Evaluation of System Accuracy and Range of Error of the Blood Glucose Monitoring System Contour Plus Elite Against the ISO 15197 Requirements for Self-Testing

Abstract

Introduction:

Accuracy of blood glucose monitoring systems (BGMS) is critical to self-detection of dysglycemia by people with diabetes (PWD). Reliable BGMS are integral to diabetes care, contributing to continuous metabolic control and patient understanding of behaviors, allowing for flexible adjustment of therapy. Accuracy, availability, and acceptability of BGMS can impact clinical decisions and outcomes.

Methods:

Conducted in accordance with DIN EN ISO 15197:2015, this study evaluated the accuracy of the Contour Plus Elite BGMS in terms of agreement between test system measurements and reference system measurements across a predetermined range of blood glucose concentrations. A diverse group of 104 adults from a single center participated. Analysis included the production of consensus error grids (CEGs) to evaluate the clinical significance of deviations of results between the test and reference values.

Results:

All reference and test BGMS measurements were performed according to their specifications and requirements. A total of 600 of 600 (100%) Contour Plus Elite measurements across 3 lots showed deviations from the reference test of less than 15 mg/dL (0.83 mmol/L) for glucose concentrations <100 mg/dL (5.55 mmol/L) and less than 15% glucose concentrations ≥100 mg/dL (5.55 mmol/L), complying with the first DIN EN ISO 15197:2015 acceptance criterion. All Contour Plus Elite measurements (100%) resided within zone A of the CEG, thus exceeding the second criterion, which permits measurements in zone A or B. An ad hoc confidence analysis provided an additional metric, demonstrating that the smallest error range within which at least 95% of differences between test device and reference measurements fell was ±8.1 mg/dL or 8.1%.

Conclusions:

The evaluation revealed that the BGMS Contour Plus Elite fully complied with and exceeded DIN EN ISO 15197:2015 system accuracy requirements. The device displays a clinically acceptable level of accuracy for use by PWD. All measurements (100%) were well within acceptability levels.

Introduction

The major goal of modern diabetes therapy is to achieve and maintain euglycemic blood glucose concentrations to prevent the development of diabetic complications. ¹ Accuracy and precision of blood glucose monitoring systems (BGMS) are critical to achievement of optimal glycemic control in a self-management regimen.^1,2 Self-detection and control of dysglycemic episodes can assist in managing HbA1c levels and reducing the risk of acute and chronic adverse diabetes-related outcomes. ³ Reliable BGMS that are suitable for patient use are an integral part of diabetes care and contribute to continuous metabolic control and patient understanding of the impact of behaviors, allowing for flexible adjustment of therapy.^1,4

The Contour Plus Elite BGMS (Ascensia Deutschland GmbH) is a CE-certified device intended for blood glucose measurement by people with diabetes (PWD) or health care professionals (HCPs) using a fingertip capillary sample. The features of this BGMS have been described previously. ⁵ Variations in performance of BGMS can be driven by several factors including sample type, reference method, and technical aspects of BGMS.^2,6 Patient-specific characteristics such as concurrent medical conditions, demographic factors, technique, and individual physiological variations like hematocrit may also influence BGMS performance.^6,7 The harmonized International Standard DIN EN ISO 15197:2015 defines specific quality standards for the accuracy of BGMS for self-monitoring. ⁶

In accordance with DIN EN ISO 15197:2015, this study evaluated the system accuracy of the Contour Plus Elite BGMS in terms of closeness of agreement between a set of representative results and their respective reference values determined by the validated Cobas c111 system (Roche Diagnostics GmbH, Mannheim, Germany). The study also included calculation of the ad hoc confidence metric, which quantifies the narrowest error margin within which at least 95% of differences between test BGMS results and reference BGMS results fall. ²

Methods

Study design and participants

This prospective study enrolled adults aged over 18 years with hypo-, eu- or hyperglycemia. The study protocol was reviewed and approved by an ethics committee and participation was voluntary and based upon informed consent. In accordance with the requirements of the standard, 100 inpatients from the Karlsburg Hospital, Germany, were recruited and provided a short disclosure of medical history conditions and medication intake. Exclusions were pregnancy or lactation, as well as existing conditions that could pose a risk to study personnel or be aggravated by the sampling procedure. Participants attended one study visit. Participants were monitored for safety events.

Test methods

Devices

Testing was conducted according to the manufacturer operating instructions for the Contour Plus Elite and Cobas c111. Two Contour Plus Elite test devices and Contour Plus test strips from 11 packs from each of three different lots were used. Cobas c111 is a hexokinase laboratory reference test that operates in the blood glucose range 1.98 to 720 mg/dL (0.11 to 40 mmol/L) (Roche Diagnostics Glucose Hexokinase enzyme kit (Gluc2) package insert). Calibration and control of accuracy and precision of the reference system were performed by use of manufacturer-issued standards with glucose concentrations of 104, 202, and 248 mg/dL, respectively, on each study day, before and after the test series.

