page contents

Section I: Meta-analyses and Systematic Reviews


coffeecomputerpensOlsen & Clausen: The first study listed, by Olsen and Clausen, was from 2012 and was published in the Cochrane Database of Systematic Reviews. It purported to be a systematic review of randomized controlled trials of home birth. This seemed to me to be a curious and difficult way to study home birth, as it would require randomly selecting women to give birth at home. Although randomized controlled trials are considered the “gold standard” of research, how could one ethically assign women to birth at home or hospital? I didn’t have to wonder long: the systematic review only determined that there are no usable randomized controlled clinical trials. They only were able to find one study that fit their criteria for inclusion, and the sample size of eleven women was too small to be of any statistical use to anyone.

Leslie & Romano: The second study listed was a systematic review of nonhospital birth studies by Leslie and Romano, published in the Journal of Perinatal Education in 2007. Now, I do think it is important to note that the Journal of Perinatal Education is the official “journal” of Lamaze International. This means that it is sponsored by an organization whose bread and butter is based on natural childbirth education. This does not mean they are incapable of publishing valid findings, but it is prudent to be aware of the money behind a publication when evaluating a study.

The Leslie and Romano study found that nonhospital birth results in far fewer interventions, such as cesarean sections, use of intravenous fluids, and use of medical pain control. They cite three studies in their finding that perinatal mortality[1] rates are “similar” to that of the hospital: Gulbranson (1997), Janssen et al (2002), and Olsen (1997). The Janssen study is covered under Section III, discussed later in this paper. The Gulbranson study was conducted in New Zealand and determined a perinatal death rate of 2.97 per thousand; I clearly needed to find out if this was truly “similar” to hospital rates, so I highlighted this number and made a mental note to come back to it.

Olsen: The third study used by Leslie and Romano for perinatal mortality comparison is also the last study in section I: Olsen from 1997, “Meta-analysis of the safety of home birth.” This is the second listing from Olsen in the first three studies listed, and it hails from 18 years ago. It was published in the journal Birth. It may interest the reader to know that Birth is published on behalf of Lamaze International, just like the Journal of Perinatal Education.

Because Olsen’s meta-analysis was completed in 1997, all of the six studies included are rather aged (ranging from 1977 to 1994) and four of the six studies are international. International studies are of limited value due to the extreme differences in midwifery training from country to country. The two U.S. studies included, Mehl (1977 in Wisconsin) and Durand (1992 in Tennessee), are not only small (sample sizes equal to or less than 1707) and one could argue outdated, but they also are not included in MANA’s list of “best evidence” for home birth safety. If MANA is going to hold up a meta-analysis as best-evidence, why would they not include the studies that powered the meta-analysis on the same list?

Olsen concluded in 1997 that, “No empirical evidence exists to support the view it is less safe for most low-risk women to plan a home birth.” It allows the reader to hope that as we move into more recent research, a stronger conclusion (such as evidence supporting that it is safe, rather than a lack of evidence that it isn’t) could be reached.

[1] Perinatal mortality means death of the fetus or newborn near the time of birth. It generally includes fetal deaths toward the end of pregnancy, deaths during labor and birth, and deaths that occur during the first week of life.

Print Friendly
Section I: Meta-analyses and Systematic Reviews (Part 4)

Post navigation

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>