Enabling personalized perioperative risk prediction by using a machine-learning model based on preoperative data.

Preoperative risk assessment is essential for shared decision-making and adequate perioperative care. Common scores provide limited predictive quality and lack personalized information. The aim of this study was to create an interpretable machine-learning-based model to assess the patient's individual risk of postoperative mortality based on preoperative data to allow analysis of personal risk factors. After ethical approval, a model for prediction of postoperative in-hospital mortality based on preoperative data of 66,846 patients undergoing elective non-cardiac surgery between June 2014 and March 2020 was created with extreme gradient boosting. Model performance and the most relevant parameters were shown using receiver operating characteristic (ROC-) and precision-recall (PR-) curves and importance plots. Individual risks of index patients were presented in waterfall diagrams. The model included 201 features and showed good predictive abilities with an area under receiver operating characteristic (AUROC) curve of 0.95 and an area under precision-recall curve (AUPRC) of 0.109. The feature with the highest information gain was the preoperative order for red packed cell concentrates followed by age and c-reactive protein. Individual risk factors could be identified on patient level. We created a highly accurate and interpretable machine learning model to preoperatively predict the risk of postoperative in-hospital mortality. The algorithm can be used to identify factors susceptible to preoperative optimization measures and to identify risk factors influencing individual patient risk.

[Implementation of discharge management in the surgical department of a university hospital: exploratory analysis of costs, length of stay, and patient satisfaction]. / Einführung des Entlassmanagements an einer Universitätsklinik für Chirurgie: Explorative Analyse von Kosten, Verweildauer und Patientenzufriedenheit.

BACKGROUND: Discharge management has been mandatory by law in Germany since October 2017, and hospitals are required to finance and implement this. Currently there are no data available on the costs and effects of discharge management on the length of hospital stay. AIMS: Determination of the costs of discharge management in the Department of Surgery at the University Hospital rechts der Isar of the Technical University of Munich, Germany, assessment of the length of stay in comparison with and without discharge management, and evaluation of patients' satisfaction to create first precedents for future negotiations about adequate financing. METHODS: Cost analysis of discharge management in the Department of Surgery at the School of Medicine at the Technical University of Munich, retrospective analysis of the mean length of hospital stays before and after implementation of discharge management, and patient surveys on the quality of the structured transition process and their satisfaction. RESULTS: The cost analysis revealed lump costs of € 43 per patient and € 391 for patients with a need for complex management. No statistically significant shorter length of hospital stay after the implementation of discharge management was found by analyzing three patient subgroups. The overall rate of patients returning to the hospital due to complications associated with the surgical procedure was 3.4%. DISCUSSION: Discharge management in the Department of Surgery at the hospital is an effective and potentially quality-enhancing but at the same time cost-driving measure, which, in the medium term, will enter G­DRG rates and may thus increase costs. A possible solution to meet various stakeholders' needs could be a case-specific financial remuneration of discharge management that is adapted to the transition qualities of the various medical departments.