In the last article in this series on natural childbirth, I reviewed evidence suggesting that routine prenatal ultrasound does not improve birth outcomes for mothers or babies, and that organizations like the American College of Obstetricians & Gynecologists recommend ultrasound scans only for specific reasons.
In this article I’m going to review evidence on the safety of routine ultrasound and Doppler scanning, and make recommendations based on that research.
The potential adverse effects of ultrasound
Ultrasound adversely affects body tissues in three primary ways:
- Acoustic streaming
The sonar beam can cause heating in the tissues beings scanned. During normal pregnancy, increases in whole-body temperature of up to 4.5 degrees F (2.5 C) are presumed to be safe, and research suggests that elevations of tissue temperature up to 1.8 – 2.7 degrees F (1.0 to 1.5 C) caused by ultrasound are also safe.
The degree to which ultrasound machines raise temperatures in the tissues depend on which tissues are scanned. Bone heats faster more than soft tissue, which in turn heats more than fluid. Heating is also dependent upon exposure time, the intensity of the machine, and whether the transducer is held stationary or moved frequently.
Doppler ultrasound, which uses continuous rather than pulsed waves, has been shown to cause significant heating – especially in the baby’s developing brain. A recent study suggests that heating in late-pregnancy fetal tissues exposed to normal pulsed and continuous Doppler ultrasound may be higher than what is regarded as safe: 2.5 to 10.4 degrees F (1.4 – 5.8 C) respectively.
A 1997 study found that significant temperature increases can occur at or near to bone in the fetus from the second trimester, if the beam is held stationary for more than 30 seconds in some pulsed Doppler applications. This in turn can heating of sensory organs incased in bone.
Though both animal and human studies have shown that temperature elevations can cause abnormal development and birth defects, so far human studies have not shown a direct causal relationship between diagnostic ultrasound exposure during pregnancy and adverse effects to the developing baby.
However, it must be pointed out that all human epidemiologic studies were conducted with commercially available devices predating 1992 with acoustic outputs not exceeding an intensity of 94 mW/cm2.
Current limits in the U.S. have risen dramatically, and now allow intensities of up to 720 mW/cm2 – more than 7 times the limit in 1992. This means we have no large, population-based studies examining the effects of ultrasound at the much higher intensities commonly used today.
This is highly problematic, because, according to a 2001 review called “Guidelines and Recommendations for Safe Use of Doppler Ultrasound in Perinatal Applications“:
When modern sophisticated equipment is used at maximum operating settings for Doppler examinations, the acoustic outputs are sufficient to produce obvious biological effects, e.g. significant temperature increase in tissue or visible motion of particles due to radiation pressure streaming effects. The risk of inducing thermal effects is greater in the second and third trimesters, when fetal bone is intercepted by the ultrasound beam and significant temperature increase can occur in the fetal brain.
A 2007 study reached a similar conclusion:
(1) thermal rather than nonthermal mechanisms are more likely to induce adverse effects in utero, and (2) while the probability of an adverse thermal event is usually small, under some conditions it can be disturbingly high.
Cavitation occurs in tissues with significant pockets of gas (such as the lung and the intestine) after birth. There is no consensus on the significance of cavitation effects in human fetal tissue, but some evidence suggests that mammalian tissue may contain microbubbles that are susceptible to cavitation effects.
Acoustic streaming involves a jet of fluid created by the ultrasound wave, which causes a mechanical shearing force at the cell surface. While the effect of this force is not fully understood, research suggests that it may change cell permeability and have adverse effects on both early and late prenatal and postnatal development.
Animal studies suggest diagnostic levels of ultrasound may cause harm
One study found brain hemorrhages in mouse pups exposed in womb to pulsed ultrasound at doses similar to those used on human babies.
Another study found exposing adult mice to dosages typical of obstetric ultrasound caused a 22 percent reduction in rate of cell division and a doubling of the rate of apoptosis of cells in small intestine.
Other research has found that ultrasound induces bleeding in the lung among other mammals, including newborns and young animals.
The American Institute of Ultrasound in Medicine concluded:
There exists abundant peer-reviewed published scientific research that clearly and convincingly documents that ultrasound at commercial diagnostic levels can produce lung damage and focal haemorrhage in a variety of mammalian species…. The degree to which this is a clinically significant problem in humans is not known.
