This recently published open-access article by Robba et al delves into the retrospective monitoring data on 70 TBI patients. The patients in the cohort were all monitored comprehensively, providing the investigators with continuous data on PbtO2, ICP and CPP, as well as on relevant respiratory indices. They found a statistically significant link between lung function and brain oximetry. Their data suggests that the PaO2/FiO2 ratio is an important independent determinant of cerebral hypoxia and mortality, which seems intuitive enough. The finding that intrigued me the most, however, is that the purported hypoxic threshold as measured by PbtO2 seems to be both highly age-dependent, and varies between 21-30 mmHg, with the strongest correlation found at 28 mmHg for patients 60 years or older. This threshold is significantly higher than those applied by most centres that monitor cerebral oximetry invasively. My own centre, for instance, operates with a lower threshold of 20 mmHg using the Licox (TM) device. It really shouldn’t surprise us that hypoxia thresholds are subject to variation, just as ICP-tolerance and optimal cerebral perfusion pressure are likely to be highly individual. These findings should nudge us further towards tailoring our treatment goals to each patient. There is some hope that large PbtO2 trials such as the ongoing BOOST 3 study will provide more data on which thresholds to apply for which patients.