No abstract

physicians seldom document these discussions. We sought to design and pilot test an intervention that increases the rate at which emergency physicians document goals of care discussions among end-of-life patients admitted to the intensive care unit. Methods: In an emergency department with annual census of approximately 88,000 patients at a tertiary urban academic medical center, electronic medical record review of intensive care unit admissions by a physician champion revealed a very low rate of emergency physician goals of care documentation. The physician champion and emergency department co-chairmen developed an intervention. From June 2015 through October 2016 the champion reviewed emergency physician documentation among intensive care unit admissions who appeared to be near the end-of-life, and submitted reports to the co-chairmen for further review. Severe trauma, acute stroke and acute coronary syndrome patients were excluded since consultants typically have these goals of care discussions in the emergency department. End of life was defined as patients with a reviewer-estimated life expectancy of six months or less, and categorized according to previously described trajectories of dying: (1) advanced cancer, (2) bed bound from severe neurologic disease such as dementia or stroke, or (3) advanced organ system failure. Emergency physician-documented review of pre-existing goals of care documentation was in cases where patients lacked capacity and no surrogate was available. Reviewers e-mailed positive feedback to emergency physicians for documenting goals of care. Co-chairmen performed academic detailing per their discretion when goals of care were not documented. Results: We reviewed 1286 intensive care unit admissions and identified 151 patients who appeared to be near the end of life. Seventy-eight (52%) had advanced cancer, 43 (28%) were bed bound from severe neurologic disease, 26 (17%) had advanced organ system failure, and 4 (3%) were severely disabled due to other conditions. Ninety-five (63%) of the end-of-life patients died, including 80 (52%) deaths within 6 months and 55 (36%) during the same admission. Mortality could not be determined for the remaining 56 patients through electronic medical record review alone, but 9 of these were discharged on hospice. Emergency physicians were receptive to the intervention. The emergency physician goals of care discussion documentation rate rose from 1 of 16 (6%) end-of-life patients in June 2015 to 9 of 17 (52%) in October 2016. Conclusions: We found it feasible to implement a quality improvement intervention that promotes emergency physician goals of care documentation for intensive care unit admissions appearing to be near the end of life. We plan to conduct a linear time-series analysis to formally evaluate effectiveness.