Predictors of Thirty-Day Morbidity and Unplanned Readmission after Induction Therapy Followed by Lung Resection for Non-Small Cell Lung Cancer: European Society of Thoracic Surgeons Database Analysis

OBJECTIVES Induction treatment followed by curative-intent surgery is a widely used approach for locally advanced NSCLC but is associated with increased postoperative morbidity. This study aimed to identify predictors of 30-day morbidity and unplanned readmission in patients who underwent lung resection for NSCLC after induction therapy, registered in the ESTS Database. METHODS We retrospectively analysed the ESTS Database, including patients who met the following criteria: (1) ≥18 years of age; (2) prior neoadjuvant therapy; (3) non-urgent major anatomical lung resection (lobectomy/bilobectomy/pneumonectomy) for NSCLC (01/2013–12/2024). Logistic regression was used to identify factors associated with 30-day morbidity and unplanned readmission. Two predictive models were developed; model accuracy was evaluated using ROC-AUC for 30-day morbidity and unplanned readmission. A risk score integrating key predictors was developed for clinical application. RESULTS A total of 4,547 patients were included. COPD (p < 0.001), previous malignancies (p < 0.001), arterial hypertension (p = 0.020), age (p < 0.001), and male sex (p = 0.026) were associated with 30-day morbidity. The predictive model developed using these factors achieved an accuracy of 92.7% (ROC-AUC = 0.70). Age (p < 0.001), major postoperative cardiopulmonary complications (p < 0.001), male sex (p < 0.001), and Clavien-Dindo complication grading (p < 0.001) were associated with 30-day unplanned readmission. The predictive model developed using these factors achieved an accuracy of 93.6% (ROC-AUC = 0.91). A predictive score for 30-day unplanned readmission was constructed. CONCLUSIONS We identified several preoperative factors associated with 30-day morbidity and unplanned readmission after postinduction lung resection. Furthermore, we developed a risk score that accurately predicts 30-day unplanned readmission. This score could be integrated into clinical practice to improve perioperative patient management.