At GeneClinicX we use the Freestyle Libre CGM device manufactured by Abbott Laboratories. This device has been approved for use by the US FDA (Food and Drug Administration), the globally accepted gold standard for medical device certification.
It's common for people to compare their CGM readings against their finger stick to compare accuracy and notice that the values are different. This can be disconcerting, but the key thing to know is that variations in readings will occur and this is expected.
The FreeStyle Libre is factory-calibrated to estimate blood glucose based on interstitial glucose - which is the fluid between skin cells. These numbers can lag behind blood glucose levels by ~15 minutes depending on how quickly things are changing. This can stack up especially when your glucose is actively rising or falling (as opposed to flat and steady), causing the difference between finger prick glucometer and CGM measurements to be more pronounced. For this, you should only compare readings in a fasting state.
The gold standard for accuracy is a blood draw measurement. But both finger sticks and CGMs have error margins to that standard known as MARD, or the mean absolute relative difference.
Finger sticks tend to be in the range of 5-10% MARD, while the Freestyle Libre has a MARD of about 9.2% over 14 days. With this measure, the lower the number, the better the accuracy.
These studies confirm the efficacy of performance and usability of the FreeStyle® Libre™ Flash glucose monitoring system manufactured by Abbott Diabetes Care for interstitial glucose results compared with capillary blood glucose results:
The performance and usability of a factory-calibrated flash glucose monitoring system
Feasibility of factory calibration for subcutaneous glucose sensors in subjects with diabetes