The Role of A1c in Light of CGMS
October 24th, 2007 by Kevin
Things are about to change in many ways for both patients and the clinicians who help these people to manage their diabetes. One notion that I’d like to state is that there are two kinds of A1c; the Physician’s A1c and the Patient’s (and in some cases the patient caregiver’s) A1c.
In the present day, the A1c is a routine blood test scheduled to coincide with the quarterly physician visit. The data from this test is sometimes available at the visit but for most people it’s a lab that comes after the visit. Regardless, appointments are not dependent on having a current A1c. Based on the A1c, the physician may increase or reduce meds, introduce insulin or sensitizers, etc… but none of those decisions have much to do with A1c variance between tests unless the results exceed some critical and arbitrary threshold.
For example, there is no hard and fast guideline for what to do when a patient goes from and A1c of 8.1 to an A1c of 7.2 (3 months later). As has been said quite often, blood sugar control is an art form not a science.
Due to several factors, the role of A1c as a somewhat arbitrary generalized retrospective on control is morphing into a very important diagnostic tool that needs to be used by the patient for behavioral feedback. After all, diabetes is all about behavior whether you’re type 1 or type 2. Almost nobody checks their blood sugar around the clock either so even the most frequent blood sugar checkers are only able to see glucose variability (swings) over 50% of the day.
Assumes 4 blood sugar checks per day with each check offering insight into the preceding 30 minutes and the subsequent hour on average. That means each check accounts for 1.5 hours times 4 checks per day and that equates to roughly 6 hours out of every 24. Even if we double it that gives us 12 hours of insight. Hardly enough to adequately control blood sugars using artificial methods like carb estimating, activity guessing and meter based blood sugar level approximations (meters are roughly 10% to 20% of - plus or minus).
Enter: the new continuous glucose monitoring system technologies.
The promise is 24 x 7 x 365 insight into the world of Diabetes Cause and Effect. Talk about an educational tool for helping patients to measure, understand and act! However, when using one of these devices, it is critical that the patient calibrate the continuous data with meter data. So, we now have 24 x 7 insight but with the caveat of the data only being within +/- 20% of the meter and the meter only being within +/- 20% of the 15 minute old actual blood sugar level.
Continuous Monitoring is for real-time trend-based management - not a precision Swiss time piece for keeping the railroad on schedule.
One of the natural tendencies of being on a cgms is to use your meter less over time based on trust regardless of what the indications say about always using a meter test before dosing insulin or taking some other critical action.
So let’s forget the math for a minute and realize that there is huge potential for the patient to end up in the weeds and when they get their quarterly A1c at the physician’s office or lab…drum roll… the patient is surprised and the physician is unphased. After all, the A1c tells no lies, right and it is what it is. There’s a physician-patient moment I don’t need to be in on.
The physician performs his/her art and counsels the patient… but not much to say beyond what the patient already knows. There are obvious exceptions to this but that’s how most encounters go especially with the kind of patient who would be willing to use a cgms.
For the patient however, the A1c is actually the only data point you can trust. The savvy patient will take the average from their cgms and compare to the translated A1c blood sugar average as a quick sanity check. What does it mean when these two numbers disagree?
Herein lies the new role of the A1c in the management of diabetes. Other than 15g of carb to treat a low, the A1c is the only number you can trust if your goal is to stave off long-term complications. It is the ultimate form of calibration for the cgms and this new role dictates several new ways to use the A1c. It also means that the A1c should be performed more frequently and anything more than a +/- 0.1 variance from the True A1c is unacceptable.
With this in mind, the Patient’s A1c becomes the primary guiding light for the patient while the Physician’s A1c is that data point used to manage patients within a practice; two very different goals with two very different sets of requirements and different use cases for utilization of the A1c in the management of diabetes.
Maybe we users of cgms should start referring to the A1c as the Master Calibrator?
Last 5 posts by Kevin
- What's in a name? - March 7th, 2010
- Steve Ponder MD, CDE - Headliner Extraordinaire - January 22nd, 2010
- Fallen Hero Now a Competitor? - January 20th, 2010
- Man Made Barriers to a Man Made Solution - January 16th, 2010
- Healthy Families of South Texas - Launch Day - December 31st, 2009
October 25th, 2007 at 9:53 am
Kevin
Interesting thoughts about the ways in which we, and caregivers, evaluate how we’re doing. There’s really no good answer here, I think.
I don’t even like to think about the +/- difference, mostly because it’s too depressing. In the light of other guidance, I have to assume it’s true. Which I know it isn’t. Oh well. Maybe the next generation devices will get closer to real values. Though that won’t really help us interpret the myriad of values and put them to good long term use.
October 25th, 2007 at 11:05 am
On the accuracy front, there are lots of choices people can make if they only knew which tools are accurate and which are less accurate. For example, our HomeCheck A1c is always within 0.1 and the closest you can get to a True A1c. Likewise, some meters are more accurate than others (ie - Wavesense and Ascenscia). Pump vs. MDI. People who weigh their food on a scale vs. people who never weigh their food or double check the accuracy of packaged foods (especially for breakfast foods when they are insulin resistant and require more insulin/carb.).
My point in this post is that as we move into a world of real-time inputs and real-time decisions, we need to find at least one highly accurate beacon in our personal diabetes care technology toolkit that tethers cgms calibration and behavioral feedback to something akin to True North (vs. false magnetic based indications of the North Pole).
Without this highly accurate anchor point, there is no way to relate cause and effect and major errors will be made much like the arctic explorer who forgets to calibrate his GPS antenna (http://tinyurl.com/yprxrp) before heading off on the expedition.