The diagnosis and classification of gestational diabetes mellitus: Is it time to change our tune?

Abstract Objective: This study was designed to determine the impact on our population of adopting the Carpenter and Coustan criteria for gestational diabetes mellitus in place of the currently used National Diabetes Data Group criteria, to review the evidence supporting replacement of the National Diabetes Data Group criteria with the Carpenter and Coustan criteria, and to propose analogous diagnostic criteria for diabetes in pregnant and nonpregnant women. Study Design: The National Diabetes Data Group criteria and the proposed Carpenter and Coustan criteria were both used to retrospectively review medical records of patients screened for gestational diabetes mellitus during 1995 and 1996 in the Kaiser Permanente Northwest Division. Computerized search was performed on automated data systems and software was used for statistical analyses. A MEDLINE review of relevant literature was conducted. Results: Of 8857 pregnant women screened for gestational diabetes in 1995 and 1996, 284 (3.21%) met the National Diabetes Data Group criteria, whereas 438 (4.95%) met the Carpenter and Coustan criteria. We estimate that in our population use of the Carpenter and Coustan criteria in 1996 could at best have reduced the prevalence of infants weighing ≥4000 g from 17.1% to 16.9% and the prevalence of infants weighing ≥4500 g from 2.95% to 2.91%. Conclusions: Replacing the National Diabetes Data Group criteria with the Carpenter and Coustan criteria would increase by 54% the number of pregnant women with a diagnosis of gestational diabetes mellitus and would also increase costs, while only minimally affecting prevalence of infant macrosomia. The medical literature does not provide compelling evidence for adopting the Carpenter and Coustan criteria. Standardization of both measurement of venous plasma glucose level and diagnostic criteria for gestational diabetes mellitus is an important goal. Parallel criteria for diagnosis and classification of diabetes mellitus in pregnant and nonpregnant women should be developed. (Am J Obstet Gynecol 1999;180:1560-71.)