It appears that you are currently using Ad Blocking software. What are the consequences? Click here to learn more.
diabetes supplement

Diabetes Update: FDA Diabetes

Number 39; June 19, 2002

By David Mendosa

This newsletter keeps you up-to-date with new articles, columns, and Web pages that I have written. I list and link most of these on my Diabetes Directory at

From time to time Diabetes Update may also include links to other Web pages of special interest.

My most recent contribution is:

on June 19, 2002


    on June 17, 2002
  • Meter Accuracy
    As soon as I can I will be updating my Web page on Blood Glucose Meters (Part 14 of On-line Diabetes Resources) with new information about meter accuracy developed from my interviews at the FDA for my ADA column referred to above. I was particularly interested in what standards the FDA set for its clearance of new blood glucose meters, in part because of a nasty argument I was involved in just last month (he was the nasty one, not me).

    He started it by writing on a newsgroup that "all meters sold in the U. S. must meet performance standards as articulated by the FDA. That standard, in general, states that the reading produced by any meter should be within 20% of a reading from that blood sample in a doctor's laboratory..." After some back and forth I told him that "there is no place [on the FDA site] that I find that even talks about '20%.' Can you at least point me to that?"

    He refused, writing "I can't do your research for you. Get in and dig it out. I'm in the business of applying computer and information science techniques for pay on the internet, so it would not be fair to my clients for me to do your research for free."

    I replied that "I do my own research. I just followed up a lead from you where you said you got some information. Obviously, you can't substantiate your statement."

    Finally, he wrote, "Want to bet $5,000 that I can't show you the substantiation? Obviously, you haven't the courage of your convictions."

    I didn't bet with him then, and it's too late now that I have done the research to prove him wrong. It's too late because he had a heart attack and died a couple of days after he challenged me to bet him $5,000.

    As my ADA column reviewing the new FDA Diabetes Web site says, the FDA doesn't have any standards that the agency applies when approving new blood glucose meters. Nothing could have surprised me more, but the statement comes from three FDA officials including the director of the FDA's Division of Clinical Laboratory Devices, Steve Gutman, M.D.

    Steve emailed me with much more information than I could use for my column. But it is important, so I will share it here.

    "Premarket review of glucose meters has traditionally not been standards based," Steve wrote me. "That's because there is not a well-defined single reference to use for establishing performance and because the relative degrees of performance required may vary with the setting."

    "The DCCT [Diabetes Control and Complications Trial] was performed in a phased in manner beginning using old generation devices and completing using with newer generation devices and so it is generally our view that this standard of performance total errors in the neighborhood of 20% or less are acceptable for current practice and are being met by virtually all meters. Unfortunately, meters are subject to both user error and random pre, analytical, or post analytical errors which can confound results.

    "This informal performance target, however, has not been codified anywhere; however the soon to be adapted ISO [International Organization for Standardization] standard on glucose meters does require a 20% maximum system error and also requires that data performance at 5, 10, and 15% of the reference method be included in labeling.

    "We did do a retrospective evaluation of our submissions and found that virtually all in the 510(k)s have error quotients in the mid to high teens. This does not meet the ADA [American Diabetes Association] standards, although these standards were never rightly defined and did not refer to a particular reference point to be used. It is probably reasonably easy for a meter with systemic error to correct for bias and probably more challenging to correct for imprecision, but I know of no formulas or consensus to define error components this finely.

    "Among the standards that do exist, many offer dual criteria—one for values above 100 mg/dl (usually max performance of 15 or 20% error) and one for values below 100 mg/dl (usually max performance of 15 to 20 mg/dl rather than % error.

    "While we are not thrilled with the limits likely to appear in the international ISO standard, we do plan to recognize this standard and do believe that the current products probably represent a reasonable balance between performance and cost.

    "It is my understanding that part of the variability that plagues monitoring of this disease is biological and not analytical and that CDC [Centers for Disease Control and Prevention] does plan to initiate studies to try to sort this out."

    That was a lot to digest, but then I followed up by asking Steve these questions:

    "What can you tell me (and put on the new FDA Diabetes Web site?) about the performance of the meters that are on the market in the U.S. now?

    "You write that 'virtually all meters' are accurate within 20%. Which ones aren't?

    "And can you say which meter or meters is the most accurate? Dr. Richard K. Bernstein insists that all his patients use the Glucometer Elite because it is the most accurate. Is it?

    "And how do the newer alternative site meters compare, aside from the lag time issue? Are the MediSense Sof-Tact (which I use), the TheraSense FreeStyle, and the LifeScan Ultra any more or less accurate than meters of any earlier generation?"

    He started his reply by referring me to two abstracts from 1998 and 1999 that looked at meter testing and accuracy. I found one article, "Newer Portable Glucose Meters—Analytical Improvement Compared with Previous Generation Devices?" online at and the other, "Clinical and Statistical Evaluation of Self-Monitoring Blood Glucose Meters" at

    Unfortunately, the meter market has changed so much that only one of these meters is still on the market (LifeScan's One Touch Basic). But, as Steve says, "the bottom line message [of the 1999 article] is that over time the newer generation of meters has been improving."

