How and Whom to Monitor for Seizures in an ICU: A Systematic Review and Meta-Analysis

Objectives: To pool prevalence of nonconvulsive seizure, nonconvulsive status epilepticus, and epileptiform activity detected by different electroencephalography types in critically ills and to compare detection rates among them. Data Sources: MEDLINE (via PubMed) and SCOPUS (via Scopus) Study Selection: Any type of study was eligible if studies were done in adult critically ill, applied any type of electroencephalography, and reported seizure rates. Case reports and case series were excluded. Data Extraction: Data were extracted independently by two investigators. Separated pooling of prevalence of nonconvulsive seizure/nonconvulsive status epilepticus/epileptiform activity and odds ratio of detecting outcomes among different types of electroencephalography was performed using random-effect models. This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and also adhered to the Meta-analyses Of Observational Studies in Epidemiology guidelines. Quality of evidence was assessed with the Newcastle-Ottawa Quality Assessment Scale for observational studies and Cochrane methods for randomized controlled trial studies. Data Synthesis: A total of 78 (16,707 patients) and eight studies (4,894 patients) were eligible for pooling prevalence and odds ratios. For patients with mixed cause of admission, the pooled prevalence of nonconvulsive seizure, nonconvulsive status epilepticus, either nonconvulsive seizure or nonconvulsive status epilepticus detected by routine electroencephalography was 3.1%, 6.2%, and 6.3%, respectively. The corresponding prevalence detected by continuous electroencephalography monitoring was 17.9%, 9.1%, and 15.6%, respectively. In addition, the corresponding prevalence was high in post convulsive status epilepticus (33.5%, 20.2%, and 32.9%), CNS infection (23.9%, 18.1%, and 23.9%), and post cardiac arrest (20.0%, 17.3%, and 22.6%). The pooled conditional log odds ratios of nonconvulsive seizure/nonconvulsive status epilepticus detected by continuous electroencephalography versus routine electroencephalography from studies with paired data 2.57 (95% CI, 1.11–5.96) and pooled odds ratios from studies with independent data was 1.57 (95% CI, 1.00–2.47). Conclusions: Prevalence of seizures detected by continuous electroencephalography was significantly higher than with routine electroencephalography. Prevalence was particularly high in post convulsive status epilepticus, CNS infection, and post cardiac arrest. Dr. Limotai had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Limotai and Thakkinstian contributed to drafting of the article and statistical analysis. Drs. Limotai, Ingsathit, and Thakkinstian contributed to study concept or design, administrative, technical, or material support. Drs. Ingsathit, McEvoy, Attia, and Thakkinstian contributed to study supervision. Drs. Limotai, Ingsathit, McEvoy, Attia, and Thakkinstian contributed to critical revision of the article for important intellectual content. All authors contributed to the acquisition, analysis, or interpretation of data. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website (http://bit.ly/29S62lw). This article is an integral part of training for Dr. Limotai who is a PhD candidate in Clinical Epidemiology and Biostatistics in the Faculty of Medicine Ramathibodi Hospital and Faculty of Graduate Studies, Mahidol University, Bangkok, Thailand. The remaining authors have disclosed that they do not have any potential conflicts of interest. For information regarding this article, E-mail: atiporning@gmail.com; atiporn.ing@mahidol.ac.th; chusak.l@chula.ac.th Copyright © by 2019 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.

http://bit.ly/2FBRtki