Background Postoperative gastrointestinal leak and venous thromboembolism (VTE) are devastating complications of bariatric surgery. The performance of currently available predictive models for... Show moreBackground Postoperative gastrointestinal leak and venous thromboembolism (VTE) are devastating complications of bariatric surgery. The performance of currently available predictive models for these complications remains wanting, while machine learning has shown promise to improve on traditional modeling approaches. The purpose of this study was to compare the ability of two machine learning strategies, artificial neural networks (ANNs), and gradient boosting machines (XGBs) to conventional models using logistic regression (LR) in predicting leak and VTE after bariatric surgery. Methods ANN, XGB, and LR prediction models for leak and VTE among adults undergoing initial elective weight loss surgery were trained and validated using preoperative data from 2015 to 2017 from Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program database. Data were randomly split into training, validation, and testing populations. Model performance was measured by the area under the receiver operating characteristic curve (AUC) on the testing data for each model. Results The study cohort contained 436,807 patients. The incidences of leak and VTE were 0.70% and 0.46%. ANN (AUC 0.75, 95% CI 0.73-0.78) was the best-performing model for predicting leak, followed by XGB (AUC 0.70, 95% CI 0.68-0.72) and then LR (AUC 0.63, 95% CI 0.61-0.65, p < 0.001 for all comparisons). In detecting VTE, ANN, and XGB, LR achieved similar AUCs of 0.65 (95% CI 0.63-0.68), 0.67 (95% CI 0.64-0.70), and 0.64 (95% CI 0.61-0.66), respectively; the performance difference between XGB and LR was statistically significant (p = 0.001). Conclusions ANN and XGB outperformed traditional LR in predicting leak. These results suggest that ML has the potential to improve risk stratification for bariatric surgery, especially as techniques to extract more granular data from medical records improve. Further studies investigating the merits of machine learning to improve patient selection and risk management in bariatric surgery are warranted. Show less
Arezzo, A.; Migliore, M.; Chiaro, P.; Arolfo, S.; Filippini, C.; Cuonzo, D. di; ... ; REAL Score Collaborators 2019
Background Anastomotic leak after rectal cancer surgery is a severe complication associated with poorer oncologic outcome and quality of life. Preoperative assessment of the risk for anastomotic... Show moreBackground Anastomotic leak after rectal cancer surgery is a severe complication associated with poorer oncologic outcome and quality of life. Preoperative assessment of the risk for anastomotic leak is a key component of surgical planning, including the opportunity to create a defunctioning stoma. Objective The purpose of this study was to identify and quantify the risk factors for anastomotic leak to minimize risk by either not restoring bowel continuity or protecting the anastomosis with a temporary diverting stoma. Methods Potentially relevant studies were identified from the following databases: PubMed, Embase and Cochrane Library. This meta-analysis included studies on transabdominal resection for rectal cancer that reported data about anastomotic leak. The risk for anastomotic leak after rectal cancer surgery was investigated. Preoperative, intraoperative, and postoperative factors were extracted and used to compare anastomotic leak rates. All variables demonstrating a p value < 0.1 in the univariate analysis were entered into a multivariate logistic regression model to determine the risk factors for anastomotic leak. Results Twenty-six centers provided individual data on 9735 patients. Selected preoperative covariates (time before surgery, age, gender, smoking, previous abdominal surgery, BMI, diabetes, ASA, hemoglobin level, TNM classification stage, anastomotic distance) were used as independent factors in a logistic regression model with anastomotic leak as dependent variable. With a threshold value of the receiver operating characteristics (ROC) curve corresponding to 0.0791 in the training set, the area under the ROC curve (AUC) was 0.585 (p < 0.0001). Sensitivity and specificity of the model's probability > 0.0791 to identify anastomotic leak were 79.1% and 32.9%, respectively. Accuracy of the threshold value was confirmed in the validation set with 77.8% sensitivity and 35.2% specificity. Conclusions We trust that, with further refinement using prospective data, this nomogram based on preoperative risk factors may assist surgeons in decision making. The score is now available online (). Show less
Mechanical bowel preparation is a long standing practice in elective open colon surgery dating from the 1970's. It has always been believed to reduce the rate of postoperative complications in the... Show moreMechanical bowel preparation is a long standing practice in elective open colon surgery dating from the 1970's. It has always been believed to reduce the rate of postoperative complications in the form of anastomotic leakage and wound infection. In this thesis we broadly and thoroughly examine the various aspects and the value of mechanical bowel preparation in elective open colon surgery. - Mechanical bowel preparation in elective open colon surgery does not reduce the rate of anastomotic leak or wound infection nor does it reduce peritoneal bacterial contamination. It is however poorly tolerated by patients and the total workload for the nursing staff and health care costs are considerable. Therefore, it is superfluous in elective open colon surgery and should be abandoned. - The intraoperative subcutaneous bacterial swab is a powerful negative predictor of a postoperative wound infection in elective open colon surgery. -Mechanical bowel preparation with polythylene glycol causes no significant additional histopathologic damage of colonic tissue, in comparison with patients receiving a normal meal the night before open colon surgery. - Mechanical bowel preparation with polyethylene glycol does not significantly interfere with colon anastomotic healing in rats. - There is no significant benefit of a single dose of intraluminal 3.0 mMn-butyrate on colon anastomotic healing in rats after polyethylene glycol bowel preparation. Show less
