C23-11 Procedural Predictors of Pneumothorax and Bleeding in the Era of Robotic and Image-Guided Bronchoscopy

Rationale Advances in interventional pulmonology, including robotic bronchoscopy, cone-beam imaging, and novel biopsy devices, have expanded diagnostic capabilities but may also alter complication risk. Pneumothorax and bleeding remain the most clinically significant adverse events, yet their independent procedural predictors are not well established. We sought to identify which bronchoscopic techniques independently increase the risk of these complications. Methods We performed a retrospective analysis of 390 bronchoscopic procedures conducted at a tertiary academic center between 2023 and 2025. The primary outcomes were pneumothorax and bleeding, each analyzed as a binary variable. Predictor variables included bronchoscopy type (robotic vs flexible), imaging modality (cone-beam CT and fluoroscopy), and procedural techniques (transbronchial needle aspiration (TBNA), endobronchial biopsy, cryobiopsy, and forceps transbronchial biopsy). Two separate multivariable logistic regression models were constructed, one for pneumothorax and one for bleeding. Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated and displayed as forest plots. Results In the pneumothorax model, forceps transbronchial biopsy was independently associated with a significantly increased risk of pneumothorax (adjusted OR ≈ 18.6; 95% CI 3.9-89.6; p < 0.001). No other procedural variable, including robotic bronchoscopy, cone-beam CT, fluoroscopy, TBNA, cryobiopsy, or endobronchial biopsy, was independently associated with pneumothorax. In the bleeding model, no procedural technique demonstrated a statistically significant association after adjustment. Both models showed acceptable calibration and overall fit (Nagelkerke R² ≈ 0.18 for pneumothorax). Conclusions In this bronchoscopy cohort, forceps transbronchial biopsy was the only independent predictor of pneumothorax after multivariable adjustment, whereas no procedural technique was associated with bleeding. Prior studies have shown minimal or insignificant differences in pneumothorax and bleeding rates across bronchoscopic techniques. Our findings contribute additional data from a broader diagnostic bronchoscopy context and suggest a possible signal toward increased pneumothorax risk with forceps biopsy. Further prospective studies are needed to validate these observations and refine procedure-specific safety strategies.