Development of a Nomogram for Predicting the Occurrence of Symptomatic Intracranial Haemorrhage after Thrombectomy: A Single-Center Retrospective Observational Study
Abstract Background Symptomatic intracranial haemorrhage (SICH) is a severe and deadly complication in patients with large vessel occlusion (LVO) who receive endovascular treatment (EVT). Recent studies have indicated that many risk factors, including pretreatment scores and the operation process, may be associated with the occurrence of SICH after thrombectomy. This study aims to identify independent risk factors and establish a novel nomogram-based model for patients with anterior LVO to predict the occurrence of SICH after direct thrombectomy or bridge therapy (thrombectomy based on intravenous thrombolysis). Methods Patients with acute ischaemic stroke after EVT to recanalize the blocked artery in anterior circulation were consecutively recruited from November 2017 to March 2019. Baseline information was collected from each patient. These data were subsequently analysed by R Project for Statistical Computing. Results A total of 127 patients with complete data were classified into the training set, among whom 37 patients (29.1%) fulfilled the criteria for SICH. The results of the multivariate analyses showed that NIHSS (P=0.024), ASPECT (P<0.001) and ASITN (P=0.017) scores were independently associated with the occurrence of SICH after thrombectomy. Ultimately, three independent pretreatment predictors were included in the NIHSS/ASPECT/ASITN (NAA) prediction model, and the receiver operating characteristic analysis results showed an area under the curve (AUC) of 0.845 (95% CI=0.763–0.928). The calibration plots showed that the actual observations were consistent with the measured and predicted results of the nomogram. Conclusions In this study, a novel model based on NAA for predicting the occurrence of SICH after thrombectomy in patients with anterior LVO was established and validated internally. The results suggest that this model can help improve perioperative evaluations and individualized treatment strategies.