About 1 in every 800 live births in the United States yields a Down syndrome baby—an infant with flat features who is likely to have mild to moderate mental retardation and perhaps heart damage or impaired hearing.

The incidence would be far higher without prenatal screening. Current tests, usually performed in the second trimester, result in early termination of 80 to 90 percent of the pregnancies that would otherwise culminate in a baby with the syndrome. The tests are not always accurate, however, and leave parents only a few weeks to decide what they want to do.

A new multicenter study in this week's New England Journal of Medicine offers evidence that testing for the genetic defect that causes Down syndrome can be done safely and effectively in the first trimester, giving parents more time to consider an abortion or to prepare for the birth.

If termination is the chosen course, first-trimester abortions carry a much lower death rate for the mother than do those performed later. Joe Leigh Simpson, chairman of obstetrics and gynecology at Baylor College of Medicine in Houston, notes that the maternal death rate drops from 7 to 10 per 100,000 procedures to 1.1 per 100,000.

The study enrolled more than 36,000 pregnant women at 15 centers around the country. In the first trimester, their blood was tested for the level of two markers that suggest Down syndrome called PAPP-A and beta hCG. An ultrasound test of nuchal translucency—fluid at the back of the fetus's neck, which can indicate a genetic defect—also was performed. Follow-up blood tests at 15 to 18 weeks were not appreciably more accurate than the combined tests in the first trimester. The key seemed to be adding the nuchal translucency test to the blood measurements.

"Pregnant women will now expect the option of first-trimester screening," wrote Simpson in an accompanying editorial. If the nuchal translucency test is not available, he added, "it is prudent to permit a patient to pursue it elsewhere." Simpson believes that many women, in fact, would benefit by undergoing chorionic villus sampling in the first trimester. CVS involves removing a small bit of fetal tissue for direct examination of the cells and the chromosomes within that tell the genetic tale.

The CVS technique is invasive, so it adds risk of injury or death to the fetus. But Simpson believes the danger is far lower than is generally thought—perhaps 1 in 300 or so in the hands of an experienced practitioner. He suggests looking for someone who has done at least 25 to 50 CVS procedures.

Within three to five years, noninvasive tests with the capability of analyzing the fetus's genetic makeup should begin to appear, says Simpson. "That's the holy grail—to assess fetal tissue directly, without risk."