Using root cause analysis to improve survival in a liver transplant program

Abstract

Despite continuous total quality management (TQM) and traditional weekly morbidity and mortality (M&M) conferences, our liver transplant program survival rates became lower than expected according to national benchmarking standards. Data from the U.S. Scientific Registry of Transplant Recipients (SRTR), the organization contracted by the federal government to manage statistics reported by all U.S. transplant centers, showed that during 7/1/98 TO 6/30/00, our 1-year graft (76.86%, P = 0.230) and patient (80.61%, P = 0.016) survival was lower than expected compared with national expected rates (graft 81.89% and patient 88.30%). In response, our program added root cause analysis (RCA) to our quality improvement process. RCA is a method of identifying causal factors that underlie variation in performance. Using RCA, two of our liver transplant surgeons performed a systematic review of all liver transplant patient deaths in our center from 1/95 TO 8/00 (86 of 372 patients transplanted). All phases of the transplant process, including recipient and donor selection, transplant procedure, and follow-up care (including psychosocial issues) were reviewed to determine specific events from each phase that led to an adverse outcome. For the 86 deaths, 162 root causes were identified. The apportionment was as follows: recipient and donor selection, 58%; transplant procedure, 24%; follow-up care, 18%. The top four root causes were obesity, surgical anatomy issues, pulmonary events, and cardiac events, all relating to patient selection. Multiple root causes in the cases reviewed led to futile liver transplants. Early in 2001, our program conducted in-services and instituted protocol changes according to RCA findings. In April 2004, SRTR data revealed that for patients transplanted between 1/01/01 and 6/30/03, our 1-year liver graft survival of 90.73% is now significantly higher ( P = 0.018) compared to the national expected rate of 84.48%. Our 1-year patient survival rate of 92.66% is higher than the expected rate of 89.29%, although not significantly ( P = 0.285). In conclusion, periodic RCA of adverse events should be added to the TQM efforts and M&M conferences of programs encompassing multiple medical services.