Novel tool utilized as a trigger for advance care planning in hospitalized oncology patients.

22 Background: Study Purpose: reduce inpatient oncology readmissions; determine if there is a need to standardize trigger for Advance Care Planning in the inpatient oncology population. METHODS Investigational Quality Improvement Study January-June 2014. Retrospective chart review of oncology readmissions on the BMC Hematology and Oncology service during a 6 month period, January-June 2013 (N = 68). Data abstracted included: demographics (age, race/ethnicity, primary language, education, marital status, ECOG); type and stage of cancer; type of consults (palliative, spiritual, social services, integrative, PT/OT, hospice, visiting nurse). Developed a Severity of Illness Scale for use in the oncology inpatient patients at BMC. RESULTS Mortality rate within 1 year of discharge: 32.8% (non-elective readmissions) 19/58; mean days to death from readmission discharge: 65 days (range 0-252); 84.2% of those who died had metastatic cancer in index admission; only 37% ever had a palliative care consult placed (all on readmit); only 17% had a documented EOL discussion on index admission; 38.5% of patients had lung CA (22% of all non-elective readmissions). CONCLUSIONS (1) There is a need to standardize triggers to goals of care discussions and access to Palliative Care in the Hematology/Oncology Inpatient Population at this Institution. (2) Readmissions may be reduced with better Advance Care Planning. (3) There is a communication gap amongst interdisciplinary teams in regards to Advance Care Planning at this institution. Currently IRB Approval for use of novel severity of illness scale on a maximum 450 patient sample (current N = 300) at daily rounds administered in all inpatients admitted to the Hematology/Oncology Service January 4, 2016-June 30, 2016. Preliminary data shows 25-30% of study sample scoring in for Advance Care Planning with a score of greater than or equal to 4 on novel tool (BMC Cancer Care Severity Of Illness Tool).