OBJECTIVETo analyze the change in postoperative complication rate after gastric cancer surgery registered in the Peking University Cancer Hospital in recent 11 years and the learning curve of complication registration, and to investigate how to improve the complication registration and evaluation in gastric cancer surgery.

METHODSPatients who underwent open or laparoscopic gastric cancer surgery between April 14, 2005 and February 15, 2016 in our institution were included in the study, and those without essential clinical and administrative data were excluded. Data were biennially collected, and linear regression was performed to investigate the changes of the following parameters, including overall complication rate, severe complication proportion (proportion of complications with Clavien-Dindo score ≥III(a in the total registered complications), re-operation rate and the major complication rate.

RESULTSA total of 5 666 patients were included in the analysis, with 4 111 males (72.56%) and 1 555 females (27.44%). The average age was (58.87±11.50) years and average BMI was(23.15±3.30) kg/m. There were 305 patients included in the 2005-2006 interval, 810 patients in 2007-2008, 957 patients in 2009-2010, 1 163 patients in 2011-2012, 1 421 patients in 2013-2014, and 1 010 patients in 2015-2016, respectively. The overall re-operation rate was 2.34%(133/5 666), postoperative mortality was 0.41%(23/5 666), registered complication rate was 19.66%(1 114/5 666), severe complication proportion was 32.28%(338/1 047), and the proportion of complication missing the Clavien-Dindo score was 6.01%(67/1 114). The linear regression showed the re-operation rate (r=0.13, P=0.801) and postoperative mortality (r=0.58, P=0.231) remained low (< 4% and < 1% respectively) since 2005, and showed no statistical significance. The registered complication rate showed evident increase from 3.93%(12/305) to 29.13%(414/1 421) between 2005 and 2014 (r=0.92, P=0.010), and slight decrease to 22.77%(230/1 010) in 2015-2016. The severe complication proportion significantly decreased from 6/9 in 2005-2006 to 22.73%(50/220) in 2015-2016 (r=0.90, P=0.014). The proportion of complication missing the Clavien-Dindo score significantly decreased from 25.00%(3/12) in 2005-2006 to 4.35%(10/230) in 2015-2016(r=0.82, P=0.044). The most common complications were infection (9.12%, 517 cases), effusions (6.26%, 355 patients), gastrointestinal motility disorder (4.45%, 252 cases), anastomotic leakage (3.19%, 181 cases) and bleeding (2.31%, 131 cases). The registered rates of these complications all increased since 2005, and the rates of leakage and effusions decreased since 2012 while the others decreased after 2014.

CONCLUSIONSAccording to the data from our institution in the recent 11 years, a learning curve exists in our institution for complication registration in gastric cancer surgery. The administrative data appears to be more reliable than registered complication data in quality and safety evaluation during the learning period. A detailed classification with the Clavien-Dindo score aids to the use of complication data for the quality and safety measurement.

Aged ; Anastomotic Leak ; etiology ; Data Collection ; methods ; statistics & numerical data ; Female ; Gastrectomy ; adverse effects ; mortality ; Humans ; Laparoscopy ; adverse effects ; Male ; Medical Records ; statistics & numerical data ; Middle Aged ; Postoperative Complications ; epidemiology ; Registries ; statistics & numerical data ; Reoperation ; statistics & numerical data ; Retrospective Studies ; Stomach Neoplasms ; complications ; surgery