I want to be clear: we can’t extrapolate the results of these animal studies to humans, and so far, many longer-term human studies have not shown harm to the fetus from diagnostic ultrasound exposure. However, when the stakes are this high (i.e. the health of our children), I believe the animal study results warrant caution and further study before plowing ahead with ultrasound technology.
Some human studies also suggest harm…
Single or small studies on humans exposed to ultrasound have shown that possible adverse effects include premature ovulation, preterm labor or miscarriage, low birth weight, poorer condition at birth, perinatal death, dyslexia, delayed speech development, and less right-handedness.1
This is especially true for Doppler ultrasound, which is used in specialized scans, fetal monitors and handheld fetal stethoscopes (sonicaids). Ordinary scans use pulses of ultrasound that last only a fraction of a second. The machine uses the interval between pulses to interpret the echo returns. Doppler, on the other hand, uses continuous waves – which leads to much higher levels of exposure than with pulsed ultrasound.
A large UK study found that healthy mothers and babies that received two or more Doppler scans to check the placenta had more than 2 times the risk of perinatal death compared to babies unexposed to Doppler.
An Australian study found babies that received more than 5 Dopplers were 30% more likely than babies that received routine (pulsed) ultrasound to develop intrauterine growth retardation (IUGR). This is ironic because Doppler is often used specifically to detect IUGR.
A randomized clinical trial published in 1996 split 2,743 women into two groups: one that received a single doppler at 18 weeks and further scans only when clinically indicated and another that received 5 Doppler readings during pregnancy. When compared with the regular group, and after adjusting for other confounding variables, babies in the intensive group tended to be shorter when measured at birth and on days 2-3 of age. There were also reductions in the circumferences of the chest, abdomen and mid-arm, and in the skin-fold thicknesses of the triceps, parascapular and subscapular regions – although these differences weren’t statistically significant.
A later study in Lancet found a similar effect on fetal growth in women receiving repeated ultrasound exams, although measures of growth and development later in childhood (up to age eight) were similar in both groups.
A case control study of 72 children who had undergone a formal language evaluation found that children with delayed speech had a higher rate of ultrasound exposure in utero than normal controls. Their findings suggested that a child with delayed speech was twice as likely to have been exposed to prenatal ultrasound. (Note that this is a correlation and doesn’t prove causation.)
…while other studies suggest ultrasound is safe
On the other hand, a recent World Health Organization (WHO) review of the literature in 2009 concluded that “exposure to diagnostic ultrasonography appears to be safe.”
However, even in this review they did express some concern about the association between left-handedness in males and exposure to Doppler ultrasound. Non-righthandedness is sometimes a marker of damage or disruption to the developing brain. 2
Another review in 2008 concluded:
At this time, there is no specific reason to suspect that there is any significant health risk to the fetus or mother from exposure to diagnostic ultrasound in obstetrics. This assurance of safety supports the prudent use of diagnostic ultrasound in obstetrics by trained professionals for any medically indicated examination.
What are we to make of these conflicting results?
One of the reasons it’s difficult to make any clear determinations from the research is that the methodology of many of the trials is faulty. For example, in a randomized controlled trial in Sweden in the late 70s that found no differences in hearing, vision, growth or learning at age 9 in kids exposed and unexposed to ultrasound, 35% of the supposedly unexposed group actually had a scan. This means there was no true control group.
In fact, there are very few studies at all comparing outcomes between women who have received no ultrasounds at all, and women who have received ultrasound, during pregnancy. This is the kind of research we need to make an accurate determination of the effects of ultrasound on mothers and developing babies.
Another problem which I mentioned earlier in the article that casts doubt on current safety assessments is that scanning intensities used today are up to 6-8 times higher than they were in the 1990s, when all of the large population-based studies assessing ultrasound safety were done. This means we have no data on the large population level indicating whether ultrasound scanning at the frequency and intensity commonly practiced today is safe.
In a 2002 review of the safety of ultrasound in the prestigious journal Epidemiology, the authors concluded:
Until long-term effects can be evaluated across generations, caution should be exercised when using this modality during pregnancy.
Weighing the risks and benefits of routine ultrasound
The evidence I’ve reviewed here does not prove that a single ultrasound scan at relatively low intensity performed by a skilled operator will cause harm to a developing baby.
