Prenatal Ultrasound Does Not Improve Perinatal Outcomes by Judy Slome Cohain

[Editor’s note: This article first appeared in Midwifery Today, Issue 102, Summer 2012.]

Abstract: Based on accurate randomized controlled studies, the correct evidence-based recommendation would be for women not to undergo prenatal ultrasound except to assist with turning a breech baby to head down, evaluating ectopic pregnancy and directing the needle during amniocentesis and fetal blood transfusions. First and second trimester organ scans, biophysical profile (BPP), amniotic fluid index (AFI), placental grading 0-III and Doppler umbilical, uterine and fetal artery velocity testing have been tested in randomized controlled studies on tens of thousands of women. They are used to attempt to predict suspected fetal growth restriction (FGR or IUGR), suspected placental insufficiency and suspected postdate pregnancy. They are unable to predict those with sufficient accuracy to direct management that will reduce the number of stillbirths or improve perinatal mortality rates and in most settings result in increased cesarean rates as a result of failed induction.


The experiences of past decades have provided considerable objective evidence that recommendations should be made to women not to undergo amniotic fluid index, biophysical profile and placental grading testing or dating ultrasounds. However, ultrasound screening continues to be promoted as the gold standard and is assumed by many to be exceptionally accurate. This assumption is contrary to what is known, and therefore, current protocols should be reexamined.

In 2009, the US Preventative Services Task Force changed mammography protocols to recommend that most women take mammograms later and less often. Two years later, the task force recommended that men stop taking the prostate-specific antigen PSA test for prostate cancer screening because it caused more harm than good. The treatments caused an increase in erectile dysfunction and loss of bladder control without lowering prostate cancer deaths significantly. The cancers are either easy to detect by manual exam, not life threatening or so aggressive that little could be done about them. Some obstetric protocols are similarly useless in improving outcomes but continue to be used routinely. Shaving pubic hair and giving enemas before birth were discarded in some places as useless. However, although no evidence supports continuous electronic fetal monitoring, IV fluids, hospital birth or delivering on a hospital table for low-risk women, these routines are widely used and promoted.

Screening is meant to diagnose postdate and IUGR pregnancies. When routine ultrasound screening or measuring fundal height suggests that the baby is in the lowest tenth percentile for growth, the mother is referred for an organ scan, amniotic fluid index and biophysical profile, and if those results are questionable, then the woman undergoes Doppler velocity studies. Like the tests for cancer, research shows that these tests are somewhat able to detect findings of questionable relevance but unable to improve outcomes.

IUGR pregnancies have higher rates of perinatal mortality (70/1000 vs. 8/1000) and stillbirth. Postdate pregnancies experience stillbirth in about the same ratio as “normal” pregnancies week by week after 37 weeks. In most countries about 1/1000 fetuses die in utero among low-risk women. Among VBACs and primiparous women the stillbirth rates are 2/1000. If you isolate women over 40, stillbirth rates after week 40 appear to be 2/1000 in small studies. These studies all combine healthy women with women who are high risk, diabetics, smokers and have high blood pressure. There are no studies that show the risk of stillbirth and infant mortality for healthy low-risk postdate women. Early dating ultrasound is used to diagnose IUGR and provide data to determine when pregnancy exceeds 41 weeks.

When postdates or IUGR is diagnosed, it is proposed that inducing the pregnancy will result in better outcomes than expectant management. This assumes that ultrasound can accurately diagnose a fetus that was originally fated to die in utero, and that inducing the pregnancy to end earlier, this fate would be avoided. This is not what science has shown us. Inducing the pregnancies diagnosed by ultrasound as being suspected of having IUGR or of being postdates has not been shown to lower the perinatal mortality and stillbirth rates. The current protocol ignores the inaccuracy of ultrasound and the fact that pregnancy outcome is frequently a result of some controllable factors such as lifestyle habits (tobacco use, alcohol consumption, diet, exercise, sexual activity) and living conditions (education, employment, income, support, safety, environmental quality, housing, water and electricity access, medical care).

