Related Experiment Video
Updated: Jun 12, 2026
Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
Development and validation of a predictive model for postoperative complications in thymoma patients
Zhendong Lu1, Fangchao Liu2, Dongjie Yan3
1Department of Cellular and Molecular Biology, Beijing Chest Hospital, Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Research Institute, No 9, Bei guan Street, Tong Zhou District, 101149, Beijing, P. R. China.
Objective:
To develop and validate an intervention model for the risk factors of postoperative complications in patients with thymoma.
Methods:
A retrospective analysis was conducted on the clinical data of patients who underwent thymoma resection surgery at our hospital between June 2006 and July 2024. Postoperative complications were systematically categorized according to the Clavien-Dindo classification (Grade II-V), including but not limited to: ① 30-day mortality, ② respiratory failure (defined as mechanical ventilation > 48 h or unplanned reintubation), ③ myocardial injury (troponin elevation > 5×ULN).Whether complications occurred was divided into two groups. The data were randomly divided into a modeling cohort (70%) and a validation cohort (30%). Logistic regression was used to identify risk factors for postoperative complications, and model performance was evaluated using ROC curves and calibration plots.
Results:
A total of 26.1% (66/253) of thymoma patients developed postoperative complications. Multivariable logistic regression analysis revealed that Masaoka-Koga stage, Myasthenia Gravis, intraoperative blood loss, forced expiratory volume in 1 second, and albumin were significant predictors in the model. The model demonstrated an area under the ROC curve of 0.84 (95% CI: 0.78-0.91) in the modeling cohort and 0.76 (95% CI: 0.63-0.89) in the validation cohort. The Hosmer-Lemeshow goodness-of-fit test showed: modeling cohort (χ2 = 5.0997, P = 0.6478), validation cohort (χ2 = 4.9946, P = 0.7582). Decision curve analysis indicated that the risk thresholds for the modeling and validation cohorts were 10%-70% and 10%-60%.
Conclusion:
This study developed a risk prediction model for postoperative complications in thymoma patients, demonstrating strong discriminative ability (C-statistic 0.82, 95% CI 0.78-0.86) and calibration (Brier score 0.11) upon internal validation. The model may aid in identifying high-risk patients for targeted preventive measures, though its impact on actual complication rates requires prospective intervention trials.
