The effectiveness of screening all pregnant women versus pregnant women with risk factors for gestational diabetes
Systematic review
|Published
The Norwegian Directorate of Health asked The Norwegian Institute of Public Health to summarize available research about the effect of screening all pregnant women for gestational diabetes compared with screening only women with risk factors for gestational diabetes.
Key message
A National guideline about gestational diabetes was published in Norway in April 2017. There has been some discussion about these new guidelines, and there is disagreement about whether to offer all pregnant women a glucose tolerance test or whether to offer the test just for pregnant women with an increased risk of gestational diabetes.
The Norwegian Directorate of Health asked The Norwegian Institute of Public Health to summarize available research about the effect of screening all pregnant women for gestational diabetes compared with screening only women with risk factors for gestational diabetes on identification of gestational diabetes and health outcomes on mother and child.
Results from one quasi-randomized controlled trial found that universal screening possibly identify more women with gestational diabetes that screening only women with known risk factors. We did not identify any studies that assessed health outcomes on mother or child.
We did not identify any other studies with control group that looked at this comparison. There is a lack of good research about the effect of universal screening for gestational diabetes compared to screening on indication.
Summary
Background
Gestational diabetes develops during pregnancy and results in increased blood sugar levels. Untreated gestational diabetes may be harmful for both mother and child. During pregnancy the need for insulin is increased. Gestational diabetes develops if the production of insulin does not match the increase in demand. Gestational diabetes can develop without given symptoms additional to normal symptoms of pregnancy, for example tiredness, and the pregnant woman may not be aware that her blood sugar levels are higher than normal.
Prevalence of gestational diabetes in Norway is in the range of 3% to 8% measured using the WHO 1999-criteria. If we use the new criteria for gestational diabetes, as many as one in ten, or maybe even more, pregnant women may be diagnosed with gestational diabetes. There is no national or international agreement about the levels of glucose that should be used as diagnostic criteria. The Norwegian guidelines define gestational diabetes as fasting blood glucose level between 5.3 to 6.9 mmol/l and/or a two hours level of blood glucose between 9.0 and 11.0 mmol/l after a glucose tolerance test. Women with gestational diabetes are recommended treatment. The aim of the treatment is to lower the blood sugar level to a level that is healthy both for the mother and child. Treatment can consist of life style changes such as increase physical activity, change in diet, regular measurements of blood sugar levels, and potentially medication. Unidentified gestational diabetes increase risk of complications, while treatment can lower the risk of complications to about the same level as for pregnancy without gestational diabetes.
The authors of a Cochrane review from 2017 conclude that more pregnant women may be diagnosed with gestational diabetes if all pregnant women are screened compared to screening of only the women who have risk factors. The authors noted that they have a low level of confidence in the effect estimates, and that they therefore cannot draw clear conclusion about the effect of universal screening compared with screening on indication. Included in the Cochrane review was only one quasi-randomized controlled trial with 3152 participants.
Objective
We have conducted a systematic review to address the following question:
What is the effect of screening all pregnant women for gestational diabetes compared with screening only women with risk factors for gestational diabetes on identification of gestational diabetes and health outcomes on mother and child?
Method
Research librarian Gyri Hval Straumann developed and conducted the systematic literature search. The search for published and planned systematic reviews was conducted in the following databases: Medline (Ovid), Embase (Ovid), Cochrane Library, Epistemonikos, and PROSPERO. No systematic reviews which also included prospective controlled trials were identified.
For this systematic review, the search was conducted in November 2018 in the following databases: Cochrane Central Register of Controlled Trials, Medline (Ovid), Embase (Ovid), CINAHL (Ebsco).
Inclusion criteria:
Study design: randomized controlled trials, prospective studies with a control group, interrupted time series
Population: pregnant women from week 24
Intervention: screening with a glucose tolerance test of women with risk factors
Comparison: screening with a glucose tolerance test of all pregnant women. Screening with a glucose tolerance test of women with other risk factors
Outcomes:
Mother, short term: gestational diabetes, pre-eclampsia, caesarean, complicated birth/ difficult delivery (including stuck shoulder), mental health
Mother, long term: diabetes, overweight/obesity, cardiovascular illness
Infant/ child short term: large for gestational age, macrosomia (birth weight >4500g), small for gestational age, asphyxia, stillbirth
Child, long term: overweight/obesity and diabetes
Two people independently assessed all references and abstracts. Potentially relevant articles were collected and read in full text, also by two people independently of each other. One author extracted information from the included study, and another author double checked the extraction. We assessed our confidence in the effect estimate for the relevant outcomes using GRADE.
Results
The literature search identified 3 416 references. After reading them all, we assessed ten of them to be potentially relevant and collected them in full text. We included one quasi-randomized trial from Ireland. This study included 3152 pregnant women where half of them were screened for gestational diabetes between week 24 and 28 (universal screening). In the other half, only women with risk factors were screened, at week 32. The screening was a glucose tolerance test. This study is the same as the one included in the above mentioned Cochrane review from 2017.
Current available documentation is based on the one study that compared universal screening for gestational diabetes with screening based on risk factors, they found that:
- It is possible that more pregnant women are diagnosed with gestational diabetes when all women are screened that when only women with risk factors are screened. We have low confidence in this effect estimate
- There are no studies that have studied the health effects on mother and child, hence, we do not know if or how the different screening strategies affect the health of mother and child
Discussion
We have included just one quasi-randomized controlled trial that compared universal screening for gestational diabetes with screening on indication. The only reason we have not included non-randomized controlled trials or observational studies is that we did not identify any that studied this comparison (that fulfilled the inclusion criteria).
In our search for systematic reviews and primary studies we have found several reviews about the effect of screening for gestational diabetes, There is no need for further systematic reviews before there has been conducted more primary studies. The need for more research on this question should be assessed in conjunction with relevant decision makers such as the population it includes, relevant health care personnel, researchers and politicians.
Conclusion
Results from one quasi-randomized trial found that universal screening may identify more women with gestational diabetes than selective screening limited to women with known risk factors. We did not find any studies that assessed the health effects on mother and child. We did not find other controlled studies that assessed this question. There is a lack of solid evidence about the effect of universal screening compared with only screening women with known risk factors for gestational diabetes.
