"It's not as dire as your presentation shows," MacDormand said in March when the research was shared at a conference. "The absolute level of infant mortality is very low for home and birthing-center births."
Fewer than 1 percent of U.S. births now occur at home, about 30,000 a year, CDC data show. However, home births have been inching up since 2004, and a handful of states - including Pennsylvania, Montana, and Idaho - have a rate two to three times higher than the national average.
Those states have lots of the kinds of women who rely on midwives: poor, underserved patients and affluent, educated ones who want to experience childbirth with minimal medical intervention.
"I'm not a crazy hippie," said Dina Jenney, whose third child was born at home in Phoenixville. "We did a lot of research, a lot of soul-searching. But there were just so many things about home birth that were wonderful."
The crux of the study is that babies born at home are 10 times more likely to be stillborn and four times more likely to suffer seizures or neurological "dysfunction" than babies born in hospitals.
The authors suggest this is because of delays in emergency care for patients who must be transferred to a hospital. Several of the authors cited this issue in a recent editorial challenging the decision by the American College of Obstetricians and Gynecologists to acknowledge that women have the right to choose where to deliver.
"In the hospital," said Amos Grunebaum, director of obstetrics at New York Weill Cornell Medical Center and lead author of the study, "if you need an emergency cesarean section, you can do it quickly. If you have to transfer the woman to the hospital, it may not happen quickly enough."
Critics say the researchers read too much into the birth certificates, which are a sketchy source for medical outcomes. For example, the researchers used an indicator that a newborn was unresponsive - namely, a zero Apgar score five minutes after birth - as evidence of a death.
Another criticism involves the concept of risk. No matter the setting, zero Apgar scores and neurological problems were rare. However, these were rarest in hospitals, so, in comparison, homes were riskier. For each 1,000 births, zero Apgar scores occurred in 0.16 deliveries by hospital doctors, 0.55 deliveries by midwives in freestanding birth centers, and 1.63 home births.
"Using tiny numbers of rare outcomes to portray risks is not appropriate," said Dana Perlman, interim director of the Midwifery Institute of Philadelphia University.
To resolve the debate, the researchers advocate making hospital maternity units more homelike, with more midwives. Indeed, the study found even better outcomes for hospital deliveries by midwives than by physicians.
"I understand why women want" home birth, said Grunebaum. "They want an experience with less intervention - no Pitocin, no episiotomy, no epidural, no cesarean section."
Midwives say that's a nice idea - one that guides hospital-connected birth centers - but not realistic.
"You can make a hospital look like a five-star hotel, but it's not going to be homelike," said Rising Moon Midwifery founder Kate Aseron, who was Jenney's midwife.
At the March conference, MacDormand suggested another way.
"Even if [home births] are a little more risky, what do you do about it? Your solution seems to be, 'OK, let's shut it down.' My solution is ... let's make it safer. And the way to make it safer is to license midwives, have uniform training, support transfers, and make it part of the health-care system."