Why You Should Worry About the Accuracy of Your Glucose Test Strips

True or false:  The FDA guarantees the safety and accuracy of blood glucose test strips currently on the market.

Answer: False. If I may quote from the website of StripSafely, a patient-run effort to ensure test strip quality, “The FDA reviews test strips presented by manufacturers before the strips can be marketed but not after. Strips can and do vary after they are marketed. There is no post market review.”

Are you surprised? I certainly was. That’s why l attended last week’s summit led by David Klonoff of the Diabetes Technology Society in Bethesda, Maryland about a proposed post-market surveillance program for diabetes test strips. Klonoff is a clinical professor in the division of endocrinology and metabolism at UCSF; he founded the Diabetes Technology Society , which is a non-profit, non-governmental organization  “committed to promoting development and use of technology in the fight against diabetes” in 2001.

The meeting was a follow-up to a previous DTS gathering in May 2013 that focused on the overall issue of test strip quality. The upshot of that first meeting? There are quality issues. According to The Diabetes Care Project’s summary of two recent studies on the issue:

1. The first study found that “of the 27 SMBG [self-monitoring of blood glucose] systems available in the United States, more than half of them (16) do not meet ISO standards, the common gold standard for meter accuracy.”
2. The second found that “only three of seven SMBG systems sold in the United States consistently met ISO accuracy standards, and since this study was conducted, even tighter ISO standards have been adopted.”

On the surface, this already sounds shocking – but let’s delve into what it actually means. First, ISO is the acronym for the International Standardization Organization, which sets the accuracy standards that the United States’ Food and Drug Administration (FDA) uses for blood glucose test strips. (The FDA is the agency responsible for regulating blood glucose monitors.)  ISO standards used for both of the studies mentioned above required that 95 percent of a blood glucose meter’s readings be within plus-or-minus twenty percent of the actual blood glucose value if the blood glucose in question were 75 mg/dl or higher. (Readings below 75 mg/dl must be within 15 percent of actual value.)

Putting this into practice, this means that if your blood glucose was actually 100 mg/dl, a meter would be considered “accurate” if 95 percent of its readings fell between 80 mg/dl (100 minus 20 percent) or 120 mg/dl (100 plus 20 percent) – resulting in an accepted range of 40 mg/dl. A little surprising, but clinically not too a big deal, since you’re unlikely to take insulin for either.

But as is always the case, the higher your initial number is, the greater the accepted margin of error becomes – and it’s not hard to imagine some potentially serious consequences. Say your blood glucose were actually 400 mg/dl. That would mean your meter would be considered accurate if it read anywhere from 320 mg/dl to 480 mg/dl – an accepted range of 160 mg/dl.

This time the clinical consequences could be severe. To play this out, let’s say your actual blood glucose is 400 mg/dl and you’re aiming for a target blood glucose of 100 mg/dl with a 1-to-50 correction factor (i.e. one unit of insulin for every 50 mg/dl you want to drop). If your meter says 480 mg/dl, you will need about 7.5 units to drop to 100 mg/dl (assuming that diabetes acts like a math problem, which it never really does – making this potentially even worse). But since your blood sugar was really 400 mg/dl, you should have taken 6 units. The difference? 1.5 units of insulin. If you really do have a 1:50 correction factor, this means that you’re going to overshoot your target of 100 mg/dl by 75 mg/dl. That means you could end up with a blood glucose of 25 mg/dl. And that, at least in my opinion, is kind of a big deal.

So it is upsetting, to put it mildly, to hear that so many of the systems that have been cleared for sale in the US – with a concentration in low-cost meters — are not meeting even these relatively lax ISO standards. (As a side note, as of April 2013 there are newer, tighter ISO accuracy standards for blood glucose meters, but those are not yet in effect in the US.  The plus-or-minus 20 percent standard is what was used for both of the accuracy studies mentioned above, and is also the standard that was used to clear all of the glucose monitors currently on the market.)