Reviewing Ultrasound Research: Biophysical Profile

“At present, there is insufficient evidence from randomized trials to support the use of BPP as a test of fetal well-being” (Lalor et al. 2008). In these randomized trials, there was no ability to predict or prevent perinatal mortality or morbidity or even low Apgar scores with BPP. BPP was associated with a significant increase in induction and caesarean section.

Amniotic Fluid Index

In studies performed, there was no difference in stillbirth, perinatal mortality, low Apgar scores or neonatal NICU admission among women who had AFI screening and those who did not. AFI testing resulted in twice as many cesareans and 60% more inductions than that of controls (Khunpradit et al. 2010). Stillbirth appears to occur with higher rates of oligohydramnios, but no protocol has lowered stillbirth or perinatal mortality rates (Haws et al. 2009).

Doppler Umbilical Artery or Uterine Artery Ultrasound

Existing research does not provide conclusive evidence that the use of routine umbilical artery Doppler ultrasound or combination of umbilical and uterine artery Doppler ultrasound in low-risk or unselected populations benefits either mother or baby (Alfirevic et al. 2010). Furthermore, despite the novelty of Doppler ultrasound testing, it has been evaluated by more random controlled trials than has any other biophysical test of fetal growth or well-being, but there is still little evidence that any form of routine Doppler velocimetry contributes to reductions in stillbirth rates (Haws et al. 2009).

Ultrasound Diagnosis of Placental Grading

There are no placental defects or qualities that can be seen on the delivered placenta (i.e., after birth) on gross or microscopic examination that are found solely in IUGR or postdate pregnancies (Apel-Sarid 2010). Since there are no placental defects unique to IUGR, and since ultrasound is too inaccurate to diagnose them properly even if they did occur, ultrasound diagnosis of placental pathology before birth is useless (Moran 2011).

Second Trimester Organ Scans

The ability of ultrasound to find gross anomalies varies from 13% to 85% (ACOG 2009). There are cases of gross defects like omphalocele, in which the intestines are outside the fetal abdomen, being diagnosed on repeated ultrasounds, but the baby was born with no defect at all. Prenatal diagnosis of serious defects may allow time for parental counseling and delivery planning, but overall it causes more hardship than it prevents because of the high false positive rates.

Babies with very severe defects detectable by ultrasound usually have other defects as well, and survival rates are less than 50% even with early diagnosis and heroic costly prenatal interventions. For a good deal of money in the US, one can pay for surgery on the fetus during the pregnancy for the following findings, but there is also a 50% chance that the surgery will kill the fetus:

  • Spina bifida—occurrence: 1/1428 pregnancies
  • Absence of part or all of diaphragm—occurrence: 1/2200 pregnancies
  • Bladder obstructions—occurrence: 1/4500 pregnancies
  • Sacrococcygeal tumor—occurrence: 1/40,000 pregnancies
  • Twin-to-twin transfusion syndrome—occurrence: 5% of monochorionic twins

Postdate Pregnancy

As much as we would like a prenatal test and subsequent intervention that would prevent postdate pregnancy stillbirth and improve perinatal mortality, there is none. Risk screening and prenatal tests have many false positive results and the majority of adverse outcomes occur in the larger population of women identified as being at low risk of stillbirth. “There is no effective screening test that has clearly shown a reduction in stillbirth rates in the general population” (Smith 2007).

Cochrane Review (Gulmezoglu et al. 2006) found an “extremely small” decrease in perinatal mortality using induction after 41 weeks compared to expectant management. In another study included in Cochrane review, one mother randomized to induction was induced with prostaglandin. Precipitate labor ensued caused by the prostaglandin, with rapid progress to full dilation, severe decelerations, forceps rotation and extraction. The baby sustained high cervical cord injury and quadriplegia. This complication did not appear in the study’s results (Hall 2002).

