Toward comprehensive multimodal neuromonitoring of surrogates of cerebral autoregulation in traumatic brain injury: insights from a hyperventilation protocol.

SIGNIFICANCE: It is hypothesized that neurocritical-care targeting cerebral autoregulation (CAR) would lead to improved outcomes. However, there is no single accepted CAR index due to limited accessibility of the methods and the complexity of interpretation. It is imperative to introduce accessible and direct methods. AIM: We aim to compare traditional CAR indices and those on hybrid near-infrared diffuse optical (DO) monitors on a clinical population undergoing a controlled intervention. APPROACH: Traumatic brain injury patients undergoing a hyperventilation protocol were monitored. Seven CAR indices based on the common correlation approach were derived. The detection of CAR alteration by hyperventilation, their variability, and the simultaneity of the detected changes were quantified. RESULTS: Seventeen patients have been monitored, showing a predominant decrease in CAR due to hyperventilation in some but not all indices. The quantified variability has demonstrated that DO-based indices could be more sensitive to small changes, and a compound "agreement score" has indicated that most recordings (18/20) displayed directional consistency, supporting the robustness of multimodal assessment. CONCLUSIONS: We have demonstrated that different CAR indices lead to different interpretations and further analysis and clinical trials are merited. We call for the release of open data-sets to enable this.