Getting back to the pop quiz question above, there is currently no system in place in the United States to ensure that test strips and meters continue to achieve this level of accuracy once the FDA has cleared the meter for market. (What’s more, the FDA’s initial approval is based on data supplied by the manufacturers themselves; not from random meters bought off store shelves.) This means that if a meter got cleared for sale ten years ago, the FDA has likely never evaluated the meter again. Suppliers change; manufacturing practices change; production line managers change; production sites change. All of these factors can affect quality and accuracy (and, if those studies are to be believed, definitely are affecting quality and accuracy) — and yet there is no requirement or system in the United States for any sort of post-market quality assurance.

And that brings us to last week’s meeting. Klonoff and DTS had assembled a conference room full of players – representatives from test strip manufacturers, endocrinologists, government officials, patient advocates, key players in ADA, JDRF and other important diabetes-related advocacy organizations, even several people from CMS (the people who decide Medicare’s coverage policies and thus heavily influence the private insurance market – more on them later) – to give him feedback on his proposed mandatory system for post-market surveillance. (Again, Klonoff and DTS are not directly affiliated with the government or the FDA; this is a volunteer project for them.)

Here’s the official summary of DTS’s proposed post-market surveillance system:

“During the meeting, DTS presented a plan to develop a mandatory post-market surveillance program that could provide independent assessment of FDA-cleared SMBG systems to ensure they maintain accuracy standards. The program would be led by a diverse steering committee and implemented in several labs across the U.S. and outside of the U.S. This program could deliver performance verification information to the FDA and generate information to assist patients, healthcare providers and payers with product selections. Next steps for the program include:

• Forming a steering committee of regulators, clinicians, payers, advocates, industry and patients to develop program protocols
• Consulting with the FDA to ensure regulatory compliance and validity of the program
• Seeking funding for the program
• Collaborating with the entire diabetes community to achieve the mutual goal of patient safety.”

Now if you’re like me – that is, if the idea of a mandatory post-market surveillance program for devices whose inaccuracies could have life-threatening consequences seems like a no-brainer – you may be very surprised that nothing like this currently exists, let alone that the proposed program is being created by non-government-affiliated volunteers.

Even odder, the idea of post-market standards is hardly a new idea – as other speakers pointed out, they already exist for a wide variety of tests and products.  Cholesterol tests, for example, used to be wildly erratic and untrustworthy before standardization. So were HbA1c results. As this meeting made clear, the absence of post-market safety standards for test strips stands out as a frightening anomaly.

As the press release makes clear, Klonoff’s/DTS’s would-be surveillance program is in its very early stages – and one of the meeting’s goals was to get a sense from the industries and organizations represented in the room of whether it’s something they’d be interested in joining or supporting. So what was the audience response? For the most part, everyone appeared supportive of the general idea of a surveillance system. Industry representatives from many of the large market players (including Roche, Abbott, One-Touch, Johnson & Johnson and Bayer) expressed cautious enthusiasm – though they rightly pointed out that the “devil is in the details” and that there need to be clear expectations in place, as well as clear consequences for meters and strips that fail. (Since many of the accuracy issues are with generic, cheaper test strips, it makes sense that the big players would want some assurance that their current attention toward accuracy would be rewarded.) The patient advocacy group StripSafely was clearly in support of the program, as were representatives from the American Diabetes Association and JDRF.

The FDA’s primary representative at the meeting was Courtney Lias, Director of the FDA’s Division of Chemistry and Toxicology Devices (which yes, test strips fall under). I was impressed by Lias – she made it clear that she was aware of the audience’s concerns over the FDA’s role in ensuring accuracy, and seemed very willing to help Klonoff and DTS move forward with their efforts. But she also made some important points regarding the limitations of FDA’s power – ones that many patients are not aware of. Most importantly, the FDA does not have the authority to launch a mandatory surveillance program. The only two ways that could happen, explained Lias, is that Congress could pass a law requiring it (unlikely, given the current political climate), or FDA could establish a set of regulations surrounding test strip accuracy. The latter approach would require a lengthy process, including a. creating proposed rules, b. submitting the rules to the public for comment, c. incorporating those comments into a new set of rules and d. getting those new rules approved. It would take five to six years at the least. Again, this is not because FDA or Lias is trying to be difficult; that is just how the agency is required by law to work. It might be possible to establish a voluntary program in the meantime, but Klonoff repeatedly stressed that he and DTS are only interested in working on a program that would be mandatory.