Acceptance criteria

In accordance with DIN EN ISO 15197:2015, ⁶ the minimum test system accuracy required at least 95% of all measurements to fall within ±15mg/dL (0.83 mmol/L) for glucose concentrations <100 mg/dL (5.55 mmol/L) and within ±15% of reference measurements for glucose measurements ≥100 mg/dL. In addition, 99% of the individual glucose measurements must fall within zones A and B of the consensus error grid (CEG) to meet the ISO criteria.

Sample acquisition and preparation

Capillary blood samples of 200 µL were collected from participants by qualified personnel following strict hygienic requirements and according to the system instructions for use. To provide test samples in the required ranges of ≤50 mg/dL (≤2.77 mmol/L) and >400 mg/dL (22.2 mmol/L), ⁶ sufficient samples were altered to increase or decrease the blood glucose concentration by adding 40% glucose solution or incubating in a shaking water bath at 37°C, respectively.

Test procedure

System accuracy was evaluated in simulated conditions of intended use, with ambient temperature 23°C ± 5°C and the relative humidity below 85%. Reference sample blood glucose measurement with the Cobas c111 system was completed immediately before and after each BGM test with a Contour Plus Elite test strip. The evaluation of each participant sample was conducted using duplicate measurements on two devices with three different test strip lots. Both reference values were evaluated to verify sample stability, requiring a difference in reference values of no more than 4 mg/dL (0.22 mmol/L) at a glucose concentration of <100 mg/dL (5.55 mmol/L), and no more than 4% at a glucose concentration of ≥100 mg/dL (5.55 mmol/L). If these values indicated an unacceptable change in the glucose concentration, the test sample was excluded and replaced with another sample from the same glucose concentration range.

Analysis

Patient characteristics including gender, age, and diabetes type were collected for the purpose of assessment of the blood sample only; demographics were not analyzed. Only physiological data relevant for device evaluation (blood glucose level and hematocrit) are reported.

Analysis and potential exclusion of data for the evaluation of Contour Plus Elite test accuracy was conducted according to DIN EN ISO 15197:2015. ⁶

A descriptive summary of analysis populations by glucose concentration ranges above and below 100 mg/dL (5.55 mmol/L), respectively, and a Bland–Altman diagram was prepared. CEGs were produced to evaluate the clinical significance of deviations of results between the index and reference tests.^6,8 The ad hoc analysis to obtain a percentage of the smallest error range, within which at least 95% of the measurements between the test device and the reference measurements fell, was calculated. This ad hoc confidence analysis is an additional metric for the accuracy of the test device compared with the reference method. ²

Results

Participants and samples

A total of 104 patients completed study visits in March 2025. The samples from four patients were excluded from analysis due to sufficient samples already being obtained in the respective glucose range (n = 1) and nonadherence to inclusion or exclusion criteria (n = 3). During the study, the mean ambient room temperature was 23.6°C (22.2°C to 25.5°C) and relative humidity ranged between 17% and 32%, within the required specifications for both the reference and index tests. Mean (SD) hematocrit was 42.43 (4.0)%. Of 100 samples, 78 met the blood glucose level range proportions naturally; the remaining samples were altered according to standard protocols to provide the required range of hypo- and hyper-glycemic conditions.

Control measures

Following the manufacturer’s instructions for control measures, all reference and test BGM systems performed according to their specifications and requirements on each study day; no tests or samples had to be replaced.

Contour Plus Elite system accuracy

Across all three Contour Plus Elite lots and the range of blood glucose concentrations, a total of 600 of 600 (100%) measurements showed deviations from the reference test of less than 15 mg/dL (0.83 mmol/L) and 15%, respectively (Table 1 and Fig. 1). The complete dataset is available in the Supplementary Data.

FIG. 1.

Bland–Altman diagram for all lots. The dotted lines indicate acceptance limits. BGM, blood glucose monitoring system Conour Plus Elite. Reference = Cobas C111.

Table 1.

Results of System Accuracy Measurements for ISO 15197 Compliance

Glucose concentrations <100 mg/dL (5.55 mmol/L)
N	Within ±5 mg/dL (0.43 mmol/L)		Within ±10 mg/dL (0.83 mmol/L)		Within ±15 mg/dL (1.25 mmol/L)
156	143	91.7%	154	98.7%	156	100.0%
Glucose concentrations ≥100 mg/dL (5.55 mmol/L)
N	Within ±5%		Within ±10%		Within ±15%
444	326	73.4%	433	97.5%	444	100.0%

All Contour Plus Elite measurements (100%) reside within zone A of the CEG (Fig. 2).

FIG. 2.

Consensus error grid for all lots. The dotted lines indicate acceptance limit bands. BGM, blood glucose monitoring system Conour Plus Elite. Reference = Cobas C111.

