A risk score for selective screening for gestational diabetes mellitus

Objective: To develop a risk score to predict women who are likely to have an abnormal glucose challenge test (GCT) for gestational diabetes mellitus (GDM) screening. Study design: A cohort of 1876 pregnant women who underwent a GCT between March 2005 and December 2005 at our institution were studied. A multivariable analysis was performed to determine the clinical features that were significantly associated with an abnormal GCT. These factors were incorporated into the equation which was subsequently transformed to the risk score. The validity of this risk score was then tested in a different cohort of 1900 women who underwent a GCT between October 2006 and July 2007. Results: Of 1876 women in the derivation cohort, 586 (31.2%) had positive GDM screening. In a multivariable analysis, age, body mass index, family history of diabetes, prior macrosomia, and history of >= 2 spontaneous abortions were significantly associated with an abnormal GCT. These five variables were added into the equation to determine the risk score. At a cutoff score of >= 380, the sensitivity, specificity, positive predictive value, and negative predictive value to predict an abnormal GCT were 86.9%, 45.0%, 41.8%, and 88.3%, respectively. When the equation with the same cutoff score was tested in the validation cohort, a similar diagnostic performance was obtained. By adopting this risk scoring approach to GDM screening, 41.3% of women could avoid GCT. Conclusion: Our risk score based on clinical data is simple, noninvasive, costless, and reliable to identify women who are likely to have an abnormal GCT. (c) 2009 Elsevier Ireland Ltd. All rights reserved.