However, there is sufficient evidence that multiple pulsed ultrasound scans, or as few as two continuous wave Doppler scans, or any ultrasound scan performed by an unskilled operator may cause harm. There is also a pressing need for large epidemiological studies to be performed using the higher ultrasound intensities commonly used today.
When making a decision to perform any medical diagnostic test or procedure, benefits must always be weighed against risks. It’s rarely a black or white issue. Clearly, if ultrasound was 100% safe with no potential for harm, there would be little medical reason not to perform routine ultrasound during pregnancy.
But the evidence indicates that ultrasound is not risk-free, so we are forced to weigh whatever benefits routine ultrasound might provide against the potential harm it could cause. That harm could be physiological – including the effects we’ve covered in this article – or it could also be psychological. And of course psychological effects like stress and anxiety very quickly produce real physiological changes in both the mother and the baby.
The authors of the 2010 Cochrane review on ultrasound remind us that:
Subjecting a large group of low-risk patients to a screening test with a relatively high false positive rate is likely to cause anxiety and lead to inappropriate intervention and subsequent risk of iatrogenic morbidity and mortality.
Translation: giving all women ultrasounds may end up introducing unnecessary stress and anxiety, which in turn can produce real complications that would not have otherwise occurred. The screening for potential abnormalities can become a self-fulfilling prophecy.
Routine ultrasound also increases the likelihood that more tests will be performed, which could also increase the risk of complications. In a trial of Doppler in 4,187 low-risk pregnancies in France, the only significant result of using doppler was an increase in the number of ultrasound and doppler examinations subsequently conducted. There were no other effects on the management of pregnancy.
And then there is the new trend of non-medical fetal ultrasound (also known as ‘keepsake’ ultrasound), which is defined as using ultrasound to view, take a picture, or determine the sex of a fetus without a medical indication. This practice involves long exposures using 3-D and 4-D ultrasound techniques, which have not been studied adequately, and do not provide the patient with medically appropriate data.
For this reason, major organizations like the American College of Obstetricians and Gynecologists, AIUM and the FDA do not support keepsake ultrasound.
Recommendations and personal experience
Based on the evidence we’ve reviewed in this article, I recommend minimizing exposure to ultrasound during pregnancy in the following three ways:
- Using ultrasound only when medically indicated, i.e. only when a problem is suspected, rather than as a routine screening to determine the sex of the baby or check on its development.
- Minimize total exposure time (by choosing a skilled and knowledgeable operator).
- Minimize exposure intensity (i.e. avoiding Doppler during the first trimester especially).
Steps 2 and 3 are especially important in light of the lax regulation of ultrasound and the incredibly high variability of skill of ultrasound operators. In the USA, UK & Australia, ultrasonography training is voluntary – even for obstetricians – and the skill and experience of operators varies tremendously. Most operators don’t follow the scientific literature and aren’t aware of the safety issues involved with repeated and high intensity exposure.
In cases where abnormalities are suspected, a woman may wish to have an ultrasound to determine whether an early termination is warranted. The moral, ethical, economic and social issues involved in that decision are far beyond the scope of this article, and cannot be answered through research alone.
I support the right of a woman and her partner to choose what is best for them in this regard; after all, it is they who have to live with the results of their decision.
My wife Elanne and I chose not to have any ultrasound scanning done during her pregnancy, even though she was 39 when she conceived and thus at higher risk for certain genetic abnormalities.
We discussed it at length. In the end, we decided that what we might lose in getting the scans was greater than what we might gain. We felt the stress that a minor or uncertain problem on the scan could produce, and the worry and concern we’d feel waiting for the next scan, and the next one… would interfere with our relationship with our growing baby.
We also decided that we would carry the pregnancy through to full term, regardless of whether an early scan (had we had one) turned up a risk for an abnormality. It took us 2 years of off-and-on attempts to get pregnant, and because of the relatively high risk of false positives and the uncertain results of those genetic tests, we were willing to live without the information a scan might have given us. If it hadn’t taken us so long to get pregnant, or if Elanne had been younger when she did get pregnant, perhaps we would have made a different decision. Perhaps not.
I’m in no way suggesting this is the right choice for everyone. I strongly recommend that you educate yourselves about the risks and benefits of ultrasound first, and then consider your own personality, circumstances and values before making a decision. No one – not me, your doctor, or any other authority – can make this choice for you.