“The Dumbest Idea Ever”

Remember that little discussion about abandoning the established A1c for a new Average Glucose (AG) measurement? Well, guess what? This is pretty much the “dumbest idea ever floated” in diabetes, according to my favorite irreverent industry expert, David Kliff of Diabetic Investor.

Indeed, the ADA and a number of other powerful health organizations — the International Federation of Clinical Chemistry (IFCC), European Association for the Study of Diabetes (EASD), and International Diabetes Federation (IDF) — now have an agreement in place to evaluate the accuracy of the A1c and potentially switch to the new AG units, contingent on results of an international study currently underway. The new standard apparently won’t change the chemistry of the test itself, but only the “reference method,” i.e. how the machines that conduct the tests are calibrated, and how the results are reported.Confused

So why is this idea so dumb? According to Kliff, it’s “purely academic” and will have no value for patients in the real world, other than to confuse the heck out of them.

The notion sounds good: average blood glucose results should be reported in the same unit measurements patients use for self-monitoring, so that everybody is “using the same language to communicate glucose goals.”

BUT, Kliff points out, the A1c as it stands is the ONLY number most patients understand. He cites the great masses of non-insulin-dependent diabetics who don’t use their home glucose monitors at all, because they don’t know what the results really mean or what they should do with those numbers.

With the A1c, on the other hand, it’s easy to understand that any number under 7 is good and any number above 7 requires action. So how in the heck will patient education be improved by replacing an easy-to-grasp test result with a new number that most patients don’t understand?

Hmm, you’ll note that the current AG study is underwritten by the likes of Abbott Diabetes Care, Bayer HealthCare, GlaxoSmithKline, LifeScan, Inc, Medtronic MiniMed, and Merck & Co. — so I’m guessing these companies must have something to gain here.

And that’s not all. According to Kliff, the great race to introduce non-invasive glucose monitoring provides even more evidence of “just how far out of touch people from academia can be with reality.”

“For years, Diabetic Investor has been saying the so-called pain factor is vastly over-rated when it comes to why patients don’t test regularly. The simple fact is that the majority of patients don’t understand what the results mean and there see no action step based on the results. Why would anyone do something that does not add value to their lives or improve their outcomes?”

Touché. But I’m not sure I agree with you on the pain factor there, David. Admittedly, I’m a Type 1 currently bleeding at least a dozen times a day to get a number that I do understand (the result, if not always the cause) and can act on. I sure do agree, however, that the current A1c is generally less confusing than daily readings. So if it ain’t broke, don’t fix it. Correct?


22 Responses

  1. Kendra
    Kendra July 17, 2007 at 7:12 am | | Reply

    Hmm..interesting indeed, but I’d file this one in the “applies more to uninformed Type IIs” (which, unfortunately, are more common than doctors would like to admit…I don’t think Type Is have luxury of not getting what the meter is saying.) All the folks w/ diabetes that I know understand the difference between a 200 and 20, and know how to act accordingly…so I’m always a little bit shocked that the majority of diabetics haven’t been taught that simple concept.

    I think the suggestion here is that for most people with diabetes, education is more important than fancy devices or fine-tuning the A1C measurement – and in principle I agree. However, I wouldn’t sniff at fine-tuning the A1C measurement, or finding a better way to gauge my overall control. Knowledge isn’t confusing, it’s POWER. As a diabetic who tests about 10 times a day and knows the subtle difference between a 75 and an 85, I would appreciate all of the academia :)

  2. riva
    riva July 17, 2007 at 7:29 am | | Reply

    Changing formulas aside, I just came back from the Children with Diabetes conference in Orlando and discovered some interesting things at the included health fair expo. One is Pelikan. They are testing a new lancing device powered by a battery which gives a less painful prick. It has 30 adjustable depths, but it’s uniqueness is a hidden lancet comes out and first touches up against your finger to assess how tough the skin is and then softly enters. They have conducted a trial in Australia and are now conducting a 90 day trial here. If you contact Pelikan you may be able to join the trial. So far, I like it, not to mention it’s cool gold color, but it’s too big to carry around. I use it at home.

    Then I discovered something confounding and very irritating. Since most of the meter companies were there and giving out meters I picked up two from One Touch (One Touch Ultra 2 and the new mini) as I already use their One Touch meter, now quite outdated, and I got a Freestyle since it uses the smallest drop of blood. I had the nurse there test on my foremarm, which I’d never done before. I hated it. First, they always tell you it’s not as accurate as testing on your finger, so why bother, and second, the rubbing she had to do on my arm to get the blood to the surface and then the pressing after she lanced to get the blood, made me feel like I was going to faint. Obviously this is not for me.

    But here’s the kicker. For the next three days I tested on all machines because I found disparity amongst all of them. The Freestyle put me 30 points above the One Touches. That’s the difference between 65 which I would treat and 95, where I’d be patting myself on the back. Or, 100 where I’d be happy or 130 where I’d be wondering what I did wrong. Ugh! Even between the three One Touch meters the spread went from 102 to 114.

    So A1c aside, how do we know whether the numbers we’re getting home testing are accurate?

