Defining high volume center for minimally invasive distal pancreatectomy

Abstract

Presenter: Roheena Panni MD, MPHS | Washington University, St. Louis Background: Minimally Invasive Distal Pancreatectomy (MIDP) is associated with reduced intra-operative blood loss, transfusion requirement, and shorter length of stay compared to open distal pancreatectomy (ODP). However, its advantages over OPD and oncologic adequacy are unclear, due to the small number of centers performing MIDP and strict patient selection criteria. Several studies have outlined the relationship between hospital volume and postoperative mortality for patients undergoing pancreatic surgery, but the exact effect of centralization of care for MIDP still needs to be determined. The purpose of this study is to evaluate the association between hospital procedure volume and mortality for patients undergoing MIDP using a large database to determine an evidence-based threshold of hospital volume associated with improvement in mortality. Methods: Patients who underwent MIDP were identified using the National Cancer Database from 2010 to 2015. The relationship between hospital volume and 90-day mortality was assessed using a logistic regression model. Logistic regression analysis and restricted cubic spline regression analysis was performed to determine the linear and non-linear association between mean hospital volume and mean 90-day mortality. Results: 2,837 patients underwent distal pancreatectomy at 487 different hospitals. 30- and 90-day mortality of the study population were 1.27% (n=36) and 2.54% (n=72), respectively. Baseline characteristics and mean annual mortality of individual hospitals were determined (fig1). A logistic regression analysis was performed, which demonstrated that institutional volume is significantly associated with decreased overall 90-day mortality. The maximum improvement in 90-day mortality is seen if the annual hospital volume was greater than 7 (p<0.0001). We further explored the non-linear association between institutional volume and 90-day mortality, which demonstrated continued improvement in 90-day mortality with an increase in average hospital volume. Conclusion: Our data suggest that the centralization of MIDP is associated with lower postoperative mortality. Based on the results, we recommend defining a high volume center as hospitals performing eight or more MIDP cases per year. The true impact of this finding on overall survival should be assessed in future studies using large databases with long-term follow-up information.