Scientists now say a third of infant deaths are due to premature births - a much larger percentage than previously thought.
In the past, "preterm birth" has been the listed cause of death in fewer than 20 percent of newborn fatalities. But that number should be 34 percent or more, said researchers at the U.S. Centers for Disease Control and Prevention.
That's because at least a dozen causes of newborn death are actually problems that go hand-in-hand with premature births, such as respiratory distress syndrome caused by underdeveloped lungs.
"This brings preterm birth, as a cause of death, to the kind of level that we think it deserves," said the CDC's Dr. Bill Callaghan, the lead author of a study appearing Monday in the journal Pediatrics.
The revised statistic may lead to greater efforts to counsel pregnant women about taking care of themselves and avoiding actions that can lead to preterm births - such as smoking and drug use.
It also may help organizations lobbying for more research into why some women who follow medical advice still have preterm babies. The March of Dimes is advocating to expand federal research into preterm labor and delivery and the care and treatment of premature infants.
At issue is how to label the causes of deaths for infants who die before they reach their first birthday.
"Preterm birth" generally describes infants who are born before 37 weeks gestation, and the term is also used as an official cause of death. Two-thirds of infant deaths occur in children who were preterm, but their cause of death is often attributed to one of the several specific problems that can occur in preterm babies.
"The only way that an infant gets assigned ("preterm birth") is if there's nothing else on the death certificate," said Callaghan, a senior scientist in the CDC's maternal and infant health branch. "That may result in an underestimation of what the real problem is."
Callaghan and other researchers reviewed about 28,000 infant deaths that occurred in 2002.
More than 4,600 of those - or 17 percent - were attributed only to preterm birth. But the researchers also grouped in roughly 5,000 other deaths that were attributed to preterm-related conditions including respiratory distress syndrome, brain hemorrhage and maternal complications such as premature rupture of membranes.
In that counting, nearly 9,600 births - or 34 percent - could be classified as preterm, Callaghan said. The researchers believe that figure is conservative and likely underestimates the true picture.
Experts have generally understood the burden of preterm birth on infant deaths, but the new study sorts out the data and provides specific numbers, said Carol Hogue, an Emory University professor of maternal and child health.
"It isn't just an exercise in moving deck chairs around on the Titanic," said Hogue.