• Laurence Kennedy, William H Herman, and GOAL A1C Study Team.
• Division of Endocrinology, Department of Medicine, Shands Hospital at University of Florida, Gainesville, Florida 32610, USA. kenneal@medicine.ufl.edu
• Diabetes Technol. Ther. 2005 Dec 1; 7 (6): 907-12.

BackgroundThe Glycemic Optimization with Algorithms and Labs At Po1nt of Care (GOAL A1C) Study assessed the effect of titration monitoring strategies and methods of A1C testing on glycemic control in patients with type 2 diabetes failing oral therapy and beginning basal insulin glargine. The availability of both point-of-care (POC) and central laboratory A1C values provided an opportunity to evaluate correlation and statistical agreement between these methods of testing. This analysis forms the basis of the current report.MethodsThis is a 24-week, randomized, four-arm, open-label study conducted in 7,758 subjects enrolled at 2,130 sites. At baseline, patients had A1C measurements both by POC testing using the A1cNow device (Metrika, Inc., Sunnyvale, CA), which applies an immunoassay method, and by central laboratory analysis using ion exchange high-performance liquid chromatography. These measures were compared statistically.ResultsAn r value of 0.72 was calculated for POC and laboratory A1C assessments. Although the mean POC A1C values were in agreement with the central laboratory values, there was a large range in individual POC A1C values.ConclusionsPOC testing of A1C in predominantly primary care settings using the A1cNow device was correlated with central laboratory results. The correlation was less than expected based on each method's reproducibility data. Although there was agreement between the average POC A1C values and the corresponding central laboratory values, the dispersion of individual POC A1C values was large. Thus, we conclude that these two methods of A1C testing should not be used interchangeably.

18 keywords...

hide…