Accuracy Discrepancies: MARD Analysis by Sensor Generation
The accuracy of continuous glucose monitoring (CGM) systems is crucial for effective diabetes management. One key metric used to evaluate the accuracy of CGM systems is the Mean Absolute Relative Difference (MARD). MARD analysis by sensor generation can provide valuable insights into the advancements and limitations of CGM technology.
Introduction to MARD
MARD is a statistical measure that calculates the average difference between CGM readings and reference glucose values, expressed as a percentage. A lower MARD value indicates higher accuracy. The FDA considers a MARD of ≤10% to be acceptable for CGM systems [1].
Sensor Generations and MARD
Studies have shown that MARD values vary across different sensor generations. Earlier sensor generations (e.g., first-generation CGM systems) had higher MARD values, typically ranging from 15% to 20% [2]. In contrast, newer sensor generations (e.g., seventh-generation CGM systems) have demonstrated significantly improved accuracy, with MARD values as low as 7.4% to 9.5% [3].
Comparison of MARD Values Across Manufacturers
A comparison of MARD values across different manufacturers reveals varying levels of accuracy. For example:
- Dexcom's G7 CGM system has a reported MARD of 7.4% [4]
- Medtronic's Guardian Connect system has a reported MARD of 8.5% [5]
- Abbott's FreeStyle Libre 2 system has a reported MARD of 9.3% [6]
Pitfalls and Limitations
While MARD analysis provides a useful metric for evaluating CGM accuracy, there are limitations to consider. These include:
- Variability in individual user results
- Sensor calibration and warm-up times
- Interference from external factors (e.g., electromagnetic fields)
Conclusion
In conclusion, MARD analysis by sensor generation highlights the progress made in CGM technology, with newer sensor generations demonstrating improved accuracy. However, it is essential to consider the limitations and pitfalls associated with MARD analysis to ensure accurate interpretation of results.