Cerebral near-infrared spectroscopy monitoring (NIRS) in children and adults:a systematic review with meta-analysis

Background Cerebral oxygenation monitoring utilising near-infrared spectroscopy (NIRS) is increasingly used to guide interventions in clinical care. The objective of this systematic review with meta-analysis and Trial Sequential Analysis is to evaluate the effects of clinical care with access to cerebral NIRS monitoring in children and adults versus care without. Methods This review conforms to PRISMA guidelines and was registered in PROSPERO (CRD42020202986). Methods are outlined in our protocol (doi: 10.1186/s13643-021-01660-2). Results Twenty-five randomised clinical trials were included (2606 participants). All trials were at a high risk of bias. Two trials assessed the effects of NIRS during neonatal intensive care, 13 during cardiac surgery, 9 during non-cardiac surgery and 1 during neurocritical care. Meta-analyses showed no significant difference for all-cause mortality (RR 0.75, 95% CI 0.51-1.10; 1489 participants; I-2 = 0; 11 trials; very low certainty of evidence); moderate or severe, persistent cognitive or neurological deficit (RR 0.74, 95% CI 0.42-1.32; 1135 participants; I-2 = 39.6; 9 trials; very low certainty of evidence); and serious adverse events (RR 0.82; 95% CI 0.67-1.01; 2132 participants; I-2 = 68.4; 17 trials; very low certainty of evidence). Conclusion The evidence on the effects of clinical care with access to cerebral NIRS monitoring is very uncertain. Impact The evidence of the effects of cerebral NIRS versus no NIRS monitoring are very uncertain for mortality, neuroprotection, and serious adverse events. Additional trials to obtain sufficient information size, focusing on lowering bias risk, are required. The first attempt to systematically review randomised clinical trials with meta-analysis to evaluate the effects of cerebral NIRS monitoring by pooling data across various clinical settings. Despite pooling data across clinical settings, study interpretation was not substantially impacted by heterogeneity. We have insufficient evide