member of the committee and deaths were categorized. Level 3 review was undertaken at monthly meetings and action items determined. Results: From March 2014 to December 2015, 246 cases were reviewed. The elements were chosen for the level 1 & 2 review and collating objective data. Whether the patient was treated with curative intent, procedures were definitive or palliative, if chemotherapy, radiation or surgery were delivered, or if there was a complication within the last 2 weeks of life were recorded. Hospitalization features such as institution of palliative care, appropriate advanced directives, code status on admission and at death, and ICU admission were noted. Issues often arose when deaths were expected and predicted at admission. Increased need for inpatient palliative care teams and hospice support have been recognized. Better communication between patients, families, and oncologists should reduce hospitalization for patients whose death is imminent. Conclusions: Submitting 100% of inpatient deaths to formal mortality review discloses opportunities to improve care and allows focus on problems that occurred even when the patient’s death was expected.