Improving the quality of self-monitoring blood glucose measurement: a study in reducing calibration errors.

BACKGROUND: Two independent studies reported that 16% of people who self-monitor blood glucose used incorrectly coded meters. The degree of analytical error, however, was not characterized. Our study objectives were to demonstrate that miscoding can cause analytical errors and to characterize the potential amount of bias that can occur. The impact of calibration error with three selfblood glucose monitoring systems (BGMSs), one of which has an autocoding feature, is reported. METHODS: Fresh capillary fingerstick blood from 50 subjects, 18 men and 32 women ranging in age from 23 to 82 years, was used to measure glucose with three BGMSs. Two BGMSs required manual coding and were purposely miscoded using numbers different from the one recommended for the reagent lot used. Two properly coded meters of each BGMS were included to assess within-system variability. Different reagent lots were used to challenge a third system that had autocoding capability and could not be miscoded. RESULTS: Some within-system comparisons showed deviations of greater than +/-30% when results obtained with miscoded meters were compared with data obtained with ones programmed using the correct code number. Similar erroneous results were found when the miscoded meter results were compared with those obtained with a glucose analyzer. For some miscoded meter and test strip combinations, error grid analysis showed that 90% of results fell into zones indicating altered clinical action. Such inaccuracies were not found with the BGMS having the autocoding feature. CONCLUSIONS: When certain meter code number settings of two BGMSs were used in conjunction with test strips having code numbers that did not match, statistically and clinically inaccurate results were obtained. Coding errors resulted in analytical errors of greater than +/-30% (-31.6 to +60.9%). These results confirm the value of a BGMS with an automatic coding feature.

Evaluation of a new blood glucose monitoring system with auto-calibration.

BACKGROUND: The Ascensia CONTOUR System (Bayer HealthCare LLC, Elkhart, IN) is a new blood glucose (BG) monitoring system (BGMS) that uses glucose dehydrogenase chemistry and that combines low sample volume, fast response time, an auto-control marking feature, and an auto-calibration function in a small package. METHODS: The system was evaluated at four diabetes clinics with ambulatory subjects who had diabetes. The BGMS was tested by both health care professionals (HCPs) and lay users. It was also evaluated at high altitude (10,200 feet) and, in conjunction with the Ascensia MICROLET VACULANCE (Bayer), for use with samples obtained from four alternative anatomical sites. RESULTS: When the system was used with blood from fingersticks, both lay users and HCPs obtained results that fulfilled the International Organization for Standardization (ISO) 15197:2003 accuracy criteria, which advocate that 95% of BG measurements should fall within +/-15 mg/dL (for BG concentrations <75 mg/dL) or +/-20% (for BG concentrations > or = 75 mg/dL) of the laboratory value. Lay users obtained 96.9% of glucose results, and HCPs obtained 96.4% of glucose results, within ISO accuracy criteria. People with diabetes who had no prior experience using the system also obtained acceptable results (97.0%), as did the HCP using fingerstick samples from subjects in the high altitude study (97.0%). Among the alternative anatomical sites tested, only the results from the palm met the ISO criteria (97.5%), although the results from all four sites (palm, thigh, abdomen, and forearm) were clinically acceptable when assessed using error grid analysis. Most subjects rated the BGMS as either excellent or very good in a questionnaire, and were able to use it properly without training. CONCLUSIONS: These findings indicate that the Ascensia CONTOUR System, which does not require calibration by the user, is a convenient and accurate instrument with useful features for people who routinely monitor BG.