Predicting tigecycline-related adverse events in Infected patients: A machine learning approach with clinical interpretability

Shiya Wu¹Yuheng Chen²Wenjie Fan²Xirong Wu¹Chaofeng ZHANG³Yuchang Lin³Qi Lin^3*

• ¹Fujian Medical University, Fuzhou, China
• ²Putian University, Putian, China
• ³Affiliated Hospital of Putian University, Putian, China

Background: Tigecycline (TGC), while effective against multidrug-resistant infections, is limited by hepatotoxicity and coagulation disorders, yet lacks robust predictive tools. Methods: We developed an online dynamic nomogram to assess these adverse events using retrospective data from 2,553 TGC-treated patients (2020-2025). Seventy-seven clinical features were analyzed using Boruta and LASSO for feature selection. Seven machine learning (ML) models were evaluated via ten-fold cross-validation as well as receiver operating characteristic (ROC) curve and calibration curves, with SHAP analysis for interpretability and an online Dynamic Nomogram for clinical translation. Results: Logistic regression (LR) outperformed other algorithms, achieving AUCs of 0.800 (95% CI: 0.727-0.874) for hepatotoxicity and 0.755 (0.665-0.845) for coagulation dysfunction. Independent risk factors for liver injury included prolonged treatment duration, high dosage, ICU admission, HBV infection, and elevated baseline levels of LDH and GGT. Risk factors for coagulation dysfunction included extended treatment duration, ICU admission, elevated baseline creatinine, sepsis, and septic shock. Notably, co-administration of meloxicillin and higher baseline RBC levels appeared to be protective. Conclusions: This study constructed an online dynamic nomogram with good discrimination and calibration, which can help to identify high-risk patients and help clinicians to perform early risk stratification and individualized treatment planning.