Four observational studies suggest that induction before 42 weeks might lower stillbirth rates in postterm pregnancies (Sue-A-Quan et al. 1999; Yuan et al. 2005). These studies observed a lowering of stillbirth from 3/1000 to 2/1000 simultaneously during a period where inductions increased. This could have been the result of many factors, such as greater availability of abortion on demand, improved hospital protocols, improved prenatal care, nutrition, less smoking and not reporting newborn morbidity due to lack of attention to high-risk women because the ward was full of women being induced or serious morbidity due to induction. The studies do not report neonatal morbidity.

Another example of damage caused by induction is a case in which the fetus was induced for suspected placental insufficiency and IUGR at 35 weeks; however, the newborn was born with a normal birth weight. Nursing problems ensued and the parents were convinced of the child’s insufficiency and overfed it to the point where the child refused to eat anything from age 11 months. It did not take anything by mouth for the next two-and-a-half years and was fed through a tube directly into its stomach. At age 3-1/2 the parents put the child in a hospital program in which the child was taken away from the parents; the child was eating normally within two weeks. It is my opinion that tragedies such as this one could have been avoided with a term birth.

“The induction of labor between 41 and 42 weeks is a very crude strategy for reducing term and post-term stillbirth rates. Although the risk of fetal death is increased after 42 weeks, many more fetuses die in utero between 37 and 42 weeks than die in the post-term period. It appears that smaller term fetuses run a far greater risk than their larger counterparts, and that current methods of antepartum assessment of the term fetus are still inadequate” (Hollis 2002).

Inaccuracy of Ultrasound

No routine ultrasound screening protocol improves outcomes (Haws et al. 2009). It can be surmised from other research that this is due to the incredible inaccuracy of ultrasound. A recent study found a full third or 34% of ultrasound fetal weight estimates to be outside of the expected +10% to -10% range (Hargreaves 2011). Another study from 2011 found that 1 in 23 (4%) of first trimester ultrasounds in which the woman was told there was no viable pregnancy were wrong and the fetus was fine and the pregnancy went to term (Abdallah et al. 2011). Ultrasound technology cannot reliably predict the presence of a nuchal cord, tell whether a nuchal cord is tight or determine anything regarding the likelihood of hypoxia, IUGR or stillbirth (Cohain 2010). Use of 3D Doppler color ultrasound specifically done to look for a cord in labor detected only 35% of cords around the neck, 60% of cords that were twice around the neck and had a 20% false positive rate, meaning the results said there was a cord around the neck in 20% of fetuses when there wasn’t (Bolten et al. 2009).

The most conspicuous feature of the placenta at full term is a circumvallate placenta. It has a raised white ring around the surface of the fetal side of the placenta formed by a doubled over layer of amnion and chorion. Focused placental sonograms interpreted by experienced sonologists failed to detect the placental edge abnormality in most cases of circumvallation and 35% of normal placentas were diagnosed false positive as probably or definitely circumvallate by one or more observers (Harris 1997). In summary, if ultrasound cannot accurately detect a circumvallate placenta at term or a cord around the neck in labor and 34% of the time is more than 10% above or below the birth weight at term, then how could it possibly measure the speed at which blood is moving and the direction of blood flow in fetal umbilical artery and uterine artery when only 35% of the time it can even find the cord with the most advanced, highest level ultrasound?

Conclusion

Although thoroughly tested and retested, prenatal ultrasound screening does not improve outcomes, except for slight improvements shown in questionable studies of Doppler velocity studies among preterm pregnancies suffering from severe preeclampsia. Since prenatal ultrasound screening does not improve outcomes, it is recommended that pregnant women distance themselves from prenatal testing and focus on lifestyle improvements. Prenatal care can focus on the 80% of controllable factors: diet, exercise, not smoking, limiting drug and alcohol use, education, being employed with adequate income, lifestyle, family and social support, community safety, environmental safety, housing, water supply and energy supply.

