To assess the performance of 4 common glucometers used for point-of-care (POC) testing both in home care and hospital care in Brazil.
Glucose results from the same specimen (whole blood, no anticoagulant) was simultaneously obtained with the Accu-Check Advantage (Roche Diagnostics), Glucocard (Menarini), Precision PCx and Precision QID (Abbott Diagnostics). These results were compared with that from a paired specimen (plasma), collected at the same time using fluoride as anticoagulant and tested in the analyzer (Roche/Hitachi 917), defined as the reference method.
The statistical analysis showed good correlation (R > 0.98) of all glucometers with the reference method in this study. When compared with the laboratory analyzer, 3 out of 4 meters (the exception was the QID), met the requirements from NCCLS C30-A guideline. Based on this, three POC instruments exhibited acceptable performance for the measurement of glucose levels both at home (Accu-Check Advantage and Glucocard) and hospital (PCx). Multiple instrument comparison (MIC) of the 4 glucometers was also performed using the MIC Module from EP Evaluator. In this study, the allowable error was set as the CLIA's proficiency test limit for glucose (10%). All glucometers had at least 1 result that exceeded the CLIA's criterion.
Despite of the acceptable performance when compared with the laboratory analyzer, there is a long way to go to harmonize glucose results from different glucometers, not only among different manufacturers, but also among different brands from the same manufacturer.