When the fetus measures large after a normal glucose screen, a repeat OGTT finds diabetes in one in five

Design

Systematic review and meta-analysis of six cohort studies

Population

2166 women with normal 24–28 week glucose screening and later suspected large fetus or polyhydramnios

Primary outcome

Detection rate of gestational diabetes on repeat OGTT after 28 weeks

Effect

15.0% (95% CI 9.9% to 21.0%); 20.8% for suspected large fetus, 4.8% for polyhydramnios

An AJOG meta-analysis pooled six cohort studies of 2166 women who had normal glucose screening at 24 to 28 weeks and then had a repeat oral glucose tolerance test after 28 weeks because ultrasound suggested a large-for-gestational-age fetus or polyhydramnios.

The repeat test diagnosed gestational diabetes in 15.0% overall (95% CI 9.9% to 21.0%). The yield was 20.8% when a large fetus was suspected and 4.8% for isolated polyhydramnios. Women diagnosed late had more neonatal hypoglycaemia (OR 1.82, 1.18 to 2.81), more caesarean delivery (OR 2.00, 1.47 to 2.72) and more induction of labour (OR 2.27, 1.32 to 3.89); birthweight itself was not significantly higher.

A normal mid-pregnancy screen does not exclude diabetes later in pregnancy. Where growth accelerates, a repeat test is worth considering — in India, where gestational diabetes is common and a single non-fasting test is often used, the question is especially live. Whether treating late-diagnosed diabetes improves outcomes was not tested.

• Consider a repeat oral glucose tolerance test after 28 weeks when ultrasound suggests a large-for-gestational-age fetus, even after a normal earlier screen.
• Expect a lower yield when polyhydramnios is the only finding — about one in twenty in this analysis.
• If late gestational diabetes is diagnosed, alert the neonatal team to plan glucose monitoring after birth.
• Record the date and type of the original screen so a later test can be interpreted against it.

Why it matters

This challenges the assumption that a normal 24–28 week screen closes the question of gestational diabetes.

Don't overread it

These are cohort data; they show how often diabetes is found, not that treating it late improves outcomes.

The statistics, in plain English

A detection rate of 15.0% means about one in six or seven women tested had diabetes. An I² of 91% means the six studies differed a lot, so the pooled rate is a rough guide. Odds ratios of about 2 mean those outcomes were roughly twice as likely in the late-diagnosed group, but that group may differ in other ways too.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.