Predicting prolonged mechanical ventilation after endovascular treatment for acute vertebrobasilar artery occlusion

Abstract

ABSTRACTBackgroundPatients with acute vertebrobasilar artery occlusion (VBAO) after endovascular treatment (EVT) frequently require mechanical ventilation (MV). This study aims to develop and validate a predictive score for prolonged mechanical ventilation (PMV) in patients with VBAO after EVT.MethodsThe derivation cohort prospectively recruited acute VBAO patients undergoing EVT from four comprehensive stroke centers (CSCs) in China. PMV was defined as continuous MV for ≥96 hours. Two multivariable logistic regressions were performed to predict predictive models based on whether adding radiographic results at follow-up. The performance of two models was evaluated for discrimination, calibration, and clinical utility. A validation cohort of 444 patients from acute Posterior circulation ischemic Stroke registry was enrolled to validate two models.ResultsThe derivation cohort consisted of 498 patients from four CSCs, among whom 204 (41.0%) required PMV. Reperfusion status, National Institutes of Health Stroke Scale scores, collateral circulation status, history of atrial fibrillation, fasting blood glucose, circulation insufficiency, the presence of malignant cerebellar edema, and symptomatic intracranial hemorrhage at follow-up imaging were found to be independent predictors of PMV in multivariable logistic regression. These variables were combined to create the basic and radiographic PAIRS scores. The basic PAIRS and radiographic PAIRS scores demonstrated adequate discrimination in derivation cohort (C-index, 0.77 [95%CI 0.73 to 0.81] and 0.80 [95%CI 0.77 to 0.84]), as well as the validation cohort (C-index, 0.70 [95%CI 0.65 to 0.75] and 0.75 [95%CI 0.70 to 0.80]), and had well-fitted calibration plot for both cohorts. Decision curve analysis showed that both PAIRS scores had satisfied net benefit in predicting PMV across various thresholds.ConclusionsThe PAIRS scores provide adequate accurate predictions of PMV in VBAO patients after EVT. These findings can assist patient families and caregivers in better understanding the treatment trajectory.