LAUSR

O ver the past 4 decades, the mortality following pancreatoduodenectomy has been reduced by an order of magnitude. Multiple advances have played a role in this achievement. One important factor is the appreciation that outcomes are best at high-volume centers in the hands of high-volume surgeons. As a result, in many countries the majority of pancreatoduodenectomies are now being performed at a relatively small number of referral centers. On the contrary, serious morbidity, including organ space infections (OSIs) and clinically relevant postoperative pancreatic fistulas (CR-POPF), remains unacceptably high. In recent years, considerable advances have been made to standardize the definitions and severity of postoperative complications. With respect to pancreatic surgery, International Study Groups have clearly defined pancreatic fistulas, delayed gastric emptying, and postpancreatectomy hemorrhage. However, methodologies describing optimal outcomes in pancreatic surgery have been lacking. Thus, the paper entitled ‘‘Benchmarks in Pancreatic Surgery’’ by Sánchez-Velázquez et al in this issue of Annals of Surgery is a welcome addition to the literature. Of note, this report is an extension of prior efforts from investigators at the University of Zurich to benchmark liver surgery, esophagectomy, and liver transplantation. ‘‘Benchmarks in Pancreatic Surgery’’ leverages prospectively collected data on 6186 pancreatoduodenectomies performed at 23 high-volume centers on 3 continents over a 4-year period. The authors define 20 ‘‘benchmark cutoffs’’ including 2 operative outcomes (operative time and transfusions), 10 complications including pancreatic fistulas, mortality, failure-to-rescue, length of stay, and readmissions, and 4 oncologic outcomes including disease-free survival (DFS). The 6month follow-up with respect to complications and data on 3-year DFS are clear strengths of the analysis. Differences across 23 centers and the 3 continents with respect to the percentage of low risk ‘‘benchmark’’ patients and pancreatic fistulas (Asia>Europe>the United States) are new and interesting observations. On the other hand, the ‘‘benchmark cutoff’’ defined as the ‘‘75th percentile of medians from 23 centers’’ is not intuitive and leads to some benchmarks which seem less than optimal. For example, the ‘‘benchmark cutoff’’ for operative time is 7.5 hours, a time that has been associated with increased morbidity and mortality. Other benchmarks such as length of stay (16 d) or R1 resection rate (38%) vary dramatically by country and continent. Also, only 224 of the patients (3.6%) at 3 centers underwent robotic or laparoscopic surgery. As a result, this subanalysis was underpowered to draw robust conclusions. Furthermore, potentially predictive variables such as gland texture, duct size, pylorus preservation, pancreatic anastomotic technique, drain management, and pathology were not utilized in the analysis. Nevertheless, benchmarking produces composite measures which have the potential to provide meaningful comparisons among various patient cohorts, surgical approaches, centers, countries, and continents as well as changes over time. Over the past decade, the concept that composite measures of surgical quality may be better than individual outcomes has been championed by several outcome experts. For example, in 2009 Dimick et al reported that the combination of hospital volume and mortality should be employed to help patients and payers identify low-mortality hospitals for major surgery. Similarly, in 2014 Dimick et al recommend combining serious complications, hospital and surgeon volume, reoperations, and readmissions to profile hospital performance with bariatric surgery. In addition, investigators utilizing the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) frequently utilize serious morbidity, which includes multiple complications, or death and serious morbidity (DSM) as composite measures. Textbook Outcome is another composite concept that has emerged in recent years. Textbook Outcomes (TOs) are ‘‘all or none’’ composite measures described in 2013 by Dutch colorectal surgeons. A ‘‘textbook outcome’’ is accomplished when all of the measured variables following an operation are achieved during and after the hospitalization. TOs have been reported for aneurysm 12 13 14