Discrepancies between current displayed and auto‐logged glucose values in FreeStyle Libre 3: Implications for clinical interpretation
Primary result · Difference between CUR and AL valuesResult — go to the resultMean difference −1.2 mg/dLAuthors' conclusion
Although differences between AL and CUR were generally small, they were systematic and more pronounced in critical contexts like hypoglycaemia and rapid glucose change. Recognising these patterns may improve CGM data interpretation and alignment between user actions, provider decisions, and automated systems. These differences can reclassify readings across clinical thresholds, affecting patient-clinician alignment despite small average biases.
Study context
- Population
- Study participants wearing FreeStyle Libre 3 continuous glucose monitoring (CGM) over 15 days, with three in-clinic sessions with glycaemic excursions
- Participants
- 24 participants
- Comparison
- Current displayed glucose values (CUR) vs auto-logged glucose values (AL) in FreeStyle Libre 3
Methodological detailspaired comparison, 15-day study, in-clinic sessions with glycaemic excursions
- Paired comparison
- 15-day study
- In-clinic sessions with glycaemic excursions
Primary result
Difference between CUR and AL values
Mean difference
−1.2 mg/dL
As reported in the abstract
- Discrepancies exceeding ±10 mg/dL
- about 10% of cases
Further findings
MARD
Mean absolute relative difference (MARD)
- MARD comparable between AL and CUR, with greater deviation in hypoglycaemia
Display errors (error message instead of CUR value)
- Display errors
- 3.9%
- Mean AL difference at display errors (higher glucose levels)
- +91.9 mg/dL
- Mean absolute rate of change (RoC) difference at display errors (rapid fluctuations)
- +1.52 mg/dL/min
- P for both associations
- P < 0.001
- Linear mixed-effects models
Original publication
Discrepancies between current displayed and auto‐logged glucose values in FreeStyle Libre 3: Implications for clinical interpretation. Diabetes Obes Metab. Published 18 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →