Commonly used blood glucose monitors may not always be accurate, study indicates

Researchers assessed the accuracy of 18 FDA-approved blood glucose monitoring systems, representing approximately 90% of those available in the U.S. from 2013 to 2015.


Only 6 of 18 commercially available blood glucose monitors (BGMs) consistently met accuracy standards, a recent study found.

Researchers assessed the accuracy of 18 FDA-approved BGM systems, representing approximately 90% of those available in the U.S. from 2013 to 2015. Study participants at three clinical sites had a capillary blood glucose level measured on six different systems and had a reference capillary sample prepped for plasma testing at a reference laboratory. Three triple-blind studies were performed at each clinical site, with a different set of six systems tested each time. The researchers defined a compliant blood glucose result as one that fell within 15% of a reference plasma value for values of 100 mg/dL or higher (≥5.55 mmol/L) or within 15 mg/dL (0.83 mmol/L) for values below 100 mg/dL (<5.55 mmol/L). The proportion of compliant readings in each study was then compared with a predetermined standard of accuracy, similar to current regulatory standards but more lenient. The researchers also measured overall compliance, extreme outliers that differed from the reference value by more than 20%, and readings with no clinical risk, among other variables. The results of the study, which was funded by a grant from Abbott Diabetes Care, were published online June 13 by Diabetes Care.

A total of 1,035 participants were recruited into the study. Over half (54.1%) were women, 82.4% were white, and the mean age was 51.5 years. Participants were excluded if they had hemophilia or another bleeding disorder, if they were pregnant, or if they had a condition that could put the participants or the study at risk. Overall, 1,032 participants completed the study (one participant withdrew before testing, and two plasma specimens were lost during shipping), and each BGM system was tested on an average of 115 people. Regardless of which accuracy metric was used, rankings of BGM systems were almost identical. The predetermined accuracy standard was met by six systems in all three studies, five systems in two studies, and three systems in one study. Four of the systems tested never met the accuracy standard.

The authors noted that their results could have been affected by the exclusion of 7.8% of the reference samples, among other limitations. However, based on their findings, they concluded that commercially available BGMs do not always meet accuracy standards and noted that this can have consequences for patients. “Since patients depend on their BGMs for day-to-day management, lack of accuracy may put patients at risk for both hypoglycemia and hyperglycemia,” the authors wrote. “We hope that this study will provide objective and validated information for patients, health care professionals, and payers to make informed product selection.” The authors also called for better evaluation of postmarketing performance of these and similar products by regulators.