    The 1998 article "is a bit complex," Steve writes, "but promotes the idea that the percentage agreement may be less important than how well a meter works by error grid analysis, since this may be more clinically relevant. At FDA for new technologies we like to consider both types of statistical analysis and that has been true in reviews of both MiniMed and GlucoWatch.

    "FDA does not do head to head comparisons between various glucose meters and so we cannot comment on the relative accuracy or repeatability of these. The best advice is to evaluate information in the literature and to make sure that for any device being used some form of quality control (either commercial quality control solution checks or regular checks against conventional laboratory tests) is being performed.

    "I do know that different clinicians often have their favorite devices or devices they particularly dislike but to my knowledge these reflect the same user preferences represented by purchase of any important device, for example a new car, and that there is no clear cut scientific evidence to favor one particular meter over others.

    "There was a piece in Consumer Reports several years ago [actually October 2001] that rated meters and found some differences, but it may already be outdated since most companies are constantly trying to improve their products.

    "FDA review of the alternative site devices has suggested so far that they are analytically quite good producing equivalent results on smaller samples. The problem that must be accounted for is that the alternative site may exhibit physiological and/or biological differences from the same measurement obtained from the finger and the finger values are better calibrated simply because they were part of the DCCT studies.

    "I believe that all cleared alternative site devices have labeling warning users about the limits of the use of these sites. At the recent panel meeting on this topic the benefits of having these available for users clearly seemed to outweigh the dangers of both lags in glucose values and possible truncation in upward and downward trends."

Book Reviews:

    on June 12, 2002
    A Child in Your Care has Diabetes
  • A Child in Your Care has Diabetes: A Collection of Information
    Elisa Hendel's daughter Robyn was six when she was diagnosed with type 1 diabetes. When Elisa recovered from the shock of that diagnosis, she diligently collected a huge amount of information on diabetes. Each year she simplified and organized it, distributing it to everyone who is involved in her daughter's life. Now she shares the wealth of her experience with the rest of the world.

    What she has created is a concise, easy to follow manual for parents, teaches, caretakers, friends, baby sitters and others who are responsible for a child with diabetes. This small booklet of about 55 pages is deceptive. It packs more of value that many books ten times its length. It is not a medical manual or a treatment plan.

    While the material in this booklet is basic, there is information here that was new to me. I was previously unaware of the importance of section 504 for people with diabetes in schools. The booklet provides detailed guidance about the "504 Accommodations". Section 504 comes from the Rehabilitation Act of 1973. It provides that "No otherwise qualified individual with a disability in the United States...shall, solely by reason of his or her disability, be excluded from the participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance or under any program or activity conducted by any Executive agency or by the United States Postal Service."

    Elisa just published the fruits of her research and experience through Hen House Press Inc., 12 Route 17 North, Suite 212, Paramus, New Jersey 07642; phone (201) 291-9199; fax (201) 291-9177. ISBN 0-9718612-0-X. The price is $21.95. It is available through at


  • Diabetes Forecast - Live!
    You may submit your questions now for the next edition of Diabetes Forecast - Live! "Healthy Fast Foods" on Tuesday, June 25. Diabetes Forecast - Live! is a monthly webcast brought to you by the American Diabetes Association and the HealthTalk Diabetes Education Network. Hear experts talk about what's important in a diabetes-friendly meal plan and get quick recipes for breakfast, lunch and dinner. Get your food questions answered! For more information, please visit .


I send out Diabetes Update e-mail in HTML format, which all Web browsers and most modern e-mail programs can display. HTML has live links to all the sites named in the text so that with a simple click of a mouse you can connect to the site you have just been reading about.

This newsletter is free and will never include advertising. Nor will I ever sell, rent, or trade your e-mail address to anyone without your permission.


I send out Diabetes Update about once every two weeks. Previous issues are online:

This is a one-way "broadcast" mailing list that is not set up to accept replies. If you have any questions or wish to unsubscribe, simply write me at [email protected]. If your friends want to receive Diabetes Update, all they have to do is write me here.

Thanks for joining!

Never Miss An Update

Subscribe to my free newsletter “Diabetes Update”

I send out my newsletter on first of every month. It covers new articles and columns that I have written and important developments in diabetes generally that you may have missed.

Your Email Address

Most Popular Articles and Blog Posts

Advice For Newbies Diagnosis of Diabetes
Incorrect Terms Glycemic Values of Common American Foods
Glycemic Index The Normal A1C Level
Glycemic Values Controlling the Dawn Phenomenon
The Biggest Diabetes Scams The Food Insulin Index Trumps Carb Counting
David’s Guide to Getting our A1C Under 6.0 Chia Seeds
What Really Satisfies Snake Oil Supplements

diabetes supplement
Never Miss An Update!