(3 of 5)
Lead author Dr. Joseph Wax cautions against alarm, noting that the absolute risk of neonatal death is still extremely small in any birthing environment in the U.S. According to the review, the rate of neonatal death was 2 to 3 for every 1,000 home births. The rate among hospital births was 1 for every 1,000 births. "Home birth is quite safe for the baby," says Wax, a maternal-and-fetal-medicine specialist at Maine Medical Center. "But not as safe as a hospital birth."
All the more reason for women to eschew home birth, say obstetricians. Wax's study found that the increase in neonatal death could be attributed in part to babies' breathing difficulties and failed resuscitation factors associated with inadequate midwife training and lack of access to hospital equipment. The obvious solution: give birth in a hospital. "During the labor process, emergencies can arise that we cannot predict. In some of those cases, you only have moments to intervene successfully," says Dr. Erin Tracy, an ob-gyn at Massachusetts General Hospital and an outspoken detractor of home birthing. "It's a tragedy in those rare instances [of infant death] where medical intervention could have saved the life of the baby."
Informing the Patients
In terms of scientific evidence, meta-analysis sets a high bar. Because it aggregates data from multiple studies, a meta-analysis is useful for revealing medical trends that cannot be picked up by individual studies. Perhaps more important, the results of meta-analyses hold great sway in doctors' offices. They are kind of like medical Cliffs Notes: doctors often prefer to read a single review paper rather than 20-odd original studies to make a judgment about a particular treatment or intervention.
It would seem that the editors of the American Journal of Obstetrics & Gynecology, who highlighted Wax's paper as an Editor's Choice, hoped the study would inform patient decisions. The 12 studies analyzed were from seven countries (two from the U.S.; the rest from Australia, Britain, Canada and Western Europe) and compared data on maternal and infant outcomes in a total of 342,056 planned home births and 207,551 planned hospital births. But two independent experts in meta-analysis who reviewed the paper for TIME concluded that it was weak and methodologically flawed. Other critics say some of the studies included are outdated or misleading, thus limiting the conclusions of the review.
One such study, published in the journal Obstetrics & Gynecology in 2002, compared the outcomes of 6,133 home births and 10,593 hospital births in the state of Washington from 1989 to 1996. But the paper did not make clear whether any of the babies who died had birth defects that would have resulted in death regardless of where they were born. The study also could not determine in every case where exactly the birth had been intended to occur; the authors relied on birth-certificate data, which indicated whether a baby was delivered at home but not whether the home birth was accidental.