Routine ultrasound in pregnancy
Systematic review
|Updated
Key message
The Norwegian Knowledge Centre for the Health Services was asked to evaluate the clinical effects and diagnostic value of routine ultrasound in the first, second and third trimesters of pregnancy. Today all women are offered an ultrasound in the 17th to 19th week of pregnancy. This routine ultrasound offers information on estimation of term, number of fetuses, detection of placenta previa and detection of structural abnormalities and fetal development.
Based on the research literature we found no additional evidence in favour of introducing additional routine ultrasound in the first trimester (week 11 to 14) or in the third trimester (week 32 til 34) of pregnancy. However, an effect of introducing an early ultrasound is the information it gives for identifying fetuses at increased risk of chromosome abnormalities. If this is the aim, the early ultrasound combined with serum markers is the most effective test. Combined ultrasound and serum markers early in pregnancy give better indication for finding fetuses at increased risk for chromosome abnormalitites (and therefore fewer miscarriages due to amniocentesis) than testing based on mothers’ age.
Based on our review of the included literature, there does not seem to be important differences between ultrasound in the first or second trimester with respect to term estimation. When it comes to twin pregnancies, ultrasound in the first trimester will provide added information about mono- or dichorionic placentas. Ultrasound with measurement of nuchal translucency (NT) in week 11+0 to 13+6 supplied with maternal blood tests (CUB) is an effective means of identifying fetuses at increased risk of Down’s syndrome and is sensitivity more sensitive method than ultrasound in the second trimester or risk assessment based on maternal age. Severe structural abnormalities (with normal chromosomes) are more effectively detected in the second trimester. There does not seem to be any evidence in favour of routine ultrasound in the third trimester of pregnancy as an addition to ultrasound in the first and/or second trimester on health outcomes.
Ultrasound is offered to all pregnant women in weeks 17 to 19 of pregnancy and has a high uptake which indicates that women are in favour of it. Based on the literature included, there screening programmes in the first or second trimester does not seem to increase levels of anxiety or worries. For women who experience positive findings (increased risk for abnormalities) the level of anxiety increases. Although further investigations or diagnostic tests indicate that a finding was ”false positive”, these women will continue to suffer from higher levels of anxiety than other pregnant, women throughout the pregnancy. Recent studies indicate that women are well informed about routine ultrasound in pregnancy, but there are challenges related to informing about nuchal translucency (NT) and serum markers.
Ultrasound is part of routine antenatal care in Norway and is offered from the 17th to the 19th week of pregnancy. Norwegian clinicians do not define the routine ultrasound as a screening test (Consensus report of 1995). In international literature, however, it is often defined as such.
When clinicians and health authorities discuss whether to introduce more ultrasound scans in routine practice they might consider requirements that are commonly seen as important in population screening programmes. These include documentation of diagnostic precision, effect of interventions, as well as information about the screening test and how findings are handled.
Background
In Norway, ultrasound screening in pregnancy involves a routine scan offered to all
pregnant women between the 17th and 19th week of pregnancy. The aim is to provide information about foetal gestational age, number of foetuses, the position of the placenta and structural development of the foetus.
Ultrasound in the first trimester of pregnancy (week 11 to 14) might:
• confirm pregnancy with vital foetus
• date the pregnancy
• find multiple pregnancies, identify mono- or diochorioic placentae
• measure nuchal translucency (NT)
• find some structural abnormalities
Pregnant women are not offered routine ultrasound screening in the first trimester. If
they satisfy the criteria for foetal diagnostic tests they will however, be offered
ultrasound assessment between week 11 and 14 with NT measurement and serum
markers (beta hCG and PAPP-A.) Pregnant women at increased risk will be offered further investigations and diagnostic tests such as chorion villus sampling or amniocentesis.
This process is based on individual judgements and might include more ultrasound,
blood tests and more invasive testing at different stages to confirm diagnosis.
Approximately two to three percent of newborns have abnormalities. Not all can be
identified in pregnancy. Below are the rates of some of the abnormalities at birth that
are relevant for ultrasound screening:
Down’s syndrome 1 per 700
Spina bifida 1 per 400 - 500
Severe heart disease 1 per 250
Gastroschisis 1 per 2200-2300
The Norwegian Knowledge Centre for the Health Services was asked by the Norwegian Directorate for Health- and Social Affairs to evaluate the effect of routine ultrasound screening in the first, second and third trimester of pregnancy. We were also asked to evaluate the combined test (ultrasound combined with serum markers) early in pregnancy, as well as the effect of routine doppler. The report includes discussion on ethical aspects of ultrasound screening with the combined test to identify foetus at increased risk for abnormal chromosomes.
