The Diagnostic Accuracy of Intra-Operative Near Infrared Spectroscopy in Carotid Artery Endarterectomy Under Regional Anaesthesia: Systematic Review and Meta-Analysis

Objective: Intra-operative near infrared spectroscopy (NIRS) is a non-invasive tool used to monitor regional cerebral oxygen saturation during carotid endarterectomy (CEA), for which accuracy remains unclear. Therefore, this systematic review and meta-analysis aimed to determine the diagnostic accuracy of NIRS in patients undergoing CEA under regional anaesthesia (RA). Data sources: MEDLINE, Scopus, and Web of Science were searched for studies that compared NIRS with the "awake test" in patients undergoing CEA under RA. Review methods: Bivariable random effects meta-analysis was performed to determine the diagnostic accuracy of NIRS to detect cerebral ischaemia. Meta-regression was performed to explore causes of heterogeneity. Metaanalysis of proportions was also performed to determine the accuracy of NIRS in predicting 30 day stroke. Study quality was evaluated using the QUADAS-2 criteria. Results: Eleven primary studieswere included, assessing 1 237 participants. Themeta-analysis obtained a partial area under the summary receiver operating characteristic curve for diagnosing brain ischaemia of 0.646, with a summary sensitivity of 72.0% (95% confidence interval [CI] 58.1 - 82.7; I-2 = 48.6%) and a specificity of 84.1% (95% CI 78.5 - 88.4; I-2 = 48.6%). In meta-regression analysis, the frequency of hypertension (p = .011) and patients with symptomatic carotid stenosis (p = .031) were significant effect modifiers. Higher frequency of arterial hypertension (z score = -2.15; p = .032) and diabetes (z score = -2.12; p = .034) were associated with lower summary sensitivity, while a higher frequency of symptomatic carotid stenosis (z score = 2.11; p = .035) was associated with higher summary sensitivity. Point estimate sensitivity and specificity for predicting 30 day stroke occurrence were 41% (95% CI 19.5 - 66.6; I-2 = 0%) and 81.4% (95% CI 74.4 - 86.9, I-2 = 65.6%), respectively. Conclusion: The results of this study suggest that NIRS as a cerebral monitoring technique does not have sufficiently high sensitivity or specificity to be used alone in the neurological monitoring of patients undergoing CEA under RA.