Background: A 50-g oral glucose challenge test (OGCT) is a widely accepted screening method for gestational diabetes mellitus (GDM), but other options are being considered.

Purpose: To systematically review the test characteristics of various screening methods for GDM across a range of recommended diagnostic glucose thresholds.

Data sources: 15 electronic databases from 1995 to May 2012, reference lists, Web sites of relevant organizations, and gray literature.

Study selection: Two reviewers independently identified English-language prospective studies that compared any screening test for GDM with any reference standard.

Data extraction: One reviewer extracted and a second reviewer verified data from 51 cohort studies. Two reviewers independently assessed methodological quality.

Data synthesis: The sensitivity, specificity, and positive and negative likelihood ratios for the OGCT at a threshold of 7.8 mmol/L (140 mg/dL) were 70% to 88%, 69% to 89%, 2.6 to 6.5, and 0.16 to 0.33, respectively. At a threshold of 7.2 mmol/L (130 mg/dL), the test characteristics were 88% to 99%, 66% to 77%, 2.7 to 4.2, and 0.02 to 0.14, respectively. For a fasting plasma glucose threshold of 4.7 mmol/L (85 mg/dL), they were 87%, 52%, 1.8, and 0.25, respectively. Glycated hemoglobin level had poorer test characteristics than fasting plasma glucose level or the OGCT. No studies compared the OGCT with International Association of the Diabetes and Pregnancy Study Groups (IADPSG) diagnostic criteria.

Limitations: The lack of a gold standard for confirming GDM limits comparisons. Few data exist for screening tests before 24 weeks' gestation.

Conclusion: The OGCT and fasting plasma glucose level (at a threshold of 4.7 mmol/L [85 mg/dL]) by 24 weeks' gestation are good at identifying women who do not have GDM. The OGCT is better at identifying women who have GDM. The OGCT has not been validated for the IADPSG diagnostic criteria.