Citation

Ahn C, Oh J, Kim CW, Lee H, Lim TH, Kang H. Sci. Rep. 2022; 12(1): e3529.

Copyright

(Copyright © 2022, Nature Publishing Group)

DOI

10.1038/s41598-022-07191-7

PMID

35241701

Abstract

We aimed to assess the evidence regarding the usefulness of brain imaging as a diagnostic tool for delayed neurological sequelae (DNS) in patients with acute carbon monoxide poisoning (COP). Observational studies that included adult patients with COP and DNS were retrieved from Embase, MEDLINE, and Cochrane Library databases in December 2020 and pooled using a random-effects model. Seventeen studies were systematically reviewed. Eight and seven studies on magnetic resonance imaging (MRI) and computed tomography (CT), respectively, underwent meta-analysis. The pooled sensitivity and specificity of MRI for diagnosis of DNS were 70.9% (95% confidence interval [CI] 64.8-76.3%, I(2) = 0%) and 84.2% (95% CI 80.1-87.6%, I(2) = 63%), respectively. The pooled sensitivity and specificity of CT were 72.9% (95% CI 62.5-81.3%, I(2) = 8%) and 78.2% (95% CI 74.4-87.1%, I(2) = 91%), respectively. The areas under the curve for MRI and CT were 0.81 (standard error, 0.08; Q* = 0.74) and 0.80 (standard error, 0.05, Q* = 0.74), respectively. The results indicate that detecting abnormal brain lesions using MRI or CT may assist in diagnosing DNS in acute COP patients.

Language: en