—We thank Dr. Nemoto for the interest that he brought to our study of the correlation between cerebral cortical oxyhemoglobin saturation and jugular bulb saturation in patients with head injury. We do not agree that jugular bulb saturation measurement is not a standard of cerebral oxygenation assessment in patients with head injury. Because of the pioneering work of Robertson et al. , 1assessment of the global adequacy of cerebral blood flow to metabolism by jugular oximetry has shown the ability of this measurement to help clinicians avoid global cerebral hypoperfusion. This is especially true when therapeutic maneuvers, such as hyperventilation, are used to control refractory intracranial hypertension. 2,3We are aware that this measurement is a global one; therefore, a normal jugular oxygen saturation value cannot rule out the possibility that some area in the brain may be hypoxic or ischemic. We agree that this is a major limitation of the method. However, the same criticism could be brought against the measurement of cerebral oxygenation by methods that monitor a very limited cerebral area, such as near-infrared spectroscopy (NIRS) or direct cerebral oxygen measurement with intracerebral probes. In practice, clinicians would benefit from simultaneous measurements (i.e. , global and local) in patients with head injury.
The aim of our study was to highlight the difficulties inherent in the interpretation of NIRS data in patients with head injury. We purposely studied conditions in which cerebral blood flow could reasonably be expected to vary in the same direction in the territories that were simultaneously monitored using three methods: NIRS, transcranial Doppler ultrasonography, and jugular oxygen saturation. Among the multiple reasons that may explain the discrepancies that were observed and discussed in our study, the clinical setting is an important reason to consider. Indeed, measurement of light transmission by NIRS may be more difficult to perform in adults than in children or in head trauma with edema than during carotid clamping. A partition between intracranial and extracranial blood and also between arteriolar and venous compartments seems to be dependent on the different therapies that are used. It may explain why clinicians who work in different domains obtain NIRS data with different levels of accuracy and clinical relevance. Furthermore, improvement in NIRS technology and in the modeling of the light pathway through an adult skull should allow for the identification of the mechanisms that underlie the discrepancies that we observed between different monitoring techniques, and may find solutions that will correct for these discrepancies.