Background and Objectives: With the current advanced data-driven approach to health care, machine learning is gaining more interest. The current study investigates the added value of machine... Show moreBackground and Objectives: With the current advanced data-driven approach to health care, machine learning is gaining more interest. The current study investigates the added value of machine learning to linear regression in predicting anastomotic leakage and pulmonary complications after upper gastrointestinal cancer surgery. Methods: All patients in the Dutch Upper Gastrointestinal Cancer Audit undergoing curatively intended esophageal or gastric cancer surgeries from 2011 to 2017 were included. Anastomotic leakage was defined as any clinically or radiologically proven anastomotic leakage. Pulmonary complications entailed: pneumonia, pleural effusion, respiratory failure, pneumothorax, and/or acute respiratory distress syndrome. Different machine learning models were tested. Nomograms were constructed using Least Absolute Shrinkage and Selection Operator. Results: Between 2011 and 2017, 4228 patients underwent surgical resection for esophageal cancer, of which 18% developed anastomotic leakage and 30% a pulmonary complication. Of the 2199 patients with surgical resection for gastric cancer, 7% developed anastomotic leakage and 15% a pulmonary complication. In all cases, linear regression had the highest predictive value with the area under the curves varying between 61.9 and 68.0, but the difference with machine learning models did not reach statistical significance. Conclusion: Machine learning models can predict postoperative complications in upper gastrointestinal cancer surgery, but they do not outperform the current gold standard, linear regression Show less
Czepiel, J.; Krutova, M.; Mizrahi, A.; Khanafer, N.; Enoch, D.A.; Patyi, M.; ... ; Garlicki, A. 2021
We aimed to describe the clinical presentation, treatment, outcome and report on factors associated with mortality over a 90-day period in Clostridioides difficile infection (CDI). Descriptive,... Show moreWe aimed to describe the clinical presentation, treatment, outcome and report on factors associated with mortality over a 90-day period in Clostridioides difficile infection (CDI). Descriptive, univariate, and multivariate regression analyses were performed on data collected in a retrospective case-control study conducted in nine hospitals from seven European countries. A total of 624 patients were included, of which 415 were deceased (cases) and 209 were still alive 90 days after a CDI diagnosis (controls). The most common antibiotics used previously in both groups were beta-lactams; previous exposure to fluoroquinolones was significantly (p = 0.0004) greater in deceased patients. Multivariate logistic regression showed that the factors independently related with death during CDI were older age, inadequate CDI therapy, cachexia, malignancy, Charlson Index, long-term care, elevated white blood cell count (WBC), C-reactive protein (CRP), bacteraemia, complications, and cognitive impairment. In addition, older age, higher levels of WBC, neutrophil, CRP or creatinine, the presence of malignancy, cognitive impairment, and complications were strongly correlated with shortening the time from CDI diagnosis to death. CDI prevention should be primarily focused on hospitalised elderly people receiving antibiotics. WBC, neutrophil count, CRP, creatinine, albumin and lactate levels should be tested in every hospitalised patient treated for CDI to assess the risk of a fatal outcome. Show less
