Objective: To determine whether the measurement error of saline PCO2, using blood gas analyzers, is relevant for the interpretation and clinical use of the gastric intramucosal pH measurement.
Design: A comparison of four different blood gas analyzers (ABL-520, Ciba Corning, IL-1302, and Nova), using tonometered saline as the reference.
Setting: Clinical laboratory of a university hospital intensive care unit.
Interventions: None.
Measurements and main results: The bias and the precision of each blood gas analyzer was determined for measurements of PCO2 in saline samples. These samples had been balanced to PCO2 levels of 30, 45, and 68 torr (4, 6, and 9 kPa, respectively). In addition, the effect of buffering the saline was evaluated. The bias of the PCO2 measurement increased (p < .001) at the higher PCO2 levels. The bias ranged from -5.2 to -25.9 torr (-0.69 to -3.45 kPa) at a PCO2 of 45 torr (6 kPa) and from -5.2 to -33.1 torr (-0.69 to -4.41 kPa) at a PCO2 of 68 torr (9 kPa), and there was a significant (p < .001) analyzer-PCO2 level interaction. The type of the analyzer also influenced the bias (p < .001). The Nova analyzer underestimated the PCO2 by 50% to 60%. The other analyzers underestimated the PCO2 by 5% to 19%. The use of the buffer reduced the bias of all analyzers (p < .001). Based on the precision of the saline PCO2 measurement, a difference in gastric intramucosal pH of 0.06 pH units can be reliably detected at a PCO2 of 45 torr (6 kPa) by all analyzers, with the exception of the Nova analyzer.
Conclusions: Measurement of saline PCO2 is an important source of error in the assessment of gastric intramucosal pH, and the error depends on both the analyzer used and the actual PCO2 level. Direct comparison of pH values obtained by different analyzers is not valid. Changes in gastric intramucosal pH of 0.06 pH units can be detected by most analyzers in the clinically relevant PCO2 level.