Objective: To explore the feasibility of assessing complications registration through medical information. Methods: A descriptive case series study was performed to retrospectively collect medical information and complication registration information of gastric cancer patients at Department of Gastrointestinal Cancer Center Ward I, Peking University Cancer Hospital from November 1, 2016 to March 1, 2017 (the first period), and from November 1, 2018 to March 1, 2019 (the second period). Case inclusion criteria: (1) adenocarcinoma confirmed by gastroscopy and biopsy; (2) patients undergoing open surgery or laparoscopic radical gastrectomy; (3) complete postoperative medical information and complication information. Patients who were directly transferred to ICU after surgery and underwent emergency surgery were excluded. Because difference of the complication registration procedure at our department existed before and after 2018, so the above two periods were selected to be used for analysis on enrolled patients. The prescription information during hospitalization, including nursing, medication, laboratory examination, transference, surgical advice, etc. were compared with the current Standard Operating Procedure (SOP, including preoperative routine examinations, inspection, perioperative preventive antibiotic use, postoperative observational tests, inspection, routine nutritional support, prophylactic anticoagulation, and prophylactic inhibition of pancreatic enzymes, etc.) for gastric cancer at our department. Medical order beyond SOP was defined as medical order variation. Postoperative complication was diagnosed using the Clavien-Dindo classification criteria, which was divided into I, II, IIIa, IIIb, IVa, IVb, and V. Medical order variation and complication registration information were compared between the two periods, including consistence between medical order variation and complication registration, missing report, underestimation or overestimation of medical order variation, and registration rate of medical order variation [registration rate = (total number of patients–number of missing report patients)/total number of patients], severe complications (Clavien-Dindo classification ≥ III), medical order variation deviating from SOP and the corresponding inferred grading of complication. The data was organized using Microsoft Office Excel 2010. Results: A total of 177 gastric cancer patients were included in the analysis. The first period group and the second period group comprised 89 and 88 cases, respectively. The registrated complication rate was 23.6% (21/89) and 36.4% (32/88), and the incidence of severe complication was 2.2% (2/89) and 4.5% (4/88) in the first and the second period, respectively. The complication rate inferred from medical order variation was 74.2% (66/89) and 78.4% (69/88), and the incidence of severe complication was 7.9% (7/89) and 4.5% (4/88) in the first and second period, respectively. In the first and second period, the proportions of medical order variation in accordance with registered complication were 36.0% and 45.5% respectively; the proportion of underestimation, overestimation and missing report were 5.6% and 4.5%, 4.5% and 4.5%, 53.9% and 45.5%, respectively; the registration rate of medical order variation was 46.1% and 54.5%; the number of case with grade I complications inferred from medical order variation was 34 (38.2%) and 25 (28.4%), respectively; and the number of grade II was 12 (13.5%) and 15 cases (17.0%), respectively. The reason of the missing report of medical order variation corresponding to grade I complication was mainly the single use of analgesic drugs outside SOP, accounting for 76.5% (26/34) and 64.0% (16/25) in the first and second period respectively, and that corresponding to grade II complication was mainly the use of non-prophylactic antibiotics, accounting for 9/12 cases and 5/15 cases, respectively. Conclusions Medical information can evaluate the morbidity of complication feasibly and effectively. Attention should be paid to routine registration to avoid specific missing report.

Anastomotic leakage is one of the most serious complications of colorectal surgery. Despite progress in available surgical techniques, the morbidity associated with anastomotic leakage remains high. In this review, we summarize the current clinical status of this complication, the problems it causes, and relevant research achievements. To date, a lack of consensus regarding the diagnosis of anastomotic leakage has resulted in varying rates of diagnosis across countries and regions worldwide. Accurately predicting the occurrence of anastomotic leakage using the established risk factors and preoperative scoring systems remains difficult. Many of the described preventive measures, including defunctioning stoma creation, positive air leak testing, and use of effective tissue adhesives, remain controversial; more evidence-based medical information is urgently needed. Delayed diagnoses of anastomotic leakage also remain common in clinical practice. To prevent catastrophic outcomes, such as reoperations or deaths, early diagnosis is critically important. Parameters local to the area of the anastomosis may facilitate early detection of leakage, but their effectiveness is subject to clinical validation. Lastly, the pathological etiology of anastomotic leakage remains to be determined, and its elucidation may inspire innovative interventions that solve this critical surgical complication.

