Prenatal strep screening works -- but not as well as hoped

This article was originally on a blog post platform and may be missing photos, graphics or links. See About archive blog posts.

Streptococcus B infections transmitted by the mother are one of the most common causes of infections in newborns, occurring in about 1 in every 3,000 births. In order to reduce the rate of infections, the Centers for Disease Control and Prevention in 2002 recommended that all pregnant women be screened for the bacterium and treated with antibiotics if it is found to be present. The first analysis of results from this screening program has shown that the screening works, reducing the number of new infections by 27% over a five-year period. Unfortunately, the analysis also showed that nearly two-thirds of the children who contracted strep B had mothers who tested negative for the germ.

An estimated one-quarter of pregnant women have strep B infections, but usually suffer no ill effects unless they have underlying conditions such as cancer or diabetes. Among healthy women, about 1 in 200 transmit the infection to the infant, typically during delivery. But if the baby is premature, the woman’s membranes are ruptured more than 18 hours before delivery, or she has a fever at the time of delivery, the risk rises to 4 in 100. Infections in the mother are typically treated with penicillin or amoxicillin or, if the woman is allergic to those drugs, with cefazolin.

Symptoms among newborns who contract the infection include pneumonia, blood infections and meningitis. About 3% to 5% of infected infants die despite treatment.

Epidemiologist Melissa K. Van Dyke of the CDC and her colleagues studied births in a database that includes strep B cases in parts of 10 states, including five counties in the San Francisco area, in 2003 and 2004, comparing the results with those of a similar study in the late 1990s. They reported today in the New England Journal of Medicine that the rate of screening rose from 48% in the early study to 85% in the current one. They also observed that the antibiotics given to the pregnant women appeared to be quite effective in preventing neonatal infections and that none of the infants born to such women suffered from an allergic reaction to the antibiotic.

But they also found that about 60% of the cases of strep B in full-term newborns occurred among mothers who had tested negative for the infection. The other 40% of the cases were among women who had not been tested or those who had failed treatment with the antibiotics.

Researchers are at a loss to explain the high rate among women who tested negative. Some experts think it might be a matter of timing: It is recommended that the test be administered at 35 to 37 weeks into the pregnancy, but some may have been tested too early. It is also possible that the test was a true negative and that the woman simply contracted the infection in the short time between testing and delivery.

-- Thomas H. Maugh II