Performance and reliability of two frequently used point-of-care blood gas analyzers at 423 and 4,559 m

Abstract

Background and ObjectivesBlood gas analyzers (BGA) aid medical decision-making. Their specified performance criteria are based on sea level conditions. However, millions of people are living at high altitude (HA) where the performance of BGAs is poorly characterized. We investigated the effect of exposure to 4,559 m on the reliability and robustness of two BGAs widely used at HA. MethodsIn this prospective study arterial blood samples from 13 volunteers (2 female) with susceptibility to the development of high-altitude pulmonary edema were collected once near sea level at 423 m (nSL423) and three times at high altitude (HA4,559). Samples were measured in triplicate with the cartridge BGAs Rapidpoint 500 (SIE; Siemens Healthcare) and the ABL90 (RAD; Radiometer) to calculate coefficients of variation (CV) and intraclass correlation coefficients (ICC) within a mixed model. ResultsAt nSL423 and HA4,559, 3% and 17% of all data were not reported with SIE, mainly due to clotting of the sample caused by delays because of the frequent automated calibration routines. No data were missing with RAD. ICCs were not significantly lower (mean (min–max) 0.87 (0.68–0.98) vs. 0.94 (0.84–1.00); p = 0.217) with SIE at nSL423, but significantly lower at HA4,559 (0.87 (0.49–1.00) vs. 0.99 (0.96–1.00); p = 0.025). All CVs, except that for arterial oxygen saturation at HA4,559,were higher with SIE . ConclusionIn this study, the reliability of RAD was superior to SIE at nSL423 and HA4,559. In contrast to RAD, the performance of SIE declined at HA4,559. SIE was more prone to not reporting all variables, especially at HA4559.