A Mathematical Formula to Determine the Minimum Continuous Glucose Monitoring Duration to Assess Time-in-ranges:...

Area of Science:

• Endocrinology
• Biostatistics
• Medical Informatics

Background:

• Continuous glucose monitoring (CGM) enables assessment of glycemic control via time in range (TIR).
• Accurate TIR estimation requires a sufficiently long monitoring period, posing challenges for study design.
• A mathematical model exists to calculate minimum trial duration for a desired TIR uncertainty.

Purpose of the Study:

• To evaluate the sensitivity of a TIR duration estimation formula to its input parameters.
• To assess the formula's robustness when parameters are derived from different populations.
• To provide guidance on parameter selection for accurate study duration determination.

Main Methods:

• Utilized two independent datasets to test the formula's predictions.
• Estimated formula parameters from one dataset and applied them to another.
• Investigated the impact of parameter variations on predicted TIR uncertainty.

Main Results:

• The formula demonstrated high robustness, with minor discrepancies (<5%) when using a fixed value for parameter α across populations.
• Parameter p_r requires adjustment based on expected population TIR, as a 20% error resulted in ~10% discrepancy.
• The formula reliably predicts minimum study duration for TIR assessment.

Conclusions:

• The proposed formula is robust to parameter setting variations, making it suitable for determining study duration in diabetes research.
• A fixed value for parameter α is recommended for broader applicability.
• Parameter p_r should be tailored to the specific population's expected TIR for optimal accuracy.