Surgical outcomes in gastroenterological surgery in Japan: Report of the National Clinical Database 2011–2019

1The National Clinical Database (NCD) has been recognized as the largest and best‐organized nationwide surgical registry in Japan and has become indispensable for surgeons, patients, and the healthcare system of Japan. The NCD was established in 2010 and started its data registration in 2011.,,, As of January 14, 2021, 5404 facilities have enrolled in the NCD and approximately 1,500,000 cases have been registered every year, constituting more than 95% of all surgical cases in Japan.

2The NCD comprises the members of surgery‐related societies including the Japan Surgical Society (JSS), the Japanese Society of Gastroenterological Surgery (JSGS), and the Japanese Society of Hepato‐Biliary‐Pancreatic Surgery (JSHBPS), and clinical data are stratified into three levels: 1) basic common variables, 2) subspecialty variables such as those in 115 selected gastroenterological surgeries or eight major gastroenterological surgeries, and 3) more specialized variables such as those in the Hepato‐Biliary‐Pancreatic surgery. Data in the gastroenterological section of the NCD are based on the JSGS’s definition of the variables and include 115 gastroenterological operative procedures considered important for the board certification system. Eight major gastroenterological surgeries were selected among these 115 procedures (esophagectomy [ESO], distal/total gastrectomy [DG/TG], right hemicolectomy [RHC], low anterior resection [LAR], hepatectomy [HEP], pancreaticoduodenectomy [PD], and surgery for acute diffuse peritonitis [ADP]); data on detailed variables including preoperative laboratory findings, comorbidities, and postoperative complications were required.,,,

3Gastroenterological surgical procedures are also classified into three groups according to their technical difficulty; low, medium, and high degree of difficulty. Some of the newly approved high‐difficulty procedures, such as laparoscopic major hepatectomy and laparoscopic and robot‐assisted pancreatoduodenectomy, are required to be registered preoperatively in the NCD for health insurance to be authorized by the Ministry of Health, Labour and Welfare in Japan.

4Thus, the importance of the NCD has been increasing rapidly as a clinical database and a means for controlling the quality of new surgical procedures. The NCD is also important as it provides medical staff and societies with rigorously collected data for quality improvement of surgery, feedback about surgical outcomes as risk calculators for morbidities,,,,,,,, and mortality,,,,,,,,, data on the comprehensive surgical landscape via the NCD website,, and data gathered from various clinical studies.

5This report intends to outline the current situation and trends to understand the standpoint of and to elucidate the future directions of gastroenterological surgery in Japan using data from the gastroenterological section of the NCD. Previously, the annual reports of the NCD were published on data from 2011 to 2018, and more than 609,589 cases were newly registered in 2019. We describe the most important findings from the data about gastroenterological surgery in the NCD between 2011 and 2019.

6As previously reported,,, the subjects were patients who collectively underwent the 115 surgical procedures stipulated by the “Training Curriculum for Board‐Certified Surgeons in Gastroenterology” and had their surgical data recorded from 2011 to 2019 in the NCD system. Data of basic and perioperative variables were collected for these cases, as described previously.,, Basic common variables were designed for JSS‐level data, comprising age, gender, preoperative and final diagnosis, date of procedure, surgical procedure (NCD code), operator and assistants, emergent or elective surgery, and participation of anesthesiologists. Subspecialty variables, such as the TNM classification for malignant diseases, Clavien–Dindo (C–D) classification, of postoperative complications, date of discharge or death, and 30‐day and in‐hospital mortality were collected for the 115 selected surgical procedures. Additionally, detailed laboratory data, preoperative comorbidities and functional status, and types of postoperative morbidities were recorded for the eight major gastroenterological procedures defined above.

7Postoperative complications of C–D grade III or greater were defined as severe complications. Anonymous data of the board‐certified gastroenterological surgeons of the JSGS were transferred into the NCD and linked with each procedure to elucidate the participation of board‐certified surgeons.

8Data were extracted in a secure system without external connection and basic statistical analysis were carried out by NCD statistic experts, and the number of surgical cases and the mortality rates related to the selected 115 gastroenterological operative procedures were calculated, as well as those for the eight major operative procedures from 2011 to 2019. The incidence of participation of board‐certified surgeons as the primary surgeons or assistants in the eight major gastroenterological surgeries was also calculated.