Lias also emphasized that the FDA has no pull or sway with the “payers” – that is, the insurance companies who actually pay the majority of the cost for glucose test strips, and whose presence at the meeting was minimal compared to the other groups represented in the room. While there may have been others that I missed, the only payer I personally saw represented in the room was CMS, the aforementioned Centers for Medicare and Medicaid Services. CMS is the organization that decides what Medicare is going to cover, which in turn influences what the private market is going to cover (usually the private market closely follows Medicare’s coverage decisions, making CMS’s point of view extremely important to people of all ages and income levels).

And that brings us to the enormous elephant in the room, part of the reason that the meeting was called, and why the big industry players are so interested in this issue: CMS recently rolled out competitive bidding for glucose test strips, a highly controversial process which set the reimbursement amounts that Medicare is willing to pay for glucose test strips. These new prices went into effect on July 1, 2013. You can read Medicare’s summary of the changes here; as the New York Times reported (in a poorly researched article that does not acknowledge any potential quality or safety issues), whereas Medicare used to pay $77.90 for 100 test strips, it now will pay $22.47 – a 72% reduction. The New York Times claims that these strips will be provided by 18 “carefully vetted companies” (that is, mail order pharmacies that contract with test strip manufacturers). But the concerning reality is that most, if not all, of the test strips that can be provided for that price come from generic test strip manufacturers – the very group of manufacturers whose strips appear to have the greatest number of quality issues. (For more detail on how test strips are actually made – which gives insight into the importance of quality control – check out this article from Diabetes Forecast.) And, as noted, Medicare’s decisions frequently affect the private market; just because you’re under 65 does not mean that you’re safe.

So, to tie this all together: the United States does not currently have any mandatory (or voluntary!) post-market surveillance program to ensure the accuracy and quality of blood glucose meters and blood glucose test strips. What’s more, a considerable number of test strips and meters currently on the market – particularly low-cost generics — have been demonstrated not to live up to the already lax ISO standards detailed above. David Klonoff and DTS are attempting to bring industry, government and insurance players together to establish a mandatory post-market surveillance program, modeled on other successful standardization efforts like those for cholesterol screenings and HbA1c tests, but even the best-case scenario is still likely to take years. In the meantime, CMS’s competitive bidding program, which heavily (if not entirely) favors low-cost, generic test strip manufacturers that have been demonstrated to have higher quality concerns, has already been rolled out. (Indeed, according to its official comments at the meeting, CMS’s official treatment goals for its Type 2 population do not include aggressive prevention of hyperglycemia; thus CMS does not see a reason that meters necessarily need to be accurate in higher ranges – a subject I will discuss in a separate article later.)

Are you concerned? You should be. To find out more about both the issue and how to express your support/get involved (including contacting your state senators and representatives), visit the patient-advocacy site Stripsafely.org and the meter quality page at http://diabetescareproject.org/.

Catherine Price

Catherine Price (www.catherine-price.com) was diagnosed with Type 1 diabetes when she was 22 years old. She has written for publications including The Best American Science Catherine Price is a professional journalist who was diagnosed with Type 1 diabetes when she was 22 years old. Her work has been featured in publications including The Best American Science Writing, The New York Times, Popular Science, The Los Angeles Times, The San Francisco Chronicle, The Washington Post Magazine, Salon, Slate, Men’s Journal, Health Magazine, The Oprah Magazine, and Outside, among others. A graduate of Yale and UC Berkeley’s Graduate School of Journalism

View Author Blog

Catherine Price

Catherine Price was diagnosed with Type 1 diabetes when she was 22 years old. She has written for publications including The Best American Science Catherine Price is a professional journalist who was diagnosed with Type 1 diabetes when she was 22 years old. Her work has been featured in publications including The Best American Science Writing, The New York Times, Popular Science, The Los Angeles Times, The San Francisco Chronicle, The Washington Post Magazine, Salon, Slate, Men’s Journal, Health Magazine, The Oprah Magazine, and Outside, among others. A graduate of Yale and UC Berkeley’s Graduate School of Journalism