A culture change initiative to reduce end-of-life systemic anticancer therapy and intensive healthcare utilization.

Abstract

607 Background: Receipt of systemic anticancer therapy (SACT) near the end of life (EOL) is associated with potentially low-value care, including increased emergency department visits, hospitalizations, and intensive care unit (ICU) utilization and lower hospice utilization. At UT MD Anderson, a large freestanding NCI-designated comprehensive cancer center, we implemented a division-wide culture change initiative aimed at promoting reflection on treatment decision-making and counseling for patients who died within 14 days of SACT administration. Methods: Beginning in May of 2024, a multipronged strategy was employed to create a culture of reflection around administration of EOL SACT across 11 disease-site specific medical oncology departments within the division of Cancer Medicine. Phases of the project included listening sessions to understand drivers of EOL SACT administration, chart reviews to assess quality of pre-treatment counseling, ongoing disease site-specific department-led mortality reviews, and implementation of disease-specific oversight of high-risk chemotherapy administration. The primary end point was changes in rate of EOL SACT and secondary endpoint was changes in EOL ICU utilization. We used a time-trend analysis to assess level and slope changes following intervention implementation on data obtained from institutional dashboards modeled on Vizient methodology before and after May 2024. Results: During the preintervention period, EOL SACT rates declined at an annualized rate of 9% (p=0.033). Following intervention implementation, the annualized decline accelerated to 16% (p<0.0001). The cumulative 12-month EOL SACT decreased from 169 patients (0.12% of unique oncology patients seen in the period) in May 2024 to 115 patients (0.07%) in April 2026. EOL ICU utilization was stable in the preintervention period and demonstrated a 22% annual decrease (p<.00001) after the intervention began. Twelve-month cumulative ICU utilization within 48 hours of death decreased from 396 patients (0.28%) in May 2024 to 272 (0.18%) in April 2026. Conclusions: A division-wide culture change initiative that empowered disease-specific departments and focused on treatment decision-making, end-of-life counseling, ongoing mortality reviews and oversight of high-risk SACT administration was associated with significant reductions in both EOL SACT use and ICU utilization. As EOL SACT is associated with increased healthcare use, the decline in ICU utilization is a promising early outcome metric. Further follow up will be needed to determine impact of these changes on rates of EOL ED visits, hospitalization, hospice use and inpatient death.