By Gene Emery

(Reuters Health) - A study comparing the risks of in-hospital and out-of-hospital births in Oregon finds the odds of infant death are more than two times higher for planned out-of-hospital births.

The results, published online December 30 in the New England Journal of Medicine differ from those of a recent, larger Canadian study that found newborn death rates were not higher among women who gave birth at home (see Reuters Health story of December 21, 2015, here: http://reut.rs/1SmV3wT).

But the authors of both studies told Reuters Health the conflicting findings probably reflect underlying differences in the health systems in Oregon and in Ontario, where the Canadian study was done and where the medical guidelines for at-home births are more mature.

"The findings are different, but I think that, on a system level, they paint a similar picture," said the chief author of the Oregon study, Jonathan Snowden of the Oregon Health and Science University in Portland.

"According to the Canadian study, when out-of-hospital births are integrated into the healthcare system and there are regulations in place for how to select women for out-of-hospital births, I think it can be a safe option," he said. "In our Oregon study we found that, absent a lot of these systems, there were small but slightly increased risks of neonatal mortality in the out-of-hospital setting."

The absolute risks were still small, researchers noted. For example, in the Oregon analysis, the odds of death were 39 per 10,000 deliveries done at home or at a freestanding birth center. With a hospital delivery, the chances were 18 per 10,000.

No extra risk was seen among Oregon women who already had a child.

The fact that the Oregon study didn't screen out higher-risk pregnancies from the home birth group may also help explain why the newborn death rate was higher there, said Eileen Hutton of McMaster University in Hamilton, Ontario, chief author of the Canadian study. In Ontario, such women are not eligible for home birth, she said by email.

Among the Canadian women, 25% attempting to give birth at home were ultimately transferred to a hospital. That's much higher than the 16% who were transferred in Oregon, where birthing centers were included in the tally.

Snowden said that's probably because "there is more of a formal system for supporting women who choose out-of-hospital birth" in Ontario, and it's easier to get them or their baby admitted to the hospital quickly if there is a complication.

"We think with more of those agreements in place (in the U.S.) we could improve our outcomes across all birth settings" to make them closer to success rates in Canada, he said.

In addition, women trying to give birth at home tended to lack health insurance, so the decision to have a home birth, and to transfer to hospital if a problem looms "may be determined in part by financial ability, rather than suitability to giving birth in an out-of-hospital setting," she said.

Unassisted vaginal deliveries were more than five times more common with out-of-hospital births. The rate of cesarean section was 5.3% in women who attempted home delivery and nearly 25% in the hospital.

However, the odds of the baby suffering at least one seizure were 3.6 times higher with out-of-hospital delivery and the likelihood that a woman would need a blood transfusion nearly doubled.

With home or birthing center delivery, the rates of other physician interventions beyond cesarean section were lower.

In an editorial in the Journal, Dr. Michael Green and Jeffrey Ecker of Massachusetts General Hospital in Boston write, "Unfortunately, there is no way to discern from these data which obstetrical interventions - if any - that were significantly more common among women with planned hospital births contributed to their reduced rates of perinatal complications and which were 'unnecessary.'"

Home births in the U.S. accounted for just 0.89% of all births in 2012 but that rate has been rising - it was 0.56% in 2004. The rate at birth centers rose from 0.23% to 0.39% over that period.