Objective To explore the feasibility of early predicting intra-abdominopelvic infectious complications by closer observation of abdominal drainage.Methods Prospectively collected and evaluated the first-three-postoperative-day drainage samples of 207 up-to-standard patients underwent gastrointestinal surgery between April and September 2017 in Peking University Cancer Hospital.Among them,159 males (76.8％) and 48 females (23.2％) were included,the average age was (59.9 ± 10.6) years and the average body mass index was (23.7 ± 3.2) kg/m2.Characteristics of the samples were daily recorded and later on matched with the clinical outcomes,including complications recorded and graded according to the Clavien-Dindo classification.The measurement data with normal distribution were presented as mean and standard deviation,the enumeration data were recorded in the form of quantity and percentage,using the x2 test and Fisher accurate test.Results One hundred and ninty-nine patients (96.1％) were recorded as normal in the drainage samples,while the other 8 cases (3.9％) as abnormal (including pungent odor,purulent color etc.).There were 53 patients (25.6％) were diagnosed with postoperative complications.Among the patients with normal samples,the complication rate was 22.6％,while the eight abnormal ones were all diagnosed with postoperative complications (complication rate:100.00％,P =O.000 012).Notably,among those 8 cases,complications were diagnosed much later than 3 days after surgery in 7 (87.5％) patients.In 1 (12.5％) case,the drainage abnormality was the only abnormal signs and there was no complications detected during the postoperative period of hospitalization.The intra-abdominal infectious complication occurred within 30 days after discharge and the patient re-admitted two times.Conclusions Early change of basic characteristics of postoperative drainage is a promising candidate for detection of postoperative complications with strong specificity.Clinical practice should be further regulated to ensure in-time recording and following interventions of those signs.

Postoperative complications are important outcome measurements for surgical quality and safety control. However, the complication registration has always been problematic due to the lack of definition consensus and the other practical difficulties. This narrative review summarizes the data registry system for single institutional registry, national data registry, international multi-center trial registries in the western world, aiming to share the experience of complication classification and data registration. We interviewed Dr. Koh from Royal Prince Alfred Hospital in Australia for single institutional experience, Dr. van der Wielen and Dr. Desideriofor, from two international multi-center trial(STOMACH) and registry (IMIGASTRIC) respectively, and Prof. Dr. Wijnhoven from the Dutch Upper GI Audit(DUCA). The major questions include which complications are obligated to report in the respective registry, what are the definitions of those complications, who perform the registration, and how are the complications evaluated or classified. Four telephone conferences were initiated to discuss the above-mentioned topics. The DUCA and IMGASTRIC provided the definition of the major complications. The consent definition provided by DUCA was based on the LOW classification which came out after a four-year discussion and consensus meeting among international experts in the according field. However, none of the four registries asked for an obligatory standardization of the diagnostic criteria among the participating centers or surgeons. Instead, all the registries required a detailed recording of the diagnostic strategy and classification of the complications with the Clavien-Dindo scoring system. Most data were registered by surgeons or data managers during or immediately after the hospitalization. The investigators or an independent third party conducted the auditing of the data quality. Standardization of complication diagnosis among different centers is a difficult task, consuming much effort and time. On top of that, standardization of the complication registration is of critical and practical importance. We encourage all centers to register complications with the diagnostic criteria and following intervention. Based on this, the Clavien-Dindo classification can be properly justified, which has been widely accepted by most centers and should be routinely used as the standard evaluation system for postoperative complications in gastric tumor surgery.
Australia ; epidemiology ; Data Collection ; standards ; statistics & numerical data ; Diagnostic Techniques and Procedures ; standards ; statistics & numerical data ; Digestive System Surgical Procedures ; adverse effects ; statistics & numerical data ; Health Care Surveys ; Humans ; Netherlands ; epidemiology ; Postoperative Complications ; classification ; diagnosis ; epidemiology ; Registries ; standards ; Risk Assessment ; methods ; standards ; Stomach Neoplasms ; complications ; surgery