Interesting situation. I had two A1c’s taken within days of each other. One at Joslin Diabetes Center with a reading of 5.8. I had some lab work done and the physician added my A1c and it came back from the lab as 6.1. So, which one is correct and should I have another?
No point in having another right now. Those readings are probably within the margin of error. I have had my AIC’s tested twice around the same time, and also had slight different results.
They are both correct. The variation is too small, and it is within range of margin of error. Our blood sugars vary, and change within our bodies… no two labs will show exactly the same result, every single time.
If you’re an accountant, neither one is correct.
If you’re a scientist, they are both within the margin of error.
Labs in the US, for A1C’s, can differ by 0.5 quite easily from lab to lab, or even from day to day at the same lab. The DCCT attempted to standardize A1C measurements but the equivalences are not perfect.
The two labs in my area, Labcorp and Quest, have reference ranges differeing by about 0.7.
Would you behave differently to 5.8 than to 6.1?
It’s a little bizarre how much people read into the digit after the decimal point given the lab uncertainties.
Despite what you might be led to beleive, the HbA1c is actually an imprecise measure. The most alarming thing about the HbA1c has been the lack of a reference standard. This means that if you have an HbA1c from different labs, you will get different answers. Generally, if you stuck with a single lab, you would find the HbA1c tolerable. The National Glycohemoglobin Standardization Program sought to fix that. However, the margins of error are still surprisingly large. In order to be certified by the NGSP you need to meet the following criteria:
- Manufacturer and Level II Laboratory Certification: The 95% CI of the differences between methods (test method and SRL method) must fall within ±0.75% HbA1c.
- Level I Laboratory Certification: The 95% CI of the differences between methods (test method and SRL method) must fall within ±0.70% HbA1c.
- Certified Level I Laboratories will continue to be monitored quarterly by a 10-sample comparison. The 40 sample range for both certification and monitoring will change from 4-12% to 4-10% HbA1c.
That is right, you only have to get the HbA1c correct to ± 0.7% for a Level 1 lab.
They have a schedule for meeting those goals, and in 2010, the goal is to have the surveyed labs meet a goal of ± 8% in 2010. This means that a measured value of 5.8% would be measured by other labs between 5.33 and 6.26% and still the lab results would be considered “fine.”
I don’t know about yous guys, but that is not fine by me that is cr*p.
And don’t bother getting another, particularly from another lab, this is about the best you will get from separate labs.
Just my thought for the day.
I appreciate all the input on the A1C issue. I grew up Catholic and always looking for the perfect numbers (haha). I find this site to be one of the most helpful in discussing these issues so thank you all for your thoughts and data, it really helps. 
Your results look great. You might also consider that your A1c itself can vary as it is an average, plus a margin of error. Science is not perfect and we shouldn’t expect it or demand it to be. Sounds like you are doing well and should continue in the same way.