Monitoring carbon dioxide in mechanically ventilated patients during hyperbaric treatment.

BACKGROUND Measurement of the arterial carbon dioxide (P(a)CO(2)) is an established part of the monitoring of mechanically ventilated patients. Other ways to get information about carbon dioxide in the patient are measurement of end-tidal carbon dioxide (P(ET)CO(2)) and transcutaneous carbon dioxide (PTCCO2). Carbon dioxide in the blood and cerebral tissue has great influence on vasoactivity and thereby blood volume of the brain. We have found no studies on the correlation between P(ET)CO(2) or P(TC)CO(2), and P(a)CO(2) during hyperbaric oxygen therapy (HBOT). METHOD We studied 10 intubated and ventilatory stable patients during HBOT. End-tidal and transcutaneous measurements provided continuous data. Arterial blood samples were collected after reaching the operational pressure of 284 kPa (2.8 ATA) and analysed outside the chamber. A total of 17 paired samples of P(ET)CO(2), P(TC)CO(2) and P(a)CO(2) were obtained. RESULTS There was a good correlation between P(ET)CO(2) and P(a)CO(2) using linear regression (r(2) = 0.83). Bland-Altman analysis showed that P(ET)CO(2) on average was 2.22 kPa higher than P(a)CO(2) with limits of agreement (LoA) at ± 2.4 kPa. P(TC)CO(2), on average, was 2.16 kPa lower than P(a)CO(2) and the correlation using linear regression was poor (r(2) = 0.24). Bland-Altman analysis revealed LoA at ± 3.2 kPa. CONCLUSION During hyperbaric conditions we found that P(ET)CO(2) as opposed to P(TC)CO(2) offered the greater precision, but there was great variability among patients. Care must be taken when using P(ET)CO(2) or P(TC)CO(2) as an estimate of P(a)CO(2).