In a triple-blinded, multisite, independent study, researchers have found that only one third of commercially available FDA-cleared blood glucose meters in the United States could consistently perform to an accuracy standard close to, but actually more lenient than, the FDA’s. Of the 18 BCGs tested, only six were consistently accurate in each of three test sites. Five were compliant in two out of three sites, three in just one, and four failed in all three sites.
The complex study included 1,035 subjects spread across three sites. Each individual participant used six different meters, with different meters assigned to different participants such that 100 participants at each site used each meter. Participants also had their glucose tested at a reference laboratory. Results were considered accurate if they were within 15 percent of the lab value, and meters were considered consistently accurate if 91 out of 100 readings or more were accurate.
The best performing meter was Ascensia Diabetes Care’s Contour Next, which was accurate 100 percent of the time at two sites and 99 percent of the time at the third. The other five consistently accurate meters were Roche Accu-Chek Aviva Plus and Accu-Chek SmartView, Walmart ReliOn Confirm Micro, AgaMatrix’s CVS Advanced and Abbott’s Freestyle Lite.
The four worst performing meters were Philosys’s Gmate Smart, Advocate Redi-Code, BioSense Medical’s Solus V2, and HDI/Nipro’s TRUEtrak.
Why it matters
The study, published in Diabetes Care, took a lot of steps to make its results as credible as possible.
“To our knowledge, this is the largest accuracy study of FDA-cleared BGMs using a consensus protocol created with input from the FDA ever reported in the literature,” researchers wrote in the study. “All strips and monitors were purchased from commercial suppliers without the manufacturers’ knowledge to avoid positive bias that could occur if a manufacturer were to have an opportunity to submit their best performing strips or monitors for testing. Also, the protocol was developed by an impartial expert panel. Testing performed by health care professionals tends to lead to more accurate results than when subjects test themselves, which could lead to a higher level of accuracy in this study compared with other studies where subjects self-test. Finally, this study was triple blinded, which eliminated the possibility of systematic bias based on BGM brand.”
While the end result does create a list of winners and losers, the bigger takeaway is that the majority of meters underperformed — an indictment of the whole category.
“This study in Diabetes Care demonstrated that FDA-cleared BGMs did not consistently perform up to the standards for which they were cleared in the cases 12 of 18 products studied,” lead author Dr. David Klonoff, medical director at the Diabetes Research Institute Mills-Peninsula Medical Center in San Mateo California, told MobiHealthNews in an email. “This type of poor postmarket performance might result in treatment errors in patients who make treatment decisions based on their BGM measurements.”
What’s the trend
In addition to just suggesting that consumers should have a limited level of confidence in their blood glucometers, the study also lends credence to the FDA’s recent focus on obtaining real-world evidence of post-market performance, since each of these devices was previously cleared by the FDA.
“The need for robust postmarket data is a cornerstone of the new FDA Digital Health Precertification Program for software which will require participating companies to provide various types of postmarket performance data,” Klonoff said. “FDA is also supporting a project known as NEST which will assess postmarket performance of various devices. At this time postmarket surveillance of many medical devices appears to be a prudent consumer-friendly practice.”
The study suggests some reasons why devices that originally passed muster might no longer do so without some kind of continuing review.
“The performance of BGMs may diminish over time (i.e., postmarket performance may deteriorate),” authors wrote. “This decline may be due to scale-up issues, manufacturing errors, changes in components between strip lots, other production issues, or improper shipping. Over time, the measured analytical accuracy might no longer represent the sponsor’s initial accuracy data that were submitted to the FDA.”
On the record
Ascensia Diabetes, whose meter came out of the study number one, put out a statement after the study was published.
“We are delighted that one of our key products, the Contour Next BGMS, has been ranked number one in this study,” Ascensia Chief Medical Officer Dr. Sabina Furber said in a statement. “At Ascensia we are very proud of the high levels of accuracy that this system has achieved, as this helps people with diabetes to successfully manage their condition. Accuracy is a critical part of the performance of blood glucose monitoring systems, as patients need accurate readings in order to make appropriate and safe therapy decisions.”