Analysis of bias in measurements of potassium, sodium and hemoglobin by an emergency department-based blood gas analyzer relative to hospital laboratory autoanalyzer results.

OBJECTIVE

The emergency departments (EDs) of Chinese hospitals are gradually being equipped with blood gas machines. These machines, along with the measurement of biochemical markers by the hospital laboratory, facilitate the care of patients with severe conditions who present to the ED. However, discrepancies have been noted between the Arterial Blood Gas (ABG) analyzers in the ED and the hospital laboratory autoanalyzer in relation to electrolyte and hemoglobin measurements. The present study was performed to determine whether the ABG and laboratory measurements of potassium, sodium, and hemoglobin levels are equivalent, and whether ABG analyzer results can be used to guide clinical care before the laboratory results become available.

MATERIALS AND METHODS

Study power analyses revealed that 200 consecutive patients who presented to our ED would allow this prospective single-center cohort study to detect significant differences between ABG- and laboratory-measured potassium, sodium, and hemoglobin levels. Paired arterial and venous blood samples were collected within 30 minutes. Arterial blood samples were measured in the ED by an ABL 90 FLEX blood gas analyzer. The biochemistry and blood cell counts of the venous samples were measured in the hospital laboratory. The potassium, sodium, and hemoglobin concentrations obtained by both methods were compared by using paired Student's t-test, Spearman's correlation, Bland-Altman plots, and Deming regression.

RESULTS

The mean ABG and laboratory potassium values were 3.77±0.44 and 4.2±0.55, respectively (P<0.0001). The mean ABG and laboratory sodium values were 137.89±5.44 and 140.93±5.50, respectively (P<0.0001). The mean ABG and laboratory Hemoglobin values were 12.28±2.62 and 12.35±2.60, respectively (P = 0.24).

CONCLUSION

Although there are the statistical difference and acceptable biases between ABG- and laboratory-measured potassium and sodium, the biases do not exceed USCLIA-determined limits. In parallel, there are no statistical differences and biases beyond USCLIA-determined limits between ABG- and laboratory-measured hemoglobin. Therefore, all three variables measured by ABG were reliable.