Methods
We searched for reports, systematic reviews and new research papers in June 2006 and November 2007. The following databases were searched: The Cochrane Library, DARE, HTA CRD, Ovid Medline, Ovid EMBASE. We ”hand searched” ”Guidelines Finder” (helsebiblioteket.no) for reports and guidelines. This report summarizes the findings from systematic reviews and is hence an ”overview of overviews”.
We based the documentation on published systematic reviews that met
some international recognized methodological criteria (explicit search strategy,
inclusion- and exclusion criteria and descriptions of how research papers were critically appraised).
Systematic reviews are reviews based on a research question for on the basis of which articles have been searched for, critically appraised and summarized all good and relevant research on the topic. A systematic review should give a balanced overview of the evidence available.
Inclusion criteria:
Study design: systematic reviews, Health Technology Assessments (HTA), new
randomised controlled trials (RCTs) and diagnostic studies,
Population: pregnant women;
Intervention: ultrasound scanning at 11-13+6, 17-19th and 32 to 34 (third trimester) weeks of pregnancy, serum markers and Doppler,
Outcomes: health (morbidity and mortality) of foetus, gestational age, parents’ views, mother’s health, quality of life, costs, harm.
Exclusion criteria: articles published in languages other than English or Scandinavian
language, reviews and systematic reviews of low methodological quality and studies
already included in included systematic reviews.
Two members of the review team independently read all identified titles and abstracts. We obtained studies and reports that met the inclusion criteria. Articles obtained in full text were again judged against the criteria. Two members of the team independently critically appraised the literature. We assessed the methodological quality on included reports, systematic reviews and primal studies using checklists developed at the Knowledge Centre. The studies and
reports were judged to be of high, moderate or low quality.
Results
We identified 890 titles and abstracts in June 2006. Eighty-one references were obtained in full text. We included eight reports (Health Technology Assessments - HTA), six systematic reviews and nine primary studies plus one conference abstract. The update of the search in November 2007 led to 1010 total hits and we included 27 new reports and primary studies.
Ultrasound scanning in the first half of pregnancy generally led to a more precise
determination of gestational age, basically meaning that term date more often matched the date of birth + 14 days (ultrasound was more precise than method based on the mother’s last menstruation in from 3,7 to 19 percent of cases in three studies). The greater discrepancy, the more accurate the gestational date set by scanning. In 60 percent of cases the ultrasound scanning led to a later term date and in 30 percent to an earlier date.
A meta-analyses of six randomised trials (RCTs) showed a significant reduction of inductions caused by post-term in the group offered ultrasound scanning (odds ratio (OR) 0,61, 95% confidence interval (CI) 0,52 to 0,72). There was no difference between groups offered scanning in the first or second trimester when assessing perinatal mortality or morbidity.
The results from two HTA-reports showed that screening for NT is more effective than screening based on mother’s age alone to identify foetuses with Down’s syndrome (73 to 93 percent in an unselected population). The rate of positive test results and sensitivity is higher. The combined test (NT plus serum markers) has a higher sensitivity for identifying the risk for Down’s syndrome than NT-scanning alone, with the same false positive rate.
Ultrasound in the second trimester (weeks 17 to 19) with respect to chromosomal
anomalies was also assessed. Based on the documentation included we found that
scanning in second trimester had lower sensitivity for chromosomal anomalies than in the first trimester (weeks 11+1 to 13+6). The sensitivity for Down’s syndrome varied from 45 to 100 percent, with a false positive rate from one to 37 percent. The sensitivity was higher when the scanning was supplemented with serum markers (triple or quadruple test). This, in addition to maternal age was the most effective way to detect foetus with increased risk for Down’s syndrome in second trimester.
With regard to structural congenital heart disease we found, based on two HTA-reports and one systematic review, that NT-screening is a moderate to low effective method to identify foetus at increased risk for severe heart disease. The sensitivity for finding heart disease by ultrasound scanning is moderate to low, both in the first and second trimester. This depends, however, on the skills of the person performing the scanning. Fewer cases of heart disease are identified in the first than in the second trimester. The sensitivity for identifying severe heart disease in the first trimester is between 0 and 58 %. For screening in second trimester it is between 0 and 66 percent. The false positive rate (when investigations conclude) is very low (less that one percent). Findings from observational studies indicate that there is a better prognosis for children where severe heart disease was found in pregnancy than for those who were identified after birth.
