Work

In Vitro Dynamic Characterization of a Continuous Glucose Monitoring System Using an Independently Measured Glucose Profile: Amplitude Transfer, Rate Transfer, and Consecutive-Day Agreement

Year: 2026
Type: article
Abstract: Continuous glucose monitoring (CGM) sensors are evaluated almost exclusively in clinical studies. We investigated the dynamic response of one factory-calibrated CGM system on a programmable in vitro flow platform, with glucose measured independently by a SUPER GL2 analyzer as comparator. Twelve sensors were studied: four during verification of the measurement procedure (Stage 1) and eight tested on two consecutive days (Stage 2) under a 90 → 270 → 90 mg/dL profile. Amplitude transfer (KA) and apparent rising- and falling-phase slope transfer (Kup, Kdown) were calculated without lag correction. Stage 1 results were exploratory owing to an unexplained CGM displacement of approximately 35 min. The delivered profile differed from the programmed one in every Stage 2 run (upper plateaus 254–268 vs. 270 mg/dL). Mean KA was 0.731 and 0.731, Kup 0.749 and 0.746, and Kdown 0.729 and 0.719 on the two days; between sensors, KA ranged from 0.61–0.84. Near-zero mean paired changes reflected opposite group shifts (KA +0.026 and −0.025). Descriptive ICC(A,1) values of 0.919–0.952 (seven or eight pairs from two shared runs) indicate preserved between-sensor differences rather than established reproducibility. Bias (≈−50 mg/dL) and MARD (31–32%) apply only to these in vitro conditions; the findings do not establish clinical performance or long-term stability.
Source: Sensors
Language: English
Field: Medicine
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7236819614
FWCI:
Cites: 26
Cited by:
Related to: 10