We were seeking to understand the causes of over-aggressive care at the end of life. We identified that most aggressive care was due to over admission of patients (as opposed to ED visits or chemotherapy treatments).
This project involved an analysis of who received over-aggressive at the end of life. We took a mixed methods approach. First we interviewed
6 pancreatic oncology doctors about their definitions, perception of and reasons for over- aggressive end of life care. Next, we extracted EHR data on patient treatment during the last year of life. Initially, we considered this among only pancreatic cancer patients. Since the sample size was small we decided to expand this to breast, prostate, and lung cancer patients as well. One proposed solution is the creating of an acute care oncology clinic. We identified characteristics of aggressive care at the end of life, at Duke, among cancer patients. We are considering expanding this analysis to other disease areas, such as Heart Failure and End Stage Renal Disease. We also plan to incorporate claims information to get a better understanding of receipt of care outside of DUHS. We are also engaged in a collaboration with Performance Services to replicate this project across DUHS (DATE: Dashboard of Aggressive Treatment at the End of life). That project uses the same methodology and leverages Death Masterfile data and North Carolina death certificates to create a 95% complete record of the end-of-life care of patients served by Duke Health. Key metrics in that dashboard will include use of inpatient care and ED visits in the last month of life, and 30-day mortality. That dashboard will be included in reports and action plans of all mortality stakeholder groups (oncology, cardiovascular, med- surg, neurosciences) and will be shared with department chairs and division chiefs.
