Section III: Cohort/Population Studies (Re-Continued)
Johnson & Daviss Continued: Still considering Johnson and Daviss, I wondered how this study could be considered by MANA to be among the best evidence for the safety of home birth, seeing as it shows a perinatal death rate higher than hospital births from the same time. In their words, “Intrapartum and neonatal death rates were compared with… comparable studies of low risk hospital births.” I looked over the list of the ten studies the authors chose, which numbers permitted the authors to conclude that their death rate was “similar” to hospital birth. The most recent study listed was dated 2002, and the name of the researcher was familiar: Janssen, the same author of the 2009 Canadian study I just looked at. Johnson and Daviss were referencing Janssen’s earlier study, also conducted in Canada, wherein the hospital perinatal death rate was 1.36 per thousand, but the sample size was only 733 women and there was exactly one death.
The other studies dated mostly to the 1980s. The oldest included study used data collected from 1969-1975. Johnson and Daviss apparently had to reach back into the 1960s to find low-risk hospital numbers that would make their death rate seem “similar.” Perhaps the fact that Daviss is a midwife, and Johnson is her husband and noted midwifery advocate, were motivating factors in the selection of comparison studies.
Janssen et al 2002: Still making my way through the MANA list of best evidence for the safety of home birth, the next study sounds familiar: Janssen et al present, “Outcomes of planned home births versus planned hospital births after regulation of midwifery in British Columbia, 2002.” Not surprisingly, this study showed that home birth with a Canadian registered midwife had a lower incidence of almost every kind of obstetric intervention. However, there were a few incidents the authors reported that were not statistically significant due to small sample size, but worth noting: the only two cases of obstetric shock occurred in home births, and three of the four blood transfusions in the study were after home births. Three cases of perinatal death occurred in the home birth group (out of 860 cases), compared to only one death in the hospital group (733 cases). Short case reports were given of all three deaths in the home birth group, suggesting that none of them were likely related to birthplace or attendant. Five babies in the home birth group received assisted ventilation for over 24 hours compared to 0 of the hospital births, a statistically significant finding.
Overall, this study is not as ringing an endorsement of Canadian home birth as the follow up in 2009 (already reviewed) would seem to be. Indeed, in a letter to the publishing journal dated June of 2011, Janssen herself responded to criticisms of this study, writing that “selection bias is unavoidable” when studying home birth, and that her study suffered from a “lack of power” to make “valid conclusions.” With such a “recommendation” from the author, it is interesting that MANA has left it in this list of “best evidence.”
Schlenzka: That brings me to the final study cited in section III: an unpublished dissertation by Schlenzka, dated 1999. An unpublished dissertation is a notable departure from the rest of this list. Although one can argue about possible bias in journals that publish research articles, all real research journals subject all articles to at least a cursory (and, in most highly regarded journals, