Testing quality of a self-monitoring blood glucose sensor with an auto-coding mechanism when used by patients versus technicians

Abstract

Background: In response to the problem of erroneous readings due to miscoding when performing self-monitoring blood glucose (SMBG), this study introduces a user-friendly SMBG biosensor with an innovative auto-coding module on the meter and strip. Actual users characterized the performance of the SMBG systems.

Methods: A total of 105 patients were incorporated in the study and Clarke error grid analysis (EGA) was administered to evaluate the clinical accuracy of the results obtained by the patients versus the technicians. All patients used the questionnaires to comment on the use of the auto-coding sensor.

Results: In the imprecision test, the total CV of the 5 BG levels was 2.1%. In the EGA plot, the results of the auto-coding sensor were 96.2%, both lots A and B, in zone A for the patients and 99.0% and 97.1% for the technician. The paired t-test demonstrated no statistically significant difference between the patient and technician measurements. Regression analysis also demonstrated that the measurements taken by the patients agreed with those obtained using the laboratory method.

Conclusions: The patients achieved satisfactory performance using the auto-coding SMBG sensor and derived similar results with both laboratory reference and operation by a technician.

Keywords: Auto-coding; Miscoding; Patient-derived; Self-monitoring of blood glucose (SMBG); User error.