This article is more than 5 years old. Information may no longer be current.
Point-of-care HbA1c monitors require careful selection
Click Here to Manage Email Alerts
With the 2009 expert consensus panel recommendation that HbA1c be used as a preferred test for the diagnosis of diabetes, there has been an increased focus on point-of-care HbA1c monitors in recent years. The availability of accurate, reliable and easy-to-use monitors is important for those practices wishing to perform tests quickly when an onsite laboratory is not available.
|
|
A recent manuscript reported on the accuracy of eight HbA1c meters. Lenters-Westra and Slingerland investigated the performance of the following meters: In2it (Bio-Rad Laboratories); DCA Vantage (Siemens); Afinion (Axis-Shield); NycoCard (Axis-Shield); Clover (Infopia); InnovaStar (DiaSys); A1CNow (Bayer); Quo-Test (Quotient Diagnostics). The researchers used several Clinical and Laboratories Standards Institute (CLSI) protocols to evaluate the imprecision, accuracy and bias of these meters. They also tested each meter with two separate reagent lot numbers to determine whether it would pass National Glycohemoglobin Standardization Program (NGSP) criteria, which state that the 95% CI of the differences between methods of checking HbA1c should fall within ± 0.75%.
In other words, if HbA1c was checked with a standard lab and a point-of-care device (assuming lab is the gold standard), at least 95% of the HbA1c results from the point-of-care meter must be within ± 0.75% of lab results. Results showed that only the Afinion and DCA Vantage meters met the performance criteria of having total coefficient of variation of less than 3% in the clinically relevant range. The researchers said all of the meters demonstrated some inconsistency and calibration concerns between lots of reagents. For example, results with the Afinion meter were consistently lower than the reference results. With the DVA Vantage meter, results were less variable with lower HbA1c levels, but variability increased as HbA1c increased.
There was also slight variability between the two reagent lots. Regardless, these two meters were the only two to pass NGSP criteria with both lots of reagents. Several others passed with one lot, but not the other. Overall, the variability in results with different lots of reagents was unacceptable, the researchers said, adding that manufacturer NGSP certification does not necessarily equate to clinical accuracy. To maintain NGSP certification, manufacturers need only perform one test annually with one lot of reagent.
Another recent article published by Petersen and colleagues compared the DCA Vantage, Afinion and In2it in 88 participants. All three point-of-care devices were found to have excellent correlation as compared with a central lab. However, there were some differences in mean HbA1c levels. The mean HbA1c for lab, Afinion, DCA Vantage and In2it were: 7.9%, 7.3%, 7.2% and 7.9%, respectively. The researchers also calculated coefficient of variation to evaluate precision. The overall imprecision was less than 2.8% for all three meters (should be <3% to accurately distinguish between HbA1c levels). However, of the three, the Afinion meter had the best precision. The researchers said all three meters were acceptable to use, although the Afinion and DCA Vantage did not match with their lab as well.
In another study, Sanchez-Mora and colleagues compared the Afinion and DCA Vantage meters with standard laboratory values in 53 samples. Both meters showed very good correlation to the lab and acceptable precision. The mean difference in HbA1c for the meter vs. lab was 0.28% for the DCA Vantage meter and 0.27% for the Afinion meter. The coefficient of variation for the DCA Vantage meter vs. lab was 3.1% for high HbA1c levels and 2.97% for low HbA1c levels. For the Afinion meter, the coefficient of variation was 1.95% and 2.66%, respectively. The researchers said the meters were acceptable for monitoring diabetes control, but they did not recommend these products for diagnosing purposes based on the level of imprecision.
If you are considering one of these meters for your practice, my suggestion is to test one or two of them in the clinic. Obtain HbA1c levels from the lab and the meters on some patients to gather some data on accuracy and ease of use. This will help in selecting the meter best for you. However, always keep in mind the limitations of any of these point-of-care meters.
It is also important to note that NGSP requirements simply state for HbA1c meters that the 95% CI of 40 samples must be within ± 0.75%. Perhaps more stringent requirements are needed given that HbA1c is now used for diagnosis and not just monitoring.
James R. Taylor, PharmD, CDE, is a clinical associate professor in the department of pharmacy practice at the University of Florida, Gainesville.
For more information:
- Lenters-Westra E. Clin Chem. 2010;56:44-52.
- Petersen JR. Clin Chim Acta. 2010;411:2062-2066.
- Sanchez-Mora C. Clin Chem Lab Med. 2011;49:653-657.
Disclosure: Dr. Taylor reports no relevant financial disclosures.