Objective: To determine how much information concerning resuscitation and outcome is provided by the end-tidal CO2 and derived variables obtained during surgery.
Design: Retrospective chart review.
Setting: Emergency hospital operating room.
Patients: One hundred critically ill or injured patients requiring major surgery and having a mortality rate of 41%.
Interventions: Standard intraoperative monitoring, including continuous capnography, plus arterial blood gas analyses every 1 to 1.5 hrs during surgery.
Measurements and main results: There was only a fair correlation between the PaCO2 and end-tidal CO2 (r2 = .14). The mortality rates in these patients were highest in those patients who had the lowest end-tidal CO2 values, the highest arterial to end-tidal CO2 differences, and the highest estimated alveolar deadspace fraction. A persistent end-tidal CO2 of < or = 28 torr (< or = 3.8 kPa) was associated with a mortality rate of 55% (vs. 17% in those patients with a higher end-tidal CO2). The mortality rate was also increased in patients with a persistent arterial to end-tidal CO2 difference of > or = 8 torr (> or = 1.1 kPa) (58% vs. 23%).
Conclusions: End-tidal CO2 and derived values should be monitored closely in critically ill or injured patients. Efforts should be made--by increasing cardiac output and core temperature and by adjusting ventilation as needed--to maintain the end-tidal CO2 at > or = 29 torr (> or = 3.9 kPa) and the arterial to end-tidal CO2 difference at < or = 7 torr (< or = 1.0 kPa).