A concept for human use of real-time and remote monitoring of diabetic subjects using intermittent scanned continuous glucose measurement.

Autor: Goez-Mora JE; Grupo GITA, Facultad de Minas, Universidad Nacional de Colombia, Carrera 80#65-223, 050001, Medellín, Colombia., Arbeláez-Córdoba N; Grupo GENMOL, Departamento de Fisiología y Bioquímica, Facultad de Medicina, Universidad de Antioquia, 050001, Medellín, Colombia., Balcazar-Morales N; Grupo GENMOL, Departamento de Fisiología y Bioquímica, Facultad de Medicina, Universidad de Antioquia, 050001, Medellín, Colombia., Rivadeneira PS; Grupo GITA, Facultad de Minas, Universidad Nacional de Colombia, Carrera 80#65-223, 050001, Medellín, Colombia. psrivade@unal.edu.co.
Jazyk: angličtina
Zdroj: Biomedical engineering online [Biomed Eng Online] 2024 Feb 28; Vol. 23 (1), pp. 26. Date of Electronic Publication: 2024 Feb 28.
DOI: 10.1186/s12938-024-01217-z
Abstrakt: Background: Flash glucose monitoring systems like the FreeStyle Libre (FSL) sensor have gained popularity for monitoring glucose levels in people with diabetes mellitus. This sensor can be paired with an off-label converted real-time continuous glucose monitor (c-rtCGM) plus an ad hoc computer/smartphone interface for remote real-time monitoring of diabetic subjects, allowing for trend analysis and alarm generation.
Objectives: This work evaluates the accuracy and agreement between the FSL sensor and the developed c-rtCGM system. As real-time monitoring is the main feature, the system's connectivity was assessed at 5-min intervals during the trials.
Methods: One week of glucose data were collected from 16 type 1 diabetic rats using the FSL sensor and the c-rtCGM. Baseline blood samples were taken the first day before inducing type 1 diabetes with streptozotocin. Once confirmed diabetic rats, FSL and c-rtCGM, were implanted, and to improve data matching between the two monitoring devices, the c-rtCGM was calibrated to the FSL glucometer readings. A factorial design 2 × 3^3 and a second-order regression was used to find the base values of the linear model transformation of the raw data obtained from the sensor. Accuracy, agreement, and connectivity were assessed by median absolute relative difference (Median ARD), range averaging times, Parkes consensus error grid analysis (EGA), and Bland-Altman analysis with a non-parametric approach.
Results: Compared to the FSL sensor, the c-rtCGM had an overall Median ARD of 6.58%, with 93.06% of results in zone A when calibration was not carried out. When calibration frequency changed from every 50 h to 1 h, the overall Median ARD improved from 6.68% to 2.41%, respectively. The connectivity evaluation showed that 95% of data was successfully received every 5 min by the computer interface.
Conclusions and Clinical Importance: The results demonstrate the feasibility and reliability of real-time and remote subjects with diabetes monitoring using the developed c-rtCGM system. Performing calibrations relative to the FSL readings increases the accuracy of the data displayed at the interface.
(© 2024. The Author(s).)
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje