Thoracoscore

Thoracoscore predicts postoperative mortality for thoracic surgery patients using a multivariable preoperative risk model.

Key Features:

• Purpose: Predicts postoperative mortality by evaluating multiple preoperative clinical variables associated with thoracic surgical procedures.
• Variables Considered: Incorporates preoperative factors including pulmonary arterial hypertension (PAH), redo (re-operative) surgery, blood loss, blood transfusion requirements, duration of anesthesia, one-lung ventilation, and type of surgery performed.
• Development and Validation: Developed on a French population and validated in various international cohorts, with variable calibration and discrimination across populations.
• Performance Assessment: Model performance is evaluated using calibration and discrimination metrics such as the Hosmer-Lemeshow test and area under the curve (AUC) of receiver operating characteristic (ROC) curves.
• Validation Findings: An external validation in an Indian thoracic surgical cohort reported poor calibration and only fair discrimination, with observed mortality 3.2% versus a Thoracoscore-predicted 0.44% and identification of PAH and re-operative surgery as independent risk factors.

Scientific Applications:

• Risk Stratification: Estimates individual preoperative risk to inform surgical decision-making and consent discussions.
• Perioperative Planning: Identifies high-risk patients to guide perioperative management strategies and planning.
• Resource Allocation: Supports allocation of clinical resources by distinguishing patients at higher risk of postoperative mortality.

Methodology:

Retrospective data collection from thoracic surgery patients (posterolateral thoracotomy) followed by unpaired t-tests for continuous variables and Chi-square tests for categorical variables, multivariate logistic regression to identify independent predictors, and model evaluation using the Hosmer-Lemeshow test for calibration and AUC of ROC curves for discrimination.