Interchangeability of blood gas, electrolyte and metabolite results measured with point-of-care, blood gas and core laboratory analyzers

Abstract

Background: The random use of point-of-care, blood gas and core laboratory analyzers to measure electrolytes and metabolites increases the variability in test results. This study was designed to determine whether these results performed on whole blood (point-of-care and blood gas) and plasma (core laboratory) platforms are interchangeable without a risk of clinically relevant discrepancies.

Methods: The interchangeability of the blood gas analysis, electrolytes, glucose, lactate and hemoglobin results performed with three stat platforms (i-STAT, Radiometer ABL 825, RapidLab 865) and two core laboratory platforms (Roche Modular P800 and Sysmex XE-2100) were evaluated using samples from critically ill patients.

Results: For pH, pCO(2), pO(2) and Ca(2+), good correlation (r-values 0.96-1.00) was observed for all comparative analyzers and the biases were within clinically acceptable limits. Potassium, sodium, glucose, lactate and hemoglobin measured with stat analyzers was highly correlated with measurements performed using the laboratory analyzers, r-values 0.89-1.00 and slopes 0.83-1.07. Mean differences with significant bias (p<0.0001) were found for sodium with blood gas analyzers and hemoglobin with i-STAT.

Conclusions: The blood gas, K(+), Na(+), Ca(2+), glucose and lactate results measured with stat and core laboratory analyzers can be used in different clinical settings for critical care management. However, when monitoring small changes in sodium concentrations, the use of single analyzer is encouraged to avoid analytical differences (acceptance limit ± 2%) and misinterpretation of results measured with multiple analyzers. The users of the i-STAT at low hemoglobin values overdiagnose anaemia. Thus, prior to transfusion, the use of hemoglobin concentrations measured with laboratory analyzers is preferable.