Background:Early bacterial infections are a common and clinically significant complication after living donor liver transplantation.However, clinically usable tools for early individualized risk stratification remain limited in routine clinical practice. Material/Methods:This single-center retrospective cohort study included 205 adult patients who underwent LDLT between October 1, 2017, and October 31, 2025.All patients had complete 30-day follow-up for outcome ascertainment.Early bacterial infection was defined as a microbiologically confirmed infection occurring within 30 days after transplantation.Univariable and multivariable logistic regression analyses were performed to identify independent predictors.Model performance was evaluated using receiver operating characteristic (ROC) analysis, bootstrapbased calibration, and decision curve analysis.Static and web-based dynamic nomograms were constructed based on the final model. Results:Early bacterial infections occurred in 45 patients (21.9%).Higher pre-transplant MELD score, post-transplant blood transfusion, bile leakage, and longer intensive care unit (ICU) stay were independently associated with infection risk.The dynamic risk stratification tool demonstrated good discrimination (AUC=0.792,95% CI 0.716-0.867),good calibration on internal bootstrap validation (MAE and MSE), and favorable net clinical benefit on decision curve analysis.The web-based dynamic nomogram enabled real-time individualized risk estimation using routinely available clinical variables. Conclusions:We developed and internally validated a dynamic risk stratification tool for early bacterial infections after LDLT, implemented as static and web-based dynamic nomograms.The tool reflects evolving postoperative risk rather than prediction at a fixed time point, with good apparent performance.However, its true performance and generalizability remain uncertain, and external validation is required before clinical application.
Ngo et al. (Tue,) studied this question.