[Portable capnographs in emergency care: a comparison of equipment]

Recently, transportable capnographs fulfilling the practical demands of emergency medicine have become commercially available (NPB-75, sidestream, Nellcor Puritan Bennett; and Tidal Wave, mainstream, Novametrix). A prerequisite for their use is an accuracy as required for clinical purposes (i.e., pCO2 +/- 2 mmHg). Additionally, environmental conditions in emergency medicine (e.g., changes in ambient temperature) should not have a significant impact on accuracy. The objective of this investigation was to analyse the accuracy of the two capnographs. The accuracy of the pCO2 measurement was evaluated under the following conditions: (1) measurement with three gas mixtures of defined concentrations (gas A: 5% CO2, 95% O2; gas B: 5% CO2, 20% O2, 75% N2; gas C: 10% CO2, 90% N2) related to STPD conditions (STPD = Standard Temperature and Pressure, Dry); and (2) exposure to changes in temperature (from +22 degrees C to -20 degrees C, and from -20 degrees C to +22 degrees C) applying the aforementioned 3 gas mixtures (STPD); and (3) in 20 patients manually ventilated with pure oxygen following endotracheal intubation (i.e., BTPS conditions = body temperature and pressure, saturated). Adequacy of the results was compared to the alveolar gas monitor AGM 1304 (Bruel & Kjaer, Copenhagen, Denmark; sidestream) which served as the reference method (providing an accuracy for the alveolar carbon dioxide partial pressure (pACO2) of +/- 1 mmHg). In the 3 dry gas mixtures, mean inaccuracy proved to be +4.5 +/- 4.1, +2.8 +/- 3.7, and +2.2 +/- 7.0 mmHg (gas A, gas B, gas C; STPD) with the Nellcor sidestream device. Using the Novametrix mainstream capnograph the results were found as follows: (1) -1.1 +/- 0.6, +2.9 +/- 0.6, and +5.6 +/- 2.3 mmHg (oxygen compensation enabled); and (2) +0.2 +/- 1.6, +2.2 +/- 0.6, and +3.2 +/- 4.2 mmHg (oxygen compensation disabled). After changing the environmental temperature (-20 degrees C / +22 degrees C), the resulting deviations (gases A-C, STPD) found with the Nellcor device averaged -12 +/- 4% and +15 +/- 3% (Nellcor); with the Novametrix mainstream device the deviations averaged -1 +/- 2% and +1 +/- 1%, and -2 +/- 1% and +1 +/- 1% (oxygen compensation enabled/disabled). Mean inaccuracy of the pCO2 measurement during ventilation of patients with pure oxygen (BTPS) was found to average -0.9 +/- 0.9 (Nellcor), and either +3.9 +/- 0.8 or +2.1 +/- 0.7 mmHg with the Novametrix (oxygen compensation enabled/disabled). Under BTPS conditions, both devices showed an acceptable deviation of the measurement accuracy up to a maximum of +/- 2 mmHg. The higher deviations of the "NPB-75" (Nellcor Puritan Bennett, sidestream) when using dry gas mixtures (STPD) may be explained by the automatic water vapour correction. Under the conditions of low and changing ambient temperature (-20 degrees C, +22 degrees C), only the "Tidal Wave" (Novametrix; mainstream) remained uninfluenced, whereas deviations of -12% and +15% were found with the "NPB-75".