David Mendosa reviewed some reviews on home blood glucose monitor accuracy and reproducibility. He was motivated by an article in Consumer Reports. You’ll want to click through his links for details. The last time I looked into this, I found that a device could receive FDA approval if it could measure accuracy to within 20% of the actual blood sugar value as determined by a laboratory machine. For a blood sugar of 200 mg/dl (11.1 mmol/l), the home device could give you a value anywhere between 160 and 240 mg/dl (8.9 to 13.3 mmol/l). That doesn’t exactly inspire confidence, does it?
I usually check my sugar before I get my blood drawn for hA1C. Usually current glucose is drawn too. My meter is normally within 10-15 or so units, sometimes right on. So I’ve found to be pretty accurate.
I’m usually right at 80 though… I wonder if the meters are less accurate at high glucose levels and more accurate at normal levels.
Ryan, Dr. Bernstein says he has found that to clearly be the case.
sorry but I cannot find a link to the article by David Mendoza
Mea culpa! I forgot to create the Mendosa link when I originally published this post. Here it is: http://www.mendosa.com/blog/?p=1387#more-1387
My home glucose monitor kinda sucks. I took it along when I did the 2-hr diagnostic test for gestational diabetes and while it was a bit off for the fasting numbers (5.8 vs 5.1 as measured by the lab) it was WAY off for the post-bolus numbers (12.9 vs 8.7, 11.7 vs 7.4) Those numbers might not be exactly right (from memory), but the point was the home monitor showed me way up in the diabetic range all over, but the actual lab had me at normal for post-bolus and nearly-normal on the fasting. I spent the Christmas holiday believing I had full-blown gestational diabetes, ate no Christmas treats at all, and lost 5 lbs.
Sarah, I’m glad it turned out better than you had thought! But frustrating nevertheless.
Pierrelegrand, here is the link to the article