Electronic Letters to:

Article: Christopher D. Saudek, William H. Herman, David B. Sacks, Richard M. Bergenstal, David Edelman, and Mayer B. Davidson A New Look at Screening and Diagnosing Diabetes Mellitus J Clin Endocrinol Metab 2008; 0: jc.2007-2174v1 [Abstract]

eLetters: Submit a response to this article

Electronic letters published:

Test Order Does Not Matter for Diagnosis

B. Bruce Zellner   (9 September 2008)

Hemoglobin A1c and physical exercise for screening diabetes.

Giuseppe Lippi, Martina Montagnana, Giovanni Targher   (5 August 2008)

Test Order Does Not Matter for Diagnosis 9 September 2008
B. Bruce Zellner, retired Florida Gulf Coast University Send letter to journal: Re: Test Order Does Not Matter for Diagnosis brucezellner{at}comcast.net B. Bruce Zellner	In a recent article in JCEM (1), the authors claimed that “Screening tests are generally distinct from diagnostic tests, favoring sensitivity over specificity.” This view underlies their conclusion that “...it would be a net health benefit to establish specific criteria [for HbAic] as screening tests for diabetes, distinct from those used to establish the diagnosis...”. (p. 2451). This is not correct. The order of thresholds will have no effect on the sensitivity or specificity of the diagnostic process. Lowering the threshold for an HbA1c test used as a screening test to 6.0% and keeping the threshold for diagnostic purposes at 7% achieves the same sensitivity and specificity of the diagnostic process as reversing the order of the thresholds. The probability of a false negative error is the sum of the probability of having a negative screening test (denote this as β1) plus the probability of having a positive screening test (1 - β1) multiplied by the probability of having a negative diagnostic test (β2). Algebraically this is: β1 + (1 - β1)β2 = β1 + β2 − β1β2. The value of this term will not change if the value of β1 is changed to the value of β2 and vice versa. By the same reasoning the probability of the diagnostic process resulting in a false positive error will also be unaffected by the order in which the screening and diagnostic thresholds are used. The valid grounds for using HbA1c as a screening test must rest on its being less costly in terms of patient time and inconvenience and other resources, and identifying a high enough percentage of undiagnosed diabetics from among all those screened. Reference 1. Saudek CD, Herman WH, Sacks DB, Bergenstal RM, Edelman, E, Davidson MB 2008. A new look at screening and diagnosing diabetes mellitus. J Clin Endocrinol Metab 93:2447-2453
Hemoglobin A1c and physical exercise for screening diabetes. 5 August 2008
Giuseppe Lippi, Associate Professor of Clinical Biochemistry Sez. Chimica Clinica, Universit¨¤ di Verona, Osp. Policlinico, 37134 - Verona, Italy, Martina Montagnana, Giovanni Targher Send letter to journal: Re: Hemoglobin A1c and physical exercise for screening diabetes. ulippi{at}tin.it Giuseppe Lippi, et al.	We have read with interest the article of Saudek et al. (1) who concluded that the new screening standards for diagnosing diabetes mellitus might also include a hemoglobin A1c (HbA1c) level ¡Ý7%, confirmed by another HbA1c or a specific test, such as fasting plasma glucose or oral glucose tolerance test. This is a crucial innovation, in that HbA1c is supposed to be less affected by short-term lifestyle changes, thus overcoming a variety of pre-analytical (especially biological) variables that influence the reliability of glucose metabolism for diagnosis diabetes. It is also important to mention, however, that HbA1c is directly correlated with the intensity of aerobic exercise. In particular, we have observed that a regular physical exercise might significantly raise HbA1c values by nearly 4%, due to increased consumption of carbohydrates prior to, during and after exercise rather than reflecting alterations of glucose metabolism (2). The new guidelines of the American College of Sports Medicine and American Heart Association emphasize that physical activity above the recommended minimum should be recommend to provide additional health benefits (3). It is expected, therefore, that a larger number of persons who wish to improve their personal fitness and reduce the risk for chronic diseases may increase the time spent for physical activity. Accordingly, a specific (higher) threshold of HbA1c should be defined for such individuals. References 1. Saudek CD, Herman WH, Sacks DB, Bergenstal RM, Edelman D, Davidson MB 2008 A new look at screening and diagnosing diabetes mellitus. J Clin Endocrinol Metab 93:2447-2453 2. Lippi G, Montagnana M, Salvagno GL, Franchini M, Guidi GC 2008 Glycaemic control in athletes. Int J Sports Med 29:7-10 3. Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, Franklin BA, Macera CA, Heath GW, Thompson PD, Bauman A 2007 Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc 39:1423-1434