  3. Anne
    Anne July 17, 2007 at 8:17 am | | Reply

    The AG vs A1c study seems to be a waste of money. This money could be better spent developing a NON-INVASIVE blood glucose meter. I for one would be thrilled to stop jabbing myself and inserting needles into my ever-scarring skin.

    Sometimes when I go into conspiracy theory mode I figure that a non-invasive meter hasn’t come along since companies haven’t figured a way to include expensive disposable parts. But then I try to be more rational and realize this has never been a problem before so the problem really must be technical.


  4. Dave
    Dave July 17, 2007 at 8:26 am | | Reply

    When I was first diagnosed with Type 1 I was in shock. It took about a solid year for me to get everything under great control and for me to understand all the different numbers and what they meant. Now I am a pro. I have to admit the BG testing isn’t enjoyable but the pain that everyone talks about is seriously overrated IMO. I do look forward to some day not having to prick my finger but not because of the pain but rather to get more results so I can do an even better job of managing my BG.

  5. Khurt Williams
    Khurt Williams July 17, 2007 at 9:34 am | | Reply

    I agree with your statement – “applies more to uninformed Type IIs”.

    This has been my experience. My cousin (Type II) keeps insisting that a BG over 200 is acceptable because it’s a lot lower than the 400 she had when diagnosed.

    Of course she is still 100 pounds overweight and her eating habits have not changed.

    The meter reading disparity thing drives me nuts. As an engineer I can’t fathom how there is not standard reference point or required accuracy rate for these devices.

  6. Rob
    Rob July 17, 2007 at 9:38 am | | Reply

    I just had an endo visit last week & discussed my results. I test ~15x per day or use a CGM. The message I got from my endo is that for anyone who tests as often post-pradially as they do pre-prandially (before/after meals), then mean monitor BG is probably statistically more reliable than A1c, regardless of how it’s calculated. According to the lab (at a major research university), A1c is considered to have a margin of error of ~10%, which I think makes it about the same as fingerstick margin of error. A1c is a superior margin only for people who do not test (or record) after meals, since this means that they are under-representing the periods of the day when they are most likely to be hyperglycemic.

    Because fingerstick tests are perfrmed more often than A1c tests, the variation in margin of error will tend to wash out, provided that the patient is testing an equal frequency both pre and post prandial.

    The other tidbit was that at least at this university was that they were moving away from whole blood A1c tests and towards exclusive use of the Rapid test. They had found less variation with the Rapid tests than they had with the tests from blood draws. The rapid tests tend to yield lower numbers but are more consistent – a patient who has a change in A1c from 7.5 to 7 on the Rapid test is more likely to have seen an improvement in control than a patient who sees the same variation with the standard tests given that this change would be well within the margin of error.

    Just thought I’d pass along.

  7. Rob
    Rob July 17, 2007 at 10:45 am | | Reply

    The AG vs A1c study seems to be a waste of money. This money could be better spent developing a NON-INVASIVE blood glucose meter.

    I think anytime you talk about non-invasive testing, you’re talking about testing something other than blood levels – skin cells, interstitial fluid etc. I think there are some real inherent accuracy challenges to this – you can get a pretty good sense of it from using a CGM for a while. It’s just not the same thing.

    Also, I tested the Glucowatch a few years back – it was so not-ready-for-primetime back then so I’m not surprised they aren’t getting much closer.

  8. Chloe B.
    Chloe B. July 17, 2007 at 12:37 pm | | Reply

    All I know is that even with the A1c, I get frustrated!. It says my a1c is 6. GREAT!! But my meter average is leaning towards an equivalent A1c of 8.5 or over. Who do I trust? Is it because I’m slightly anemic or my meter is that unreliable.That’s well over a 20% discrepency and that bothers me. My endo adjusts the insulin according to the meter, and at the same time praises the A1c that doesn’t match it.

  9. mollyjade
    mollyjade July 17, 2007 at 12:40 pm | | Reply

    I’ve never understood why they can’t give you both numbers at once. “Your A1c is 7% which is equivilent to a meter reading of 170.” Patients who understand the A1c scale can latch onto that number and patients who are more familiar with meter readings can pay attention to that.

  10. Craig Williams
    Craig Williams July 17, 2007 at 1:17 pm | | Reply

    Kind of off topic, just a response to the previous comment. Converting between A1c and mmol/L is simple. It’s either multiply or divide by 18.
    A1c*18 = mmol/L
    7*18 = 126 mmol/L
    170/18 = 9.4 A1c

  11. mollyjade
    mollyjade July 17, 2007 at 1:35 pm | | Reply

    Craig Williams, were you referring to me? I was using this table http://www.dlife.com/dLife/do/ShowContent/blood_sugar_management/testing/a1c_conversion.html

    A1c and mmol/l are a bit different. The ranges are similar which makes it confusing. Giving readings in mg/dl is a US thing.

  12. Hannah
    Hannah July 17, 2007 at 1:47 pm | | Reply

    I agree with you Amy, that the “pain factor” can actually be a pain. No matter how long you’ve been a human pincushion, it’s not exactly something to strive for. Plus, the noninvasive glucose reading is one of those “holy grail” situations if you get it right.

