Lei Lynn, M.D.*1;Dana Kay, D.O.1;Naji Younes, PhD, RD, CSO2;Sabyasachi Sen, MD, FRCP1 and Vivek Jain, MD1, (1)George Washington University, Washington, DC, (2)George Washington University Milken Institute School of Public Health, Washington, DC

Meeting: Hospital Medicine 2017, May 1-4, 2017; Las Vegas, Nev.

Abstract number: G

Categories: Research Abstracts, Translating Research into Practice

Background: Diabetes Mellitus is one of the most common chronic diseases in the United States and worldwide.1-2 Hemoglobin A1C is a surrogate marker for the average blood glucose level over the previous 3 months.3 In 2009 the American Diabetes Association (ADA) recommended using A1C greater than 6.5% to diagnose diabetes.4 Hemoglobin A1C can be measured by laboratory and Point-of-Care (POC) methods. It is important to compare results obtained by the two modalities in order to determine the proper use of these tests.

Methods: One hundred and fifteen patients were identified to have both POC and laboratory A1C values obtained on the same day, between August 2013 and August 2014. The differences were analyzed and the average difference and coefficient of determination (R2) was calculated. To further investigate the difference between the two methods, a Bland-Altman plot was generated; kappa statistic calculated to assess the percentage of agreement adjusted by chance. We also computed the sensitivity and specificity of the POC method in diagnosing diabetes while using laboratory analysis as the gold standard.

Results: The differences between the two measurements ranged from 0 to 3.2 with an average difference of 0.2% of A1C. The Kappa statistic is 0.761 and R2 was 0.938. Laboratory results were greater than POC in 84% of cases. Furthermore, 78% of the POC measurements fell within 0.5% of A1C from those of in-lab analysis. We also found that in only 4% of cases, medical management was changed based on the difference of these values. Interestingly, the Bland-Altman graph (Graph 1) also indicates that the variability between the two sets of data increases with higher values of A1C especially for A1C greater than 10%. The POC method has a sensitivity of 88.6% and specificity of 96.3% in diagnosing diabetes (Table 1).

Conclusions: Per ADA guidelines, the POC method is not recommended to be used to diagnose diabetes.5 Based on our findings 10.3% of individuals would be missed if one used the POC method to diagnose diabetes. However, given its accuracy supported by our findings it is adequate to be used during routine diabetes follow up visits. The finding from our study can help improve healthcare quality by reducing unnecessary blood draws and clinic visits.  Medicare reimburses $13.21 for each POC hemoglobin A1C test and $24.58 for laboratory analysis. The proper use of these testing modalities can lead to major savings in our health care costs especially considering the prevalence of diabetes and the universal utility of these tests.

To cite this abstract:

Lynn, L; Kay, D; Younes, N; Sen, S; Jain, V . ACCURACY AND THE PROPER USE OF THE POINT-OF-CARE ANALYSIS IN HEMOGLOBIN A1C MEASUREMENT. Abstract published at Hospital Medicine 2017, May 1-4, 2017; Las Vegas, Nev. Abstract G. Journal of Hospital Medicine. 2017; 12 (suppl 2). Accessed April 1, 2020.

« Back to Hospital Medicine 2017, May 1-4, 2017; Las Vegas, Nev.