9The NCD system is modified annually to adapt to the change in operative procedures, new surgical techniques, or fit the definition and choices of variables better. In 2019, over 50 modifications were made, including the addition of six new variables and the modification of over 40 parts in the definition and other sections of gastroenterological surgery in NCD.

10The following points need to be considered in the interpretation of the data reported here: 1) As a maximum of eight operative procedures can be recorded per case in the NCD, the total number of surgeries in the results describing the 115 gastroenterological surgical procedures for the board certification system does not represent the actual number of surgical cases; 2) Cases with errors in patient age, sex, and postoperative 30‐day status were excluded; 3) Cases in which several operative methods were performed simultaneously were recorded according to all operative methods; 4) Postoperative 30‐day mortality included all cases of mortality within 30 days after surgery, regardless of pre‐ or postdischarge status. The calculation of operative mortality included all patients who died during hospitalization, including hospital stays up to 90 days, and any patient who died after hospital discharge within 30 days of the operative date.

11The total number of cases represented by the 115 selected gastroenterological surgical procedures, recorded in the NCD from January 1 to December 31, 2019, was 609,589, and 5,029,764 cases were registered between 2011 and 2019 (Figure 1). Regarding organ involvement, the stomach and duodenum decreased slightly to 63,160 (10.3% of all 115 procedures) in 2019 from 65,52 (10.8%) in 2018, and the rectum and anus increased slightly to 57,706 (9.5%) in 2019 from 56,162 (9.3%) in 2018. The involvement of other organs (esophagus, small intestine and colon, liver, gallbladder, pancreas, spleen, and other organs) was approximately the same as in 2018 (Table 1).

12Most cases were performed in the certified or related institutions of JSGS and included a notably high number of operations involving the esophagus (certified 94.3%, related 3.8%), liver (certified 89.7%, related 6.8%), and pancreas (certified 91.9%, related 6.2%), indicating that very few cases, fewer than 2–4%, underwent these surgeries in nonrelated or noncertified institutions in Japan.

13Regarding the surgeons’ credentials, it was clear that an increased number of board‐certified surgeons participated in the surgeries; these were in the esophagus (94.2%), stomach and duodenum (83.8%), small intestine and colon (74.0%), liver (94.1%), gallbladder (75.7%), pancreas (95.1%), spleen (86.8%), and other organs (74.0%) (Table 2). Meanwhile, the rate of surgeries by nonboard‐certified surgeons decreased slightly but remained at more than 50% in the gallbladder (67.7%), small intestine and colon (66.8%), stomach and duodenum (53.9%), and rectum and anus (51.1%), while they were below 35% in the liver (33.6%), pancreas (30.8%), and esophagus (23.6%).

14The rate of postoperative complications and 30‐ and 90‐day mortalities are described in Table 3. The rate of complications in the esophagus, stomach and duodenum, pancreas, and spleen increased slightly toward 2019, while the mortality decreased in these organ groups. The rate of complications and 30‐ 90‐day mortality in the rest of the organs were approximately the same as before.

15Among the procedures performed in over 50 cases in 2019 in 115 gastroenterological procedures, those with the highest 90‐day mortality rate were 1) acute pancreatitis surgery (23.2%), 2) esophagus bypass (13.5%), 3) esophageal fistula construction (12.8%), 4) gastrointestinal perforation surgery (11.9%), 5) ADP surgery (11.4%), 6) total colectomy (11.7%), 7) gastric fistula construction (excluding PEG) (11.6%), 8) external cholecystectomy (10.6%), and 9) hepatorrhaphy (10.0%). Among these nine procedures, the degree of difficulty was high in one procedure (esophagus bypass) and either moderate or low in the other eight Tables S1‐1–9).

16The number of surgeries carried out annually for the eight major operative procedures, the percentage by gender, and the percentage according to age group between 2011 and 2019 are shown in Table 4 (Figure 2).

17The number of procedures increased in RHC, LAR, and PD and decreased in DG, TG, and HEP.

