Key result
Logistic regression outperforms other machine learning algorithms in predicting prolonged CABG stays with ~0.76 AUC.
Why the study?
Prolonged hospitalization after CABG increases healthcare resource use and reflects poor recovery, necessitating early identification of high-risk patients to improve management and resource allocation.
Observational (n=572)
No
Effect estimate: AUC 0.759 (95% CI 0.669-0.850)
A machine learning-based logistic regression model using early perioperative variables can effectively predict prolonged hospital stay after CABG, providing a tool to optimize resource allocation and postoperative management.
May support preoperative risk stratification in CABG; hypothesis-generating and requires prospective validation before adoption.
Prolonged postoperative hospitalization after coronary artery bypass grafting (CABG) increases healthcare resource use and reflects poor recovery. Early identification of high-risk patients may improve perioperative management and resource allocation. This study aimed to develop and validate a machine learning-based model for predicting prolonged hospital stay after CABG. A total of 572 patients who underwent CABG at the Second Xiangya Hospital of Central South University between January 2023 and March 2024 were retrospectively enrolled and randomly divided into training and validation sets (7:3). Prolonged hospitalization was defined as length of stay above the 70th percentile. Feature selection was performed using Boruta and LASSO regression, with multicollinearity assessed by variance inflation factor. Ten machine learning models were constructed and compared. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), accuracy, positive predictive value, recall, F1 score and specificity, with integrated ranking to determine the optimal model. Decision curve analysis, calibration curves, and the Brier score were used to assess clinical utility and predictive accuracy. Shapley Additive Explanations (SHAP) analysis was applied to interpret the Logistic Regression model. Finally, a web-based prediction tool was developed based on the optimal model. Logistic regression showed the most stable validation performance (AUC = 0.759, 95%CI: 0.669–0.850), with good calibration and net clinical benefit. SHAP analysis identified preoperative myoglobin, mechanical ventilation duration, preoperative prothrombin time, and activated partial thromboplastin time as the main predictors. A nomogram and online tool enabled individualized risk assessment. A Logistic Regression model for predicting prolonged hospital stay after CABG was developed and validated in this study. The nomogram and web-based tool provide an initial approach for identifying patients at high risk of prolonged hospital stay during the early postoperative period. Further prospective and external validation studies are needed to confirm the validity and clinical applicability of this tool.
No takes yet. Share an insight, caveat, or question.
Shuai et al. (2026) conducted an observational in Coronary artery disease undergoing coronary artery bypass grafting (n=572). Logistic Regression prediction model vs. Other machine learning models was evaluated on Prolonged length of hospital stay (>21 days) (AUC 0.759, 95% CI 0.669-0.850). A logistic regression model effectively predicted prolonged hospital stay after coronary artery bypass grafting with an AUC of 0.759, outperforming other machine learning algorithms.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: