Predicting bloodless aortic aneurysm repair: A simple point system

Abstract

Introduction Patients who decline blood transfusion for religious or personal reasons present a unique challenge in cardiac surgery. We developed and validated a preoperative point-based tool to predict the likelihood of achieving transfusion-free success in open proximal thoracic aortic aneurysm repair.

Methods We retrospectively analyzed 1074 patients who underwent open proximal thoracic aortic aneurysm repair at a single academic center from 2012 to 2023. A multivariable logistic regression model was developed using cross-validation to identify independent predictors of bloodless success. Model performance was evaluated using area under the curve (AUC), Brier score, and F-score. A point system was derived from weighted coefficients of significant preoperative variables.

Results Multivariable analysis identified 8 factors (p < 0.05) associated with bloodless aneurysm repair for the point system: body surface area (BSA), hypertension, reoperation, heart failure, baseline hemoglobin, extent of repair, concomitant procedure, and degree of hypothermia. The model demonstrated good discrimination (AUC = 0.761), calibration (Brier score = 0.192), and balance (F-score = 0.758). BSA ≥1.9 m² and preoperative hemoglobin ≥13.1 g/dL (+8 points each) had the greatest positive influence on likelihood of bloodless operation, while proximal + distal extension in repair (−8 points) had the greatest negative influence on likelihood of bloodless operation.

Conclusions This novel point-based scoring tool estimates the probability of transfusion-free success using readily available preoperative factors. It may assist in counseling patients who refuse blood products, guiding perioperative planning, and prioritizing resource-intensive optimization strategies for those at highest risk.