Frontiers in Neurology (Aug 2023)

Prediction models for post-thrombectomy brain edema in patients with acute ischemic stroke: a systematic review and meta-analysis

  • Lei Liu,
  • Chun-yu He,
  • Jia-xin Yang,
  • Si-ting Zheng,
  • Jun Zhou,
  • Ye Kong,
  • Wen-bo Chen,
  • Yan Xie

DOI
https://doi.org/10.3389/fneur.2023.1254090
Journal volume & issue
Vol. 14

Abstract

Read online

ObjectiveThe objective of this study is to systematically evaluate prediction models for post-thrombectomy brain edema in acute ischemic stroke (AIS) patients. This analysis aims to equip clinicians with evidence-based guidance for the selection of appropriate prediction models, thereby facilitating the early identification of patients at risk of developing brain edema post-surgery.MethodsA comprehensive literature search was conducted across multiple databases, including PubMed, Web of Science, Embase, The Cochrane Library, CNKI, Wanfang, and Vip, aiming to identify studies on prediction models for post-thrombectomy brain edema in AIS patients up to January 2023. Reference lists of relevant articles were also inspected. Two reviewers independently screened the literature and extracted data. The Prediction Model Risk of Bias Assessment Tool (PROBAST) and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines were employed to assess study bias and literature quality, respectively. We then used random-effects bivariate meta-analysis models to summarize the studies.ResultsThe review included five articles, yielding 10 models. These models exhibited a relatively high risk of bias. Random effects model demonstrated that the AUC was 0.858 (95% CI 0.817–0.899).ConclusionDespite the promising discriminative ability shown by studies on prediction models for post-thrombectomy brain edema in AIS patients, concerns related to a high risk of bias and limited external validation remain. Future research should prioritize the external validation and optimization of these models. There is an urgent need for large-scale, multicenter studies to develop robust, user-friendly models for real-world clinical application.Systematic review registrationhttps://www.crd.york.ac.uk, unique Identifier: CRD42022382790.

Keywords