Intensive Care Medicine Experimental · Published 2026-04-21 · DOI 10.1186/s40635-026-00899-0
Abstract Background Post-induction hypotension (PIH) is associated with acute perioperative organ injury. We quantified the added value of echocardiographic parameters and ventriculo-arterial coupling (VAC)-related variables for the prediction of PIH. Methods A prospective observational cohort study conducted between July 2023 and November 2024 enrolled adults undergoing elective non-cardiac surgery. The ventriculo-arterial coupling index (E a/E es) was derived pre-operatively using transthoracic echocardiography combined with non-invasive blood pressure measurements. PIH was defined as the first occurrence of mean arterial pressure < 65 mmHg between anaesthesia induction and surgical incision. Predictors were analysed using a generalised additive model to account for potential non-linear associations. Incremental prognostic value was assessed using the fraction of new information (FNI), a reclassification-based metric, and changes in the area under the receiver operating characteristic curve (AUC). Results PIH occurred in 161/405 patients (39.8%). An E a/E es ratio > 1 was associated with PIH (OR 2.95; 95% CI 1.08–8.03; p = 0.034). The clinical model showed an AUC of 0.720 (95% CI 0.669–0.771). The addition of echocardiographic parameters increased the AUC to 0.768 (95% CI 0.720–0.816; Holm-adjusted p = 0.018) and provided 37% FNI (Holm-adjusted p = 0.199). The incorporation of VAC-related variables further increased the AUC to 0.785 (95% CI 0.739–0.831; Holm-adjusted p = 0.002) and yielded 46% FNI (Holm-adjusted p = 0.030). Compared with the clinical–echocardiographic model, incorporation of VAC-related variables provided an additional 14% FNI (Holm-adjusted p = 0.018) without a statistically significant AUC increase (0.017; 95% CI −0.004–0.039; Holm-adjusted p = 0.199). Conclusion PIH was common and was independently associated with impaired VAC, as reflected by an E a/E es ratio > 1. Incorporation of echocardiographic parameters improved the discriminatory performance of a clinical prediction model, and the further addition of VAC-related variables provided meaningful improvement in risk reclassification, despite only modest gains in overall discrimination.
Abstract from DOAJ. Public domain (CC0 1.0).
Read the article at the publisher →