Six studies were included (n=3,563). All were prospective cohort studies and five enrolled consecutive patients. Studies generally scored well on the QUADAS assessment. Items rated as no or unclear by more than one study were blinding of the index test to the reference standard results and vice versa, availability of clinical information to the person interpreting the test result and reporting of uninterpretable results and withdrawals. None of the studies were considered to be of high quality.
Four studies assessed random glucose testing performed once during pregnancy. Sensitivity ranged from 15% (95% CI 8 to 25%) to 100% (95% CI 75% to 100%). Corresponding specificities ranged were 98% (95% CI 97% to 98%) to 37% (95% CI 35% to 37%).
One study assessed screening in the first and second trimesters. In the first trimester sensitivity was 71% (95% CI 46% to 88%) and specificity was 80% (95% CI 80% to 81%) In the second trimester sensitivity was 38% (95% CI 14% to 69%) and specificity was 82% (95% CI 82% to 83%).
One study included patients with risk factors for gestational diabetes mellitus who received five plasma glucose measurements in 24 hours during the second trimester. The lowest sensitivity was 25% (95% CI 18% to 27%) and the highest sensitivity was 47% (95% CI 37% to 56%). Corresponding specificities were 97% (95% CI 91% to 99%) and 74% (95% CI 66% to 81%).