Bensaidane, Mohamed R.Turgeon, Alexis F.Lauzier, FrançoisEnglish, Shane W.Leblanc, GuillaumeFrancoeur, Charles L.2025-11-182025-11-182025-11-14Critical Care. 2025 Nov 14;29(1):487https://doi.org/10.1186/s13054-025-05701-3http://hdl.handle.net/10393/51054Abstract Purpose Near-infrared spectroscopy (NIRS) is a non-invasive, real-time and continuous cerebral oximetry monitoring with potential applications in the management of aneurysmal subarachnoid hemorrhage (aSAH) and in the detection of delayed cerebral ischemia (DCI). The aim of this study was to evaluate the association between NIRS and outcome in aSAH and its diagnostic accuracy for DCI. Methods Systematic review and meta-analysis of studies involving adult aSAH patients monitored with NIRS during index hospitalization. Primary outcome was the functional outcome at 90 days or more. Secondary outcomes included any functional outcome, mortality and diagnostic accuracy for DCI. Random effects meta-analyses were performed, and for diagnostic accuracy, forest plots and random effects meta-analyses were used to determine pooled sensitivity, specificity, diagnostic odds ratios and to generate a receiver operating characteristic (ROC) curve. Results Of the 28,296 citations identified, 35 satisfied inclusion criteria. Three studies (202 patients) were included for meta-analysis of the primary outcome. Cerebral desaturation events or loss of autoregulation as detected with NIRS were associated with higher risks of unfavourable outcome at 90 days (RR 4.29 95% CI [2.10;8.79]). Significant associations were also observed with mortality (RR 4.24, 95% CI [2.43;7.41]). Diagnostic accuracy analysis demonstrated moderate sensitivity (0.85), specificity (0.65), and diagnostic odds ratio (10.42), with a receiver operating characteristic (ROC) curve area of 0.68. The certainty of evidence was moderate for the association between NIRS and patient outcomes, and low for its diagnostic accuracy in detecting DCI.The overall quality of evidence was limited by small sample sizes, high heterogeneity in study methods and patient populations, and potential publication bias. Conclusion Cerebral desaturation events and impaired autoregulation detected by NIRS are consistently associated with poor outcomes and mortality in aSAH. However, NIRS alone provides only moderate diagnostic accuracy for DCI, with a considerable risk of false positives. The evidence is weakened by methodological limitations, heterogeneous thresholds, and the absence of a universally accepted reference standard for DCI diagnosis. Importantly, no interventional data are available to demonstrate an effect on patient outcomes. While incorporation into multimodal neuromonitoring strategies appears promising, robust prospective trials are needed before NIRS can be reliably adopted in routine clinical practice.Journal Article2025-11-18enThe Author(s)