Canadian midwives are university trained and provincially regulated caregivers with advanced resuscitation skills that most in the American study were unlikely to have. (Carolyn Ryan/CBC)

Planned home births are linked to a tripling of infant mortality compared with planned hospital births, according to a U.S. review that drew sharp criticism from Canadian researchers whose data was used in the study.

The article appears in Friday's online issue of the American Journal of Obstetrics & Gynecology. Lead investigator Dr. Joseph R. Wax of the obstetrics and gynecology department at Maine Medical Center and his colleagues analyzed data on 342,056 planned home births and 207,551 planned hospital deliveries in North America and Europe.

The authors found that when babies were born at home there were fewer medical interventions like epidurals and episiotomies, and women in home deliveries tended to have fewer tears or lacerations, fewer cases of postpartum hemorrhage and fewer infections.

The team also came to the conclusion that there was a tripling of neonatal deaths for home births once babies with congenital abnormalities were left out. The tripling of fatalities represents a very small number — from four in 10,000 to 15 in 10,000, acknowledged Dr. Jay Iams, the journal's associate editor.

But Iams said the findings provide an appropriate caution to women that giving birth in a hospital can be safer.

"The study results are consistent with what common sense would tell you," Iams said.

But Dr. Michael Klein, an emeritus professor of family practice and pediatrics at the University of British Columbia, said the U.S. conclusions are "crap" that don't consider the facts.

Data, methods questioned

"It's a politically motivated study that was motivated by the American College of Obstetrics and Gynecology who is unalterably opposed to home birth, and they probably were quite happy to publish this article because it fits with their political position," Klein said.

UBC epidemiologist Patricia Janssen, who collaborated with Klein, said the American study mixes their work with research dating back to the 1970s, and surveys with as few as five recorded births.

"The data and the methods taints the quality of the study," Janssen said, noting the review includes studies where the qualifications of the caregiver were not known.

In the past decade, midwives in Canada have become a profession of university trained and provincially regulated caregivers who have advanced resuscitation skills that most in the American study were unlikely to have.

The journal's editor defended the work as "appropriate," but the study's authors themselves declined repeated requests for interviews from CBC News.