Although they would likely show interesting results, there are no randomized controlled studies comparing the outcomes of women who do not undergo prenatal ultrasound, clients of midwives who educate themselves in pregnancy and women who have intense nutritional counseling at home in every trimester, compared to outcomes from women who undergo prenatal ultrasound.

References:

  • Abdallah, Y, A Daemen, E Kirk, et al. 2011. “Limitations of current definitions of miscarriage using mean gestational sac diameter and crown-rump length measurements: a multicenter observational study.” Ultrasound Obstet Gynecol 38 (5):497–502.
  • ACOG Practice Bulletin No. 101: Ultrasonography in Pregnancy. 2009. OBGYN 113 (2):451–61.
  • Alfirevic, Z, T Stampalija and GM Gyte. 2010. “Fetal and umbilical Doppler ultrasound in normal pregnancy.” Cochrane Database Syst Rev Aug 4 (8): CD001450.
  • Alfirevic, Z, T Stampalija and GM Gyte. 2010. “Fetal and umbilical Doppler ultrasound in high-risk pregnancies.” Cochrane Database Syst Rev Jan 20 (1): CD007529.
  • Apel-Sarid, L, A Levy, G Holcberg, et al. 2010 “Term and preterm (<34 and <37 weeks gestation) placental pathologies associated with fetal growth restriction.” Arch Gynecol Obstet 282 (5): 487–92. Bolten, K, FC Chen, N Salomon, et al. 2009. “Antenatal diagnosis of nuchal cord.” Ultrasound Obstet Gynecol 34 (Suppl. 1): 146.
  • Cohain, JS. 2010. “Nuchal cords are necklaces, not nooses.” Midwifery Today 93:46–48, 67–68.
  • Harris, RD, WA Wells, WC Black, et al. 1997. “Accuracy of prenatal sonography for detecting circumvallate placenta.” AJR Am J Roentgenol 168 (6): 1603–08.
  • Hall, P. 2002. “Routine induction of labour at 41 weeks gestation: nonsensus consensus.” BJOG 109: 485–91.
  • Hargreaves, K, M Cameron, H Edwards, et al. 2011. “Is the use of symphysis-fundal height measurement and ultrasound examination effective in detecting small or large fetuses?” J Obstet Gynaecol 31 (5): 380–83.
  • Haws, RA, MY Yakoob, T Soomro, et al. 2009. “Reducing stillbirths: screening and monitoring during pregnancy and labour.” BMC Pregnancy Childbirth 9 (Suppl 1): S5.
  • Hollis, B. 2002. “Prolonged pregnancy.” Curr Opin Obstet Gynecol 14 (2): 203–07.
  • Khunpradit, S, P Lumbiganon and M Laopaiboon. 2011. “Admission tests other than cardiotocography for fetal assessment during labour.” Cochrane Database Syst Rev Jun 15 (6): CD008410.
  • Lalor, JG, B Fawole, Z Alfirevic, et al. 2008. “Biophysical profile for fetal assessment in high risk pregnancies.” Cochrane Database Syst Rev June (1): CD000038.
  • Moran, M, J Ryan, M Higgins, et al. 2011. “Poor agreement between operators on grading of the placenta.” J Obstet Gynaecol 31 (1): 24–28.
  • Smith, GC and RC Fretts. 2007. “Stillbirth.” Lancet 370 (9600): 1715–25.
  • Sue-A-Quan, AK, ME Hannah and MM Cohen. 1999. “Effect of labour induction on rates of stillbirth and cesarean section in post-term pregnancies.” CMAJ 160:1145–49.
  • Yuan, H, RW Platt and L Morin. 2005. “Fetal deaths in the United States 1997 vs. 1991.” Am J Obstet Gynecol 193:489–95.

Judy Slome Cohain is an independent researcher and private practitioner. She enjoys getting feedback and questions:


If you enjoyed this article, you’ll enjoy Midwifery Today magazine! Subscribe now!