The mainstay of modern obstetric diagnosis is the ultrasound scan (or sonography). Similar to radar, real-time scanners employ very-high-frequency pulsed sound waves (3.5-7 mHz, or 3.5-7 million cycles per second) that are sent to the fetus via a transducer placed on the abdomen. Echoes of the sound waves create moving images on the monitor screen.
Virtually all pregnant women are now scanned and given photos or videos to take home as a pleasantly packaged souvenir of their first baby picture.
What is it supposed to do?
It's meant to determine whether your baby is healthy and when you are likely to give birth. Scans can assess gestational age, size and growth, rule out multiple or tubal pregnancies, or ovarian cysts, locate the position of the baby in the womb, and show whether the baby is growing properly or whether it has died.
A scan before 20 weeks also looks for abnormalities such as hydrocephalus, anencephaly, spina bifida, cleft lip or palate and congenital heart problems. Ultrasound is increasingly being used to pick up so-called 'soft markers', or subtle defects which may or may not be serious. It can identify club foot, low-set ears and even problems with facial development.
Ultrasound is now the first port of call for checking chromosomal abnormalities such as Down's syndrome (see box below). At the bare minimum, women are scanned at 12, 18-20 and 34 weeks of pregnancy. Many are scanned 10 or 12 times before giving birth, starting as early as seven weeks into the pregnancy.
In late pregnancy, it is used to rule out placenta praevia, when a low-lying placenta blocks the birth passage.
Is it accurate?
Not especially. In one of the largest studies (33,000 babies), conducted over six years in Oxford, ultrasound picked up only about half of the 725 babies with birth defects. Some 175 fetuses were given a false-positive result - labelled abnormal when they were healthy (Lancet, 1998; 352: 1567-8, 1577-81). A Swedish study found that ultrasound picked up only a third of babies born with serious defects (Acta Obstet Gynecol Scand, 1998; 177: 635-42).
In another study, only 31 per cent of growth-retarded babies were diagnosed before birth, whereas 2 per cent were wrongly identified as such. Only 38 per cent were correctly identified as growth-retarded, even though their mothers had nearly five scans apiece (Acta Obstet Gynecol Scand, 1998; 77: 643-89).
False positives have increased 12-fold as ultrasound is increasingly used to pick up subtle defects (Lancet, 1998; 352: 1567-8, 1577-81). In one study of 400 women, ultrasound diagnosed 250 women with placenta praevia, but it was only present in four (Lancet, 1990; 336: 387-91).
Does it have a positive impact on health?
In fact, scanned babies do worse, according to the research, possibly because they invite more invasive procedures, which don't appear to aid survival. A Norwegian study of babies with defects found that more of the scanned babies died, were delivered sooner, and spent more time in hospital and on ventilators than babies not scanned. Of those with abdominal wall defects, the scanned group were operated on sooner, but had the same outcomes as unscanned babies whose operations were delayed. Of the babies not scanned, 96 per cent survived vs 77 per cent of those scanned (Acta Obstet Gynecol Scand, 1998; 177: 635-42). A German study found that caesarean section and preterm delivery was five times more frequent, and admission into intensive care three times higher, for babies diagnosed by ultrasound before birth (Acta Obstet Gynecol Scand, 1998; 77: 643-89).
Is it safe?
Scientific evidence suggests otherwise. Sonography can cause heating and cavitation (bubbles) which, in turn, can cause genetic and cell damage, especially to the immune system. Thus far, the evidence shows that ultrasound scans:
* restrict growth, causing low birthweight babies (Lancet, 1993; 342: 887-91)
* disrupt brain development, resulting in lefthandedness in boys (but not girls). Of 177,000 Swedish men, those whose mothers had scans were 32 per cent more likely to be lefthanded (Epidemiology, 2001; 12: 618). In Britain, the rate of lefthandedness has more than doubled - from 5 per cent in the 1920s to 11 per cent today. Neurologists believe that slight brain damage can cause righthanded people to become lefthanded
* trigger premature birth, doubling the rate in at-risk women given weekly scans (Am J Obstet Gynecol, 1990; 162: 1603-10)
* increase the risk of miscarriage (Lancet, 1990; 336: 387-91), even among women exposed to occupational sonography for more than 10 hours a week (J Epidemiol Commun Health, 1990; 44: 196-201)
* may delay speech (Can Med Assoc J, 1993; 149: 1435-40)
* May cause dyslexia (Obstet Gynecol, 1984; 63: 194-200)
* expose the fetus to noise, similar to the highest notes on a piano, of 100 decibels - as loud as a subway train arriving at a station (presentation at the Annual Meeting of the Acoustical Society of America, Fort Lauderdale, Florida, 2001)
* affect hormone levels (Br J Obstet Gynaecol, 1982; 89: 694-700).