End Tidal CO2 in TBI
- Aug 18th, 2014
- Daniel Ostermayer
Does End Tidal CO2 correlate with PaCO2 in Traumatic Brain Injury?
Your neurosurgeons and trauma team have accepted a transfer to your hospital for intensive management of a trauma patient who presented to a small community hospital with a traumatic subarachnoid hemorrhage and epidural hematoma after being involved in a motorcycle accident.
Upon arrival with the critical care transport team, the patient is already intubated and stable on a a ventilator with appropriate sedation and stable hemodynamics. However, the neurosurgeons are in the operating room managing a spontaneous intraparenchymal hemorrhage and there are no available ICU beds due to multiple gun shot victims from a gang fight that you finished admitting.
While the patient is in the ED, the neurosurgeons recommend maintaining eucapnea for the patient since while there are no acute signs of herniation.(1).
Can you use the end tidal CO2(etCO2) or do you need to rely on arterial blood gas (ABG) measurements to maintain PaCO2 between 35-40 mm Hg?
In trauma patients the most robust evidence for the correlation between etCO2 and PaCO2 comes form a prospective observational study in Emergency department patients at a single center conducted by Lee et. al in 2009.(2) The median difference of PaCO2 and etCO2 was 3.6 mm Hg and greater in patients with severe hypotension and lactates > 7 mm/L. To have maximal safety it is safe to assume that the etCO2 generally underestimated the PaCO2 by at least 5 mm Hg and the PaCO2 can be at least equal but possibly higher than the etCO2.
However in poly trauma patients especially those with severe chest and abdominal trauma there was as little as a 29% acceptable correlation of 5mmHg between the etCO2 and the paCO2.(3) In those cases Warner et al. in 2009 concluded that there is an unacceptable correlation between etCO2 and PaCO2 in the very sick and severely injured trauma patients. It is more likely that the etCO2 is artificially low and is a measure of relative perfusion and less correlated with the PaCO2 and ventilation. These patients should have arterial PaCO2 measurements performed by ABG and most likely will benefit from an arterial line for monitoring resuscitation efforts.
Case conclusion: Since the patient remained hemodynamically stable on the ventilator and only suffered from isolated TBI, you performed an initial ABG at found a PaCO2 of 37 mmHg and observed an etCO2 of 39-41 mm Hg by waveform capnography. For the next 3 hours in the ED you continued to monitor the etCO2 and did not perform any repeat ABGs.
Bottom Line: End tidal CO2 will differ most from PaCO2 in the severely injured and in patients with shock. In isolated hemodynamically stable Traumatic Brain Injury, the etCO2 should correlate relatively well by at least 5mm Hg.