18The age distributions mirrored the tendency in recent years to shift toward older patients, and the percentage of cases with the age >70 years in 2019 (increase from 2011) was 48.3% (+12.0%), 64.0% (+12.3%), 63.5% (+10.8%), 69.5% (+7.1%), 46.7% (+5.9%), 57.3% (+10.2%), 60.3% (+9.7%), 55.4% (+7.6%) in ESO, DG, TG, RHC, LAR, HEP, PD, and ADP, respectively.

19The rate of involvement of board‐certified surgeons for the 115 procedures remained the same as in 2018, and the rates have been continuously increasing in recent years; these were approximately as follows: ESO (96.4%), HEP (96.3%), and PD (95.5%), and the lower rates ranged from 73.3–86.8% in DG/TG, RHC, LAR, and ADP (Table 5).

20The morbidities and mortalities of these eight major gastroenterological procedures remained relatively constant in all procedures. The morbidities increased in ESO, HEP, and PD compared to those in 2016 or 2017; in contrast, the mortalities decreased in these procedures. In the other five procedures, the rate of morbidities and mortalities remained approximately the same from 2011 to 2019 (Table 6).

21The rate of laparoscopic surgery in 2019 continuously increased from 2011 in LAR (70.3%, +40.8%), ESO (66.8%, +34.8%), RHC (52.5%, +25.4%), DG (51.9%, +20.3%), TG (27.5%, +15.4%), HEP (nonlateral segment) (12.9%, +9.6%), and ADP (21.2%, +14.9%), with the exception of PD, which remained low at 2.6% (+1.8%) (Table 7, Figure 3).

22Data of gastroenterological surgery in Japan using the gastroenterological section of the NCD were summarized, and the trends in the 115 gastroenterological procedures and eight major gastroenterological surgeries were reported. The numbers, demography, morbidities, and mortalities of the procedures comprised the main contents of this report, summarized as follows: 1) Operative numbers gradually increased in all procedures except for gastrectomy and hepatectomy, which decreased in these years; 2) age distributions shifted toward older patients in all eight major gastroenterological surgeries; 3) morbidities of ESO, HEP, and PD increased, but the mortalities were minimized in all procedures; 4) laparoscopic procedures have been increasing in all eight major gastroenterological procedures; and 5) the involvement of board‐certified surgeons increased. These trends in recent years were more prominent in 2019.

23The 30‐ and 90‐day mortalities have been steadily decreasing in these years, while the postoperative morbidities classified as C–D grade III or higher remained almost unchanged in the gallbladder, rectum, and anus (LAR) and ADP surgery and increased slightly in the esophagus (ESO), stomach and duodenum (DG/TG), small intestine and colon (RHC), liver (HEP), and pancreas (PD). The reason for the discrepancy between postoperative morbidities and mortalities was unclear, but possible reasons include an improved “failure to rescue ratio,”, reduction in severe morbidities with higher mortality risk, and more accurate registration of postoperative morbidities into the NCD system than before. These possibilities should be further evaluated in future studies.

24The rate of laparoscopic surgeries, including robotic surgeries, demonstrated their rapid increase, reflecting the current trends in gastroenterological surgeries, although the usage rate of laparoscopy varied greatly. Among the major GE procedures, LAR and ESO had the highest rate (more than two‐thirds), while PD had the lowest laparoscopic rate of only 2.6%. Detailed analyses including the safety and other outcomes of laparoscopic procedures should be performed in future studies.

25Seven of the eight major gastroenterological surgeries, apart from ADP, can be divided into two categories: 1) basic gastroenterological surgeries (DG/TG, RHC, and LAR) and 2) advanced gastroenterological surgeries (ESO, HEP, and PD). Postoperative complications and mortalities were higher, and the rate of involvement of board‐certified surgeons as the primary surgeons was higher in advanced than in basic gastroenterological surgeries. It is not difficult to imagine that young, noncertified surgeons initially performed basic surgeries and improved their skills, thereafter proceeding to advanced surgeries. Further detailed analyses of the NCD data would clarify this point in the future.

