Abstract
The possible interference of haematocrit was studied in an evaluation of the analytical performance of the NOVA 11 analyzer. For lithium concentrations between 0.56 and 0.68 mmol/L, the within-run imprecision was less than 1.06%, and less than 2.8% in day-to-day imprecision, measured in quality control solutions. Comparison of whole blood with plasma measurement showed satisfactory correlation with deviations up to 0.06 mmol/L for haematocrits from 35–53% (y = 0.73 + 0.12). Measurement on a blood specimen not containing lithium, whose haematocrit was varied, allowed us to quantify a specific red cell interference. With samples from treated patients, and a sample spiked to various lithium concentrations, we obtained differences dependent on haematocrit and on lithium concentrations. In subtracting the specific red cell effect, we could systematically observe that apparent lithium levels diminish as haematocrit increases. This observation should allow development of an appropriate algorithm for obtaining exact values on whole blood samples.
Key Words: