Related Experiment Videos
An interpretable prognostic model for early unfavorable outcomes after bronchiectasis surgery
Yichen Hou1, Qibatu Wu1, Long Ma1
1State Key Laboratory of Pathogenesis, Prevention and Treatment of High Incidence Diseases in Central Asia, Department of Thoracic Surgery, The First Affiliated Hospital of Xinjiang Medical University, No. 137, South Liyushan Road, Xinshi District, Urumqi, Xinjiang Uygur Autonomous Region, 830011, P. R. China.
BMC Pulmonary Medicine
|July 17, 2026
Summary
A new nomogram predicts unfavorable outcomes after bronchiectasis surgery using preoperative data. This tool aids in personalized risk assessment and follow-up planning for patients undergoing lung resection for bronchiectasis.
Area of Science:
- Pulmonary Medicine
- Surgical Oncology
- Medical Informatics
Background:
- Bronchiectasis is a chronic airway disease necessitating lung resection in refractory cases.
- Postoperative recovery after bronchiectasis surgery is highly variable.
- A predictive tool for individualized risk stratification is currently unavailable.
Purpose of the Study:
- To develop and validate an interpretable nomogram for predicting early unfavorable outcomes after bronchiectasis surgery.
- To identify key preoperative predictors of postoperative complications and reduced quality of life.
Main Methods:
- Retrospective analysis of 142 patients undergoing bronchiectasis-related lung resection.
- Elastic Net-penalized logistic regression for model development and nomogram construction.
- Internal validation using nested cross-validation, assessing discrimination (AUC) and accuracy (Brier score).
Main Results:
- The incidence of early unfavorable outcomes was 57.7%.
- The final nomogram incorporated predictors: pCO₂, LVEF, FEV1%pred, MVV%pred, and BSI score.
- The model demonstrated good discrimination (AUC 0.820) and calibration, with potential clinical utility.
Conclusions:
- An Elastic Net-based nomogram effectively predicts early unfavorable outcomes after bronchiectasis surgery.
- The model utilizes routine preoperative variables for risk stratification and follow-up planning.
- External validation is recommended to confirm generalizability and clinical impact.