26Among the procedures performed in over 50 cases in 2019 in 115 gastroenterological procedures, nine procedures had a mortality of over 10%, and most of these were low‐difficulty procedures, indicating that the causes of mortality were more likely not technical problems but rather the poor general conditions of the patients. In this report, the NCD data between 2011 and 2019 were utilized. It is important to maintain accurate clinical data for analysis and interpretation. To ensure accuracy, systematic audits comparing the registered data and corresponding clinical charts in 20 cases per facility were initiated, and members of the JSGS committee and its subcommittee for NCD database quality improvement investigated 5% of all JSGS‐certified hospitals (about 40 facilities), which were randomly selected every year. This will be continued by the JSGS database committee. The reports of these audits have so far revealed the high accuracy of the NCD data in the JSGS section.,

27The usage of the NCD for clinical studies has been also expanding. JSGS leads the joint council for the society‐specific registries since 2009 and has promoted clinical studies using the NCD. The joint council currently consists of 16 societies as of January 2021; JSGS, JSS, JSHBPS, The Japan Esophageal Society, Japanese Gastric Cancer Association, Japan Pancreas Society, Japan Society for Endoscopic Surgery, Japanese Society of Abdominal Emergency Medicine, Japanese Hernia Society, The Japanese Society for Treatment of Obesity, The Japanese Association for Thoracic Surgery, Japanese Liver Transplantation Society, The Japanese Society for Cancer of the Colon and Rectum, Liver Cancer Study Group of Japan, Endoscopic Liver Surgery Study Group, and Japanese Society for Endoscopic and Robotic Pancreatic Surgery.

28The JSGS approved 77 studies from 2013 to 2020 with regard to the joint council, and many high‐impact articles have been published so far. Fifteen articles were accepted and published in 2020, where preoperative risk models for postgastrectomy intraabdominal infectious complications related to gastric cancer, morbidities after total pancreatectomy, bile leakage after hepatectomies for hepatocellular carcinoma, and emergency surgery for gastrointestinal cancer were reported. The importance of board‐certified surgeons was reported in RHC and PD. The Endoscopic Surgical Skill Qualification System certification by the Japan Society for Endoscopic Surgery did not affect the postoperative mortality following laparoscopic DG and LAR. Additionally, hospital volume affected postoperative mortality after TG and PD, and laparoscopic liver resection was safely developed with a low mortality and complication rate relative to open liver resection in Japan. A geriatric surgery pilot study was conducted from 2017 to 2020, and the specific variables and outcome predictors in geriatric surgery were implemented in the NCD system in 2021. Thus, in the NCD, a robust nationwide registry on surgical outcomes is important to elucidate the performance of surgeries, to provide tools for future studies, and to improve the surgical outcomes.

29The database itself is only a result of the clinical treatment, but it is important for establishing the four pillars of surgical quality improvement that the American College of Surgeons–National Surgical Quality Improvement Program (ACS‐NSQIP) has identified. These four pillars are setting standards, creating the infrastructure required to achieve these standards, commitment to measuring performance against those standards and remaining accountable for those measurements, and agreeing to a peer review against those standards. The JSGS and ACS‐NSQIP have collaborated since the foundation of the NCD in 2010, and collaborative studies are ongoing., Rigorous data collection is required with respect to the third pillar. While the other pillars are important, establishing standards could be the essential step.

30Besides the data presented in this annual report, many other variables were also available, and future studies are expected to elucidate the current situation and implications for the future. Further active clinical studies will discover new evidence using the assets of the NCD data, which all surgeons, medical staff, and surgical clinical reviewers contributed to, in most facilities in Japan. We continuously take care to promote the value of the database and to encourage the usage of feedback and clinical studies using the NCD now and in the future.

31The protocol for this research project has been approved by the Ethics Committee of the NCD as of November 18, 2020, and it conforms to the provisions of the Declaration of Helsinki as revised in Fortaleza, Brazil, October 2013. The opt‐out method to obtain patient consent was utilized at each institution.

32Funding: The department is a social collaboration department supported by grants from the National Clinical Database, Johnson & Johnson KK, and Nipro Co.

33CONFLICT OF INTEREST

34Arata Takahashi, Hiroyuki Yamamoto, and Hiroaki Miyata are affiliated with the Department of Healthcare Quality Assessment at the University of Tokyo. The other authors have no conflicts of interest.