Abstract
OBJECTIVE: In a recent clinical study on the reliability of a point-of-care (POC) analyzer, we described a downward bias in hematocrit measurement during cardiopulmonary bypass leading potentially to overtreatment with packed red cells. We hypothesized that the detected deviation is caused by variations in electrolyte concentration rather than by colloids used. METHODS: Blood was sampled from patients before cardiac surgery to obtain undiluted anticoagulated whole blood samples (n = 53). From each sample, four dilution series covering a hematocrit range of 15-30% were made using NaCl (0.9%), modified gelatine (4%), hydroxyethylstarch (6%), or a potassium-based (16 mEq/l) solution, respectively. In each dilution series, hematocrit was measured by POC and via the "golden standard" microcentrifugal method to determine whether the deviation of the POC-analyzer to the microcentrifuge was dependent on the type and dilution level of the solution used. RESULTS: In contrast to the colloid-based dilution series, the crystalloids revealed a significant downward bias of the POC-analyzer with respect to the microcentrifuge (p < 0.05). Due to the correction algorithm for sodium in the POC-analyzer, this deviation was nearly constant for NaCl (mean of difference: -1.8 +/- 0.1%), but increased significantly in case of the potassium-based solution (up to -8.2 +/- 0.4% after 1.5-times dilution). The starch- and gelatine-based solutions led to a significant upward bias (p < 0.05) that increased with progressing dilution (up to 1.2 +/- 0.1% for hydroxyethylstarch and up to 1.3 +/- 0.1% for modified gelatine after 1.5-times dilution). CONCLUSIONS: Conductivity-based POC hematocrit measurement suffers from biases due to changes of the plasma constituents. The downward bias in hematocrit as often seen during cardiopulmonary bypass is driven by changes of different electrolyte concentration rather than by colloids used per se.
