Outstanding read, Chris! Confirms my anecdotal experience using the Abbott FreeStyle Lite for a long time (biased slightly low but excellent precision) and the LifeScan Verio recently (all over the place!). With the Verio, I’ve been taking 2–4 fingersticks and averaging them together for CGM calibration.

It’s also worth noting how terrible of a metric A1c is. See slides 9–12 here for a few thoughts on this: https://closeconcerns.box.com/s/jtalv5znq44vb3s9g2pk.

If you go to the 2008 estimated A1c study, you’ll notice two things:

• The huge dispersion of A1c values around a given mean glucose, and the huge dispersion of means around a given A1c value. Your 7% and my 7% are different because we may glycate hemoglobin differently. Irl Hirsch has talked a lot about this, particularly in people with anemia. T1D Exchange is also doing a study to look at racial differences in CGM.
• That 2008 study used super old CGM! Would love to see it repeated with more accurate CGM.

It’s a good reminder that “truth” in diabetes comes in shades, and patients are blamed way too often when it’s the tools that make things challenging (infusion sets are another example of that). whew.