Capnography monitoring in nonintubated patients with respiratory distress

Objective: The objective was to assess agreement between end-tidal carbon dioxide values measured by a handheld capnometer (PETCO2) and values measured by a blood gas analyzer (PaCO2) in nonintubated patients with respiratory distress in an out-of-hospital setting. Methods: This prospective study compared PETCO2 values obtained by an end-tidal capnometer (Microcap Plus; Oridion Capnography Inc, Needham, Mass) to PaCO2 values by the Bland and Altman statistical method. Results: A total of 50 patients were included. Continuous PETCO2 monitoring was easily performed in all 50 patients during ambulance transport, but blood gas analysis failed in I patient. Agreement between the 2 methods was poor with a bias (mean difference) between PETCO2 and PaCO2 measurements of 12 mm Hg and a precision (SD of the difference) of 8 mm Hg. The gradient between PETCO2 and PaCO2 was greater than 5 and 10 mm Hg in 41 and 25 patients, respectively. Conclusions: PETCO2 measurements poorly reflected PaCO2 values in our population of nonintubated patients with respiratory distress of various origins. (C) 2009 Elsevier Inc. All rights reserved.