A Predictive Score for Bronchopleural Fistula Established Using the French Database Epithor

Area of Science:

• Thoracic Surgery
• Pulmonary Medicine
• Surgical Complications

Background:

• Bronchopleural fistula (BPF) is a rare but fatal complication following thoracic surgery.
• Pulmonary resection, including lobectomy, bilobectomy, and pneumonectomy, carries a risk of BPF.
• Identifying patients at high risk for BPF is crucial for improving surgical outcomes.

Purpose of the Study:

• To develop and validate a predictive model for bronchopleural fistula (BPF) after pulmonary resection.
• To identify key factors associated with the occurrence of BPF.
• To provide surgeons with a tool for risk stratification in patients undergoing lung surgery.

Main Methods:

• Utilized data from 34,000 patients undergoing major pulmonary resection from the French National database Epithor (2005-2012).
• Employed logistic regression with backward stepwise variable selection to build the predictive model.
• Validated the model using the C-index and Hosmer-Lemeshow test for calibration.

Main Results:

• The overall incidence of BPF was 0.94%, with higher rates after bilobectomy (2.2%) and pneumonectomy (3%) compared to lobectomy (0.5%).
• Mortality rates associated with BPF ranged from 16.7% to 25.9% depending on the type of resection.
• Nine variables were identified as significant predictors of BPF: sex, BMI, dyspnea score, comorbidities, type of resection (bilobectomy, pneumonectomy), emergency surgery, sleeve resection, and resection side. The model demonstrated good predictive accuracy (C-index=0.8).

Conclusions:

• A robust predictive model for bronchopleural fistula (BPF) has been developed and validated.
• The model effectively identifies patients at high risk for BPF following pulmonary resection.
• Prospective validation in an independent cohort is recommended to confirm the model's utility in clinical practice.