Study objective: To evaluate the accuracy and precision of estimation of partial pressure of carbon dioxide (Pa(CO2)) using end-tidal or transcutaneous CO2 (TcP(CO2)) measurements during mechanical ventilation in the elderly patient.
Design: A prospective, observational study was conducted.
Settings: The study was done in the anesthesia department of a university hospital.
Patients: Seventeen anesthetized, mechanically ventilated patients older than 60 years were studied.
Interventions and measurements: During standard sevoflurane anesthesia, and after proper calibration and an equilibration time of 30 minutes with stable hemodynamic and respiratory variables, arterial (Pa(CO2)), end-tidal (Pet(CO2)), and transcutaneous (TcP(CO2)) CO2 partial pressures were determined. In each patient, 1 to 5 sample sets (Pa(CO2), Pet(CO2), and TcP(CO2)) were obtained.
Main results: A total of 45 sample sets were obtained from the patients studied. The Pa(CO2) values ranged between 21 and 58 mm Hg. The Pa(CO2) - Pet(CO2) tension gradient was 6 +/- 5 mmHg (95% confidence interval, -3 to 16 mmHg), whereas the Pa(CO2) - TcP(CO2) tension gradient was 2 +/- 4 mmHg (95% confidence interval, -6 to 9 mmHg) (P = 0.0005). The absolute value of the difference between Pa(CO2) and Pet(CO2) was 3 mm Hg or less in 7 of 45 sample sets (15%), whereas the absolute value of the difference between Pa(CO2) and TcP(CO2) was 3 mm Hg or less in 21 of 45 sample sets (46%) (P = 0.003). Linear regression analysis for TcP(CO2) versus Pa(CO2) showed a slope of 0.84 (r(2) = 0.73), whereas the linear regression analysis for Pet(CO2) versus Pa(CO2) showed a slope of 0.54 (r(2) = 0.50).
Conclusion: Transcutaneous monitoring of CO(2) partial pressure gives a more accurate estimation of arterial CO(2) partial pressure than does Pet(CO2) monitoring.