Ad hoc confidence metric

The smallest error range within which at least 95% of differences between test device and reference measurements fell was ±8.1 mg/dL or 8.1% of the reference results, based on the ad-hoc analysis. The mean absolute relative difference (MARD) in performance was calculated at 3.3%.

The Ascensia BGMS Contour Plus Elite was deemed to fully comply with or exceed the DIN EN ISO 15197:2015 system accuracy acceptance criteria.

Safety

Within the scope of this study, there were no recorded instances of adverse events, serious adverse events. There were no device deficiencies—including any malfunctions or other device-measurement or patient-related incidents that would require vigilance actions.

Discussion

The Contour Plus Elite BGMS was evaluated for system accuracy of glucose measurement in a manner that simulates the intended use by PWD. All measurements with the Contour Plus Elite (600 of 600; 100%) showed blood glucose deviations from the reference Cobas C111 system of less than 15 mg/dL (0.83 mmol/L) and 15%, respectively. Therefore, the test system reached a clinically acceptable level of accuracy, complying with DIN EN ISO 15197:2015 acceptance criterion 1. ⁶ Furthermore, all measurements fell tightly within zone A of the CEG. This indicates no effect on clinical action, exceeding acceptance criterion 2, which permits measurements in zone A or B (no effect or altered clinical action with little or no effect on clinical outcome, respectively). ⁸

Study participants had mean (range) hematocrit of 42.43% (28%–56%). Seven samples were outside the suggested standard range for packed cell volumes (35%–50%) for self-testing of blood glucose. ⁶ Lower hematocrit values (<35%) can result in overestimates of glucose levels using blood glucose meters, whereas higher hematocrit values (>45%) can result in underestimates. ⁹ Despite this, 100% of measurements were within the required range for system accuracy, consistent with reports that CONTOUR® PLUS test strip results are not significantly affected by hematocrit in the range of 0%–70%. ⁵

In the second edition of the international standards for self-monitoring BGMS (DIN EN ISO 15197:2015), acceptance criteria were made more stringent by inclusion of the CEG. This change was possible following improvements in glucose meter technology since the first edition.^5,6 FDA home blood glucose meter accuracy standards require that 95% of measurements fall within 15% for all blood glucose tests in the usable range and 99% of measurements fall within 20% for all blood glucose tests in the usable range. ¹⁰ These standards reflect the importance of accuracy of available self-monitoring devices for PWD.^10,11 The integration and improvement of technology in diabetes care continue to transform the interaction between individuals with diabetes and their health care providers. ¹¹ In addition to meeting and exceeding these standards, the Contour Plus Elite BGMS is designed to synchronize with the CONTOUR® DIABETES app on a smartphone or tablet. ⁵

The narrowest error range within which at least 95% of inaccuracies between test device and reference measurements was 8.1 mg/dL/%, consistent with the ad hoc confidence metric of 8.1 mg/dL/% previously reported by our group for a related Contour® device. ² The MARD of 3.3% in this study was also similar to the previously reported value of 3.2. ² Our study has added to the consistently excellent compliance and usability profile of the Contour® family. ¹²

Previous studies have demonstrated that BGMS bias may affect clinical outcomes. When the error in self-monitored blood glucose escalated to previously acceptable levels of up to 15%–20%, the risk of missing hypoglycemic readings (blood glucose ≤70 mg/dL [3.88 mmol/L]) climbed to 3.5%–10%. Concurrently, the incidence of hypoglycemia increased to 0.1%–5.5%, representing a substantial risk. ¹³ Glucose variability and the risk of hypoglycemia increased as well. ¹³

Accuracy, availability, and acceptability of BGMS can all impact clinical decisions and outcomes.^1,3,4,10,14 PWD should be provided with BGMS as indicated by their circumstances, preferences, and treatment. ¹⁰ HCPs should be aware of potential differences in the accuracy of BGMS and take steps to ensure PWD, particularly those on insulin therapies, gain benefits from accurate and precise blood glucose measurement. ¹⁰ A key component supporting the empowerment of PWD is comprehensive education about why to monitor their blood glucose as well as how; this knowledge complements consistent and reliable blood glucose testing results in promoting adherence to therapy protocols.^15,16

Strengths and limitations

This study provided an independent evaluation and adhered to applicable international standards. A statistically adequate sample size was used with a diverse patient population. The Cobas C111 reference system is validated, reliable, and highly accurate with metrological traceability. The results of this study pertain to the specific devices, test strip lots, and control solutions used, although all study protocol steps followed DIN EN ISO 15197:2015 requirements. No test results were excluded for protocol violations.

This single-center study took place in a controlled environment and may not generalize to real-world conditions. Assessment of interfering substances or medical conditions relied on patient disclosure and may be incomplete and was not objectively verifiable.

Conclusions

This study revealed that the BGMS Contour Plus Elite fully complies with and exceeds DIN EN ISO 15197:2015 requirements. The device performs at a clinically acceptable level of accuracy with all measurements well within the required range.