In clinical practice where increased NT and normal chromosomes is detected, the women will be offered further investigation regarding the fetal heart. A Swedish randomised trial compared the sensitivity for heart disease in first end second
trimester. The sensitivity was 11 percent and the test-positive rate was 0,09 percent in the first trimester. In second trimester the sensitivity was 15 percent and the rate testpositive rate was 0,10 percent. The difference was not statistically significant. The authors note that the level of skills for detecting heart disease in first trimester was low in this study.
The number of structural abnormalities that may be identified in the first trimester of
pregnancy is still limited. The sensitivity when screening for structural abnormalities in routine practice scanning in week 11 to 14 of pregnancy varies between 9 to 54 percent, while the rate of false positive (individuals with no abnormalities, but test-positive (FPR)) is low. An exception is screening for anencephaly which has a sensitivity of over 85% in the first trimester. Ultrasound screening for structural abnormalities in the second trimester has a sensitivity of 5.8 to 85 percent depending on several factors such as definition of abnormality, gestational age, repeated ultrasound scans and the skill of the person performing the investigation. The final false positive rate (FPR) is low, less than one percent.
Ultrasound scanning in the first trimester can identify chorionicity (sharing the same
placenta) in multiple pregnancies. There is a possible risk for twin to twin transfusion
when twins share a placenta (monochorionicity). We lack recent studies that show
whether early detection influence perinatal morbidity or mortality. Studies included in our report (from before 1991) found no difference in perinatal mortality (OR 0,81; 95% CI 0,36 to 1,80), morbidity in new-born period, complications around birth or length of pregnancy between those screened early or not. We lack clinical trials of multiple pregnancies where the diagnostic value of screening in the first trimester is compared with screening in second, and the effect on birth outcomes and the childrens’ health.
Regarding the effect of routine ultrasound scanning in the third trimester, we found little evidence indicating that routine ultrasound give important improvement in health
outcomes. Only when it comes to assessment of foetal position near the time of birth is there some evidence that it might be of importance. Ultrasound based on clinical
indication is today part of the standard care if a deviant position or low placenta is
suspected.
There is limited evidence to suggest any positive health outcomes when using routine
Doppler in pregnancy. There is some evidence that investigation with Doppler (in second and third trimester) can be part of risk estimation for pre-eclampsia, intrauterine growth restriction or perinatal mortality. It can, however, not be ruled out that Doppler in the first trimester might be harmful to the foetus.
High uptake of the Norwegian ultrasound scanning programme indicates that women are in favour of it. When asked, they respond that they want information from ultrasound early in pregnancy. Routine ultrasound screening programmes do not lead to higher rates of anxiety, worries or depression. Regarding these outcomes there is no difference between women who have been offered the scan routinely in the first, second or third trimesters of pregnancy. For women who experience positive findings (increased risk for abnormalities) the level of anxiety increases. Although further investigations or diagnostic tests showed that the finding was ”false positive”, these women will continue to suffer from a higher level of anxiety than other pregnant women throughout the pregnancy.
Another finding is that there is a lack of proven effective methods to improve the way women are given information about screening tests and how they can be supported to make well informed choices. On the other hand, routine ultrasound scanning is well established and asked for, and the screening does not seem to increase anxiety or worries in most women. We know less about the effect this has on her partner.
Conclusion
There do not seem to be any clinically important differences in the effect of offering
routine ultrasound screening in the first or second trimester on gestational age.
Regarding multiple pregnancy, an early ultrasound might rule out that twins share the
placenta, this is important information regarding possible twin to twin transfusion.
Screening for NT supplemented with serum markers (combined test) in the first
trimester is an effective method to identify foetuses with an increased risk of Down’s
syndrome. This test gives a higher sensitivity than ultrasound scanning in the second
trimester or screening based on mother’s age.
Serious structural abnormalities with
normal chromosomes, are more effectively detected in the second than in the first
trimester.
We found little evidence in favour of routine ultrasound in the third trimester
on health outcomes for the mother or child. Routine ultrasound does not increase
anxiety, worries or depression in most women offered the test.