Related Experiment Video
Updated: Oct 17, 2026
Cardiac Loading using Passive Left Atrial Pressurization and Passive Afterload for Graft Assessment
Published on: August 2, 2024
Global Longitudinal Strain as a Predictor of Challenging Separation from Cardiopulmonary Bypass
Giulia Franceschini1, Andrea Farinaccio1, Alessia Zingaro1
1Cardio Thoracic Anesthesia Department, University Hospital Tor Vergata, Rome, Italy.
Objectives:
Although reduced left ventricular ejection fraction is a risk factor for adverse postoperative outcomes, its ability to predict challenging cardiopulmonary bypass (CPB) separation remains limited. We aimed to determine whether preoperative left ventricular global longitudinal strain (GLS), measured with a fully automated speckle-tracking echocardiographic software (LV AutoStrain) before induction of general anesthesia, can predict challenging separation from CPB.
Design:
Single-center prospective observational study.
Setting:
Adult patients undergoing cardiac surgery with CPB.
Participants:
Fifty-two consecutive patients were scheduled for cardiac surgery requiring CPB, with adequate preoperative transthoracic echocardiographic windows. Patients with preoperative atrial fibrillation, inotropic and mechanical circulatory support, or mechanical ventilation were excluded.
Interventions:
GLS was obtained non-invasively by transthoracic echocardiography using LV AutoStrain.
Measurements And Main Results:
CPB separation was classified as easy or challenging. Compared with the easy-weaning group, patients in the challenging weaning group showed more impaired preoperative GLS values (-11.8% [IQR -14.0% to -10.3%] v -16.1% [IQR -18.6% to -15.4%]; p < .01) and longer CPB (132.5 minutes [94.0-164.0] v 100.0 minutes [69.2-132.5]; p = .036) and aortic cross-clamp times (89.0 [67.5-116.0] v 61.0 [40.0-79.0]; p = .004). On multivariable analysis adjusted for EuroSCORE II, left ventricular ejection fraction category, CPB duration, and type of surgical procedure, GLS remained the only independent predictor of challenging CPB separation (odds ratio 1.40, 95% confidence interval [CI] 1.09-1.79; p < .01). Receiver operating characteristic curve analysis showed good discrimination for challenging weaning (area under the curve 0.864, 95% CI 0.740-0.969; p < .001); the optimal threshold of -14.0% for preoperative GLS yielded 80.0% sensitivity and 90.9% specificity.
Conclusion:
Fully automated, operator-independent preoperative GLS was strongly and independently associated with challenging separation from CPB, with a clinically applicable threshold. GLS may represent a rapid, reproducible adjunct to perioperative risk stratification.