Not all blood glucose monitoring systems (BGMSs) that meet ISO 15917:2013 or FDA 2016 accuracy criteria yield accurate results in the low blood glucose range (LBGR), defined as BG below 70 mg/dl. Accuracy in the LBGR is important for hypoglycemia detection/management. This post hoc analysis utilizes probability methodology to estimate BGMS performance in the LBGR.
All data were computed from capillary blood samples from diabetes patients. CONTOUR®PLUS BGMS (CP) data were derived from a previous study comparing the accuracy of five BGMSs; CONTOUR®PLUS ONE BGMS (CPO) data were from a separate study. To estimate likelihood of accurate BGMS results (±15% of reference values) in range of 20–450 mg/dl, probability curves were computed using linear regression models. with BGMS results expressed as a function of laboratory data. The 95% ranges of BGMS results expected at specific reference values (40, 54, 60 and 70 mg/dl) were also computed.
In the LBGR probability curves demonstrated that CP and CPO BGMSs maintained high accuracy. The 95% ranges of BGMS results at specific reference values indicated that some BGMSs would have a low likelihood of having 95% of results within ±15 % of reference values, particularly in the LBGR.
In this analysis, the CP and CPO BGMSs were highly accurate in the LBGR, which is important for safe and effective diabetes management, especially in insulin-treated patients, patients with history of severe hypoglycemia or hypoglycemia unawareness, diabetes during pregnancy and patients using CGM and FSL, when BGM is recommended.