Extended clinical trial and evaluation of glucose determination with the Eastman Kodak Ektachem GLU/BUN Analyzer

We followed the "abbreviated precision protocol" of the National Committee for Clinical Laboratory Standards for the evaluation of precision, accuracy, and carryover in analyses for glucose with the "Ektachem." We analyzed 760 clinical samples by this technique, by the FDA Proposed Class Standard glucose reference method, and with the Beckman System I GLU/BUN Analyzer. Precision and accuracy were estimated for 500, 1000, 1200, 1500, and 3000 mg/L glucose concentrations in 100, 30, or 20 microL of serum or plasma. Potential interference of 19 compounds was evaluated. Random error (1.965 X SD) was 22, 30, 34, 40, and 88 mg/L. Systematic error was 8, 1.5, -2, -5, and -27 mg/L. Total analytical error was 30, 32, 36, 46, and 110 mg/L for analysis of 100 microL of serum at the above-stated glucose concentrations. The greatest interference (-39 mg/L) in the glucose (at 1200 mg/L) determination was caused by L-ascorbate (40 mg/L). Glucose concentrations as determined with the Ektachem were found to be linearly related to the expected concentration up to at least 5660 mg/L. Carryover was not statistically significant.