| Abstrakt: |
We evaluated the accuracy of multiple point-of-care devices used for glucose determination in hospital inpatients compared with a plasma hexokinase reference method. The point-of-care whole blood glucose methods included the i-STAT G cartridge (i-STAT Corporation, East Windsor, NJ), Roche Accu-Chek Inform meter (Roche Diagnostics, Indianapolis, Ind), HemoCue -glucose and HemoCue 201 glucose analyzers (HemoCue, Angelholm, Sweden), and the Radiometer ABL 725 (Radiometer Medical A/S, Copenhagen, Denmark). Plasma glucose on the Integra 400 (Roche Diagnostics) was used as the reference method. The Radiometer and i-STAT displayed the best overall performance as compared with the reference method based on slope, intercept, median bias, and percent values within 10% (for reference values >100 mg/dL) or 10 mg/dL (for reference values ≤100 mg/dL) of the reference method. Both HemoCue -glucose and Roche Accu-Chek Inform produced more values that exceeded 10% or 10 mg/dL of the reference value compared with Radiometer or i-STAT. The HemoCue 201 produced consistently high glucose values compared with the reference method and demonstrated poor overall performance. |
