Screening for gestational diabetes mellitus: a systematic review for the U.S. Preventive Services Task Force.

BACKGROUND In 2003, the U.S. Preventive Services Task Force concluded that evidence was insufficient to advise for or against routinely screening all pregnant women for gestational diabetes mellitus. PURPOSE To review evidence about the benefits and harms of screening for gestational diabetes. DATA SOURCES Databases (MEDLINE, Database of Abstracts of Reviews of Effects, Health Technology Assessment Database, National Institute for Health and Clinical Effectiveness, and Cochrane Library) were searched for reports published from January 2000 to 15 November 2007 (and from 1966 to 1999 for additional studies on screening at less than 24 weeks' gestation), citations in the 2003 evidence report, and studies identified through consultation of experts and searches of bibliographies. STUDY SELECTION English-language studies that used standard 1- or 2-step testing for gestational diabetes and that evaluated at least 1 of the following outcomes: neonatal mortality; brachial plexus injury; clavicular fracture; admission to a neonatal intensive care unit for hypoglycemia, hyperbilirubinemia, or the respiratory distress syndrome; maternal mortality; and preeclampsia or pregnancy-induced hypertension. DATA EXTRACTION 2 reviewers evaluated 1607 abstracts, critically appraised 288 articles, and qualitatively synthesized 13 studies. DATA SYNTHESIS No randomized, controlled trials that directly evaluated the risks and benefits of gestational diabetes screening were found. One good-quality randomized, controlled trial of treatment of mild gestational diabetes in a screening-detected population supported a reduction in serious neonatal complications and showed that gestational diabetes treatment also reduced the risk for gestational hypertension. Very limited evidence was found to evaluate early screening for gestational diabetes (before 24 weeks' gestation). Limited evidence suggests that serious maternal hypoglycemia is rare with treatment and that overall quality of life is not worse among women receiving gestational diabetes treatment compared with women not receiving treatment. LIMITATION The literature is limited by lack of a consistent standard for screening or diagnosis of gestational diabetes. CONCLUSION Limited evidence suggests that gestational diabetes treatment after 24 weeks improves some maternal and neonatal outcomes. Evidence is even more sparse for screening before 24 weeks' gestation.