    I think a non-invasive device is really more of a matter of convenience. No trail of test strips around the house, no wasting $2 every time your reading is inaccurate, 3 less objects to pack in your purse every day (no lancing device, no lancets, no strips), and best of all, if they can get a non-invasive continuous monitor, NO SCAR TISSUE from having to insert sensors under your skin.

    Non-invasive sounds like the way to go to me!

    However, that A1c should stay right where it is. We’ve got enough confusing crap to deal with.

  13. Rob
    Rob July 17, 2007 at 1:49 pm | | Reply

    All I know is that even with the A1c, I get frustrated!. It says my a1c is 6. GREAT!! But my meter average is leaning towards an equivalent A1c of 8.5 or over.

    I think that means that you are probably experiencing a lot of overnight lows/post-prandial lows. You would get that result if you clustered your fingersticks at times of day when your blood glucose values tend to be elevated.

  14. Rob
    Rob July 17, 2007 at 1:52 pm | | Reply

    and best of all, if they can get a non-invasive continuous monitor, NO SCAR TISSUE from having to insert sensors under your skin.

    The Glucowatch, for one, left horrendous burn marks that lasted (for me) a couple of weeks. I’m sure there have been some improvements in technology, but there are probably still some tradeoffs involved.

  15. Craig Williams
    Craig Williams July 17, 2007 at 2:01 pm | | Reply

    Yes, mollyjade.
    The labels in my post were wrong, should have been mg/dl.
    And now I’m confused that the dLife table doesn’t support the rule of 18 which I thought was true… I’ll have to research that again sometime.

  16. Allison
    Allison July 17, 2007 at 2:32 pm | | Reply

    The problem with the whole BG average/A1C average comparison is the fact that the A1C accounts for all the blood sugars your meter didn’t catch. That’s the whole reason the CGM is so good. If we knew what our blood sugars were 24/7, I’m sure our meter average would be much different. Meter discrepancies aside, you would never be able to get a 24/7/365 day average because no one tests 24/7/365, so obviously the numbers are not going to be the same. It’s impossible!

  17. Richard
    Richard July 17, 2007 at 2:37 pm | | Reply

    I think it is an attempt to “dumb down” yet another thing in our society.

    Folks need to understand that HbA1c and Average Blood Glucose are not (not) the same thing and stop trying to make one into the other in order to “make them easier to understand.”

    This is nonsense, they will never be the same.

  18. olive
    olive July 17, 2007 at 2:56 pm | | Reply

    The dLife table is giving an mg/dl approximation of the A1c percentage. A 7% A1c is approximately equivalent to an average BG of 170 mg/dl.

    The rule of 18 converts the mg/dl measurement that we use in the US to the mmol measurement of BG that is used in many other countries. Using the rule of 18, a 7% A1c is approximately equivalent to an average BG of 9.4 mmol, (170 mg/dl.) The table is also rounding to the nearest .5 mmol.

  19. Rob
    Rob July 17, 2007 at 3:59 pm | | Reply

    If you test a lot (~15x) and test at wildly varying intervals, you’ll end up with a pretty close correlation between monitor average & a1c theoretically. That’s been the case for me at least. I think Standard Deviation equations are useful for this as well, in terms of telling you how reliable the mean bg levels are (the monitor levels that is).

    If you have a high standard deviation, then it probably means that your A1c is going to be more of an accurate reflection than your monitor averages. If you have a low standard deviation, then it’s a wash (or possibly the monitor avg is more accurate, given margin of error on A1c).

  20. Angela Biggs
    Angela Biggs July 17, 2007 at 5:08 pm | | Reply

    Personally, I’ve always found the A1C a bit opaque. “Under 7 is good, over 7 is bad”: well, what does that really mean? It leaves an awful lot of wiggle room! I tend to go online and use a web-based calculator to convert from A1C to mg/dL because I feel it gives me a better grasp of the situation – after all, my CGM gives me my readings in mg/dL all day, every day. An A1C taken once every three months is just a random number floating in space as far as I’m concerned; however, it doesn’t take much effort to convert that floating number, so I’m fine with the way things are.

    So it’s true that Type II diabetics really do tend to be so under-educated? We have a local paper with an “Ask the Doctor” section, and last week the doctor told a Type II diabetic that a blood sugar over 300 was completely OK, as long as it didn’t last for more than a few days…. SCARY!

  21. Vicki
    Vicki July 17, 2007 at 8:20 pm | | Reply

    So, Angela…did you write to the paper and insert, in your most diplomatic way, comment about 300 NOT being acceptable at all?? Your chance to educate a few people – including that doc.

  22. Kendra
    Kendra July 18, 2007 at 10:31 am | | Reply

    Angela, I’d say that was pretty poor medical advice from that doc. Yeah, if I catch a 300 on my meter say…1hr post prandial, I don’t freak out. I bolus and get it down and don’t spend anymore time worrying about it unless it becomes a pattern.

    Telling someone that it’s okay to run at 300 all..day…long…or maybe for two days…shoooo.

Leave a Reply