Surgical outcomes in gastroenterological surgery in Japan: Report of National Clinical database 2011–2016

Abstract The National Clinical Database (NCD) of Japan started its registration in 2011 and over 9 000 000 cases from more than 5000 facilities were registered over a 6‐year period. This is the report of NCD based upon gastrointestinal surgery information in excess of 3 200 000 cases from 2011 to 2016 adding data of complications. About 70% of all gastrointestinal surgeries were carried out at certified institutions, and the percentage of surgeries done at certified institutions was particularly high for the esophagus (92.4% in 2016), liver (88.4%), pancreas (89.8%), and spleen (86.8%). The percentage of anesthesiologist participation was more than 90% for almost all organs, except 85.7% for the rectum and anus. Approximately, more than two‐thirds of the surgeries were carried out with the participation of a board‐certified surgeon. Although patients have been getting older, mortalities have not been increasing. There were differences in the incidence of complications according to organ site and procedure. Remarkably, mortality rates of low anterior resection were very low, and those of hepatectomy and acute diffuse peritonitis surgery have been gradually decreasing. Although the complication rates were gradually increasing for esophagectomy or pancreaticoduodenectomy, the mortality rates for these procedures were decreasing. Nationwide, this database is expected to ensure the quality of the board‐certification system and surgical outcomes in gastroenterological surgery.

T A B L E 1 Annual changes of surgeries by sex, age group, and organ for the 115 selected gastrointestinal operative procedures in the training curriculum for board-certified surgeons in gastroenterology representing the performance of surgery in each specialty (esophagectomy, distal gastrectomy, total gastrectomy, right hemicolectomy, low anterior resection, hepatectomy, pancreatoduodenectomy, and surgery for acute diffuse peritonitis). Risk models of mortality for each procedure were created using approximately 120 000 cases registered in 2011, and each model has been accepted and published in peerreviewed journals. [3][4][5][6][7][8][9][10] All reports were the first-risk stratification studies, based on a Japanese nationwide Web-based database. Mortality rates were almost satisfactory compared to those in the Western world. In case of esophagectomy, risk models may not be markedly influenced by choice of open or laparoscopic esophagectomy. 3 The 30-day mortality may underestimate the true risk for death, and operative mortality is recommended as a standard outcome measure after colorectal surgery. 7 As for acute diffuse peritonitis, 38.7% of the 8482 patients were admitted to a hospital by direct ambulance transport. 10 Based on these studies, we can use a real-time feedback system, which includes a risk calculator for the mortality (predicted postoperative 30-day mortality and operative mortality) of preoperative patients and performance reports of each participating hospital. 11 The latter shows each facility's severity-adjusted clinical performance (benchmark) in comparison with the national data and the risk-adjusted cumulative expected-observed death. Better or worse outcomes can be detected by the monitoring report. Furthermore, we are proceeding with papers on complications related to each of the eight operative methods for the evaluation of medical standards using data from 2011 and 2012. [12][13][14][15][16][17] To assure collection of high-quality data, the Japanese

| SUBJECTS AN D METHODS
Subjects were surgical data recorded in the NCD, which were stipulated by the "Training Curriculum for Board-Certified Surgeons in Gastroenterology", using the "New classification of surgical difficulty". The board- We also clarified the changes over time in the number of surgical cases and mortality rates related to the eight main operative procedures from 2011 to 2016. We also comparatively studied patient sex, age groups, institution groups, and percentage contribution of certified surgeons related to the eight main operative procedures.
The following points need to be considered in the interpretation of the data reported here.  certified surgeons that were operators was high for the esophagus (70.0% in 2016), liver (59.6%), and pancreas (62.4%). The total number of recorded surgeries increased each year (Figure 1). Postoperative complications, 30-day mortality rates, and operative mortality rates are shown in Table 3. Complication rates were comparatively higher for the esophagus and the pancreas; however, the mortality rates for these organ procedures were not so high. Figure 1 shows

| Eight main operative procedures
The respective number of surgeries carried out annually for the eight main operative procedures, mortalities and complications between 2011 and 2016 are shown in Table 13 and Figure 2. Subsequently, the male : female ratio leaned toward males for all operative methods, with males particularly predominant with esophagectomy, gastrectomy (distal and total), and hepatectomy. In addition, the percentage of those patients who were ≥80 years was high for gastrectomy (distal and total), right hemicolectomy, and acute diffuse peritonitis surgery (Table 13). Regarding the institution groups in which surgeries were carried out, more than 70% of the surgeries  (Table 14). Increase in the incidence of endoscopic surgery is shown in

| DISCUSSION
Since the start of NCD registration in 2011, surgeons in Japan, especially JSGS members, have constructed a robust nationwide database. We can see the real clinical status of surgical outcomes in Japan. The number of registered surgeries has been increasing year by year. Mortality rates for all of the procedures seem to be acceptable as a nationwide outcome, as they are satisfactorily lower than those reported from other countries. 24,25 These results may be explained by the high participation rate of board-certified surgeons.
Board-certified surgeons in gastroenterological surgery contribute to favorable outcomes in Japan. 21 32 It is necessary to analyze with explanations the tendency of surgical outcomes over time. A risk-adjusted analysis based on nationwide data allows personnel to establish and provide feedback on the risks that patients face before undergoing a procedure. 11 The NCD also provides data on each facility's severity-adjusted clinical performance (benchmark), which can be compared with national data. We can trace periodically where we are in the national standard.
Thinking of the future development of NCD, long-term clinical outcomes will be demanded, especially in cancer registries. The NCD generalizes site-specific cancer registries by taking advantage of their excellent organizing ability. Some site-specific cancer registries, including pancreatic, breast, and liver cancer registries have already been combined with the NCD. 33 Aggregation of the cancer registration system and NCD would definitely produce a novel and important database, not only in the field of clinical medicine but also public health. Another possible linkage to NCD is the medical insurance database including diagnosis procedure combination (DPC) data, which includes not only clinical information on disease but also the medical costs by disease or treatment. 34 After the first stage of the establishment of the national database, NCD has been proceeding to the second stage, development and utilization. Many studies are in progress to improve quality control of surgical procedures using the NCD. Future evolution of the NCD will be promising with impacts to the public.