Comparative System Accuracy of Six Blood Glucose Monitoring System for Self-Use

Ongoing standardized post-market evaluation of blood glucose monitoring systems (BGMS) is both clinically important to confirm appropriate performance and mandatory under the European Union In Vitro Diagnostics Regulation. Publication and accessibility of studies on device performance has therefore become a key component of evidence-based purchasing decision for distributors and users.

In the present study, a comparative post-market evaluation of 6 BGMS was performed in accordance with ISO 15197:2015 ¹ protocols and requirements (Table 1). All devices, test strips, and solutions were supplied by the respective manufacturer. Duplicate glucose measurements of native and manipulated samples (ie, glycolysis at 37°C for 3-4 hours, glucose supplementation of a 40% glucose solution) of capillary whole blood or plasma were performed using the blood glucose monitor and one of two reference methods (see Table 1 for details). For the data analysis, whole blood measurements were converted to plasma equivalents via hematocrit based numeric conversion, if needed. Trueness and precision of the BGMS and reference assays were verified using certified bioanalytical standards and manufacturer issued control solution. Based on the in- and exclusion criteria established by the manufacturers and appendix A of DIN EN ISO 15197:2015, 600 subjects with diabetes totaling 3600 measurements were included in the analysis.

OPEN IN VIEWER

Table 1.

Test Systems Evaluated and System Accuracy Results.

BGMS	Reagent	Sample	Reference method	Within ± 15 mg/dL/15%	Glucose < 100 mg/dL			Glucose ≥ 100 mg/dL			Bias	MARD
					±5 mg/dL	±10 mg/dL	±15 mg/dL	±5%	±10%	±15%
				%	%	%	%	%	%	%	%	%
JT100, Tyson Bioresearch	GDH-FAD	Whole blood	Glucoseoxidase, YSI 2300	98.2	53.1	92.2	99.0	21.1	64.5	97.8	-7.1	8.1
Dextra, IME-DC	GDH-FAD	Plasma	Glucoseoxidase, YSI 2300	97.5	58.9	84.5	97.6	50.2	83.1	97.5	2.5	6.3
GL60, Beurer	GDH-FAD	Plasma	Glucoseoxidase, YSI 2300	99.2	73.2	94.0	99.4	63.4	89.4	99.1	2.5	4.9
GlucoCheck Gold, aktivmed	GDH-FAD	Plasma	Hexokinase, Cobas c111	98.5	88.2	100	100	47.1	87.3	98.0	-2.1	5.5
GlucoCheck Excellent, aktivmed	GOD	Plasma	Hexokinase, Cobas c111	98.3	47.2	88.3	98.9	43.1	79.8	98.1	-2.8	6.9
GlucoCheck XL, aktivmed	GDH-FAD	Plasma	Hexokinase, Cobas c111	98.8	50.0	96.2	100	41.1	77.1	98.3	-5.6	6.9

Abbreviation: BGMS, blood glucose monitoring system; MARD, mean absolute relative difference; GDHFAD, glucose dehydrogenase-flavin adenine dinucleotide; GOD, glucose oxidase.

All six devices fulfilled the acceptance criteria pursuant to the directive guidelines. The measurements of each device lie in zones A and B of the Parkes/consensus error grid for patients with type 1 and type 2 diabetes. ² Between 97.5% and 99.2% of measurements showed deviations of less than ±15 mg/dL and ±15% for glucose concentrations <100 mg/dL and ≥100 mg/dL, respectively (Table 1). All devices performed marginally better at lower concentrations. The BGMS evaluated with whole blood had the highest bias (-7.1%) and mean absolute relative difference (MARD) to the reference method (8.1%).

Yet, a direct comparison of BGMS using plasma or whole blood is not always readily practicable for a variety reasons. Foremost is the often poor overlap of sample material from manufacturer’s development and performance studies with the patient population used in post market follow-up studies with the actual patient population, be it demographic (age), physiological (hematocrit), or others. ³ Also, the basis of calibration differs between monitors depending on the manufacturers enzymatic assay for whole blood or plasma measurements. However, we found no significant effect of sample type, test strip reagent, or reference method on mean absolute relative difference (MARD, Table 1), bias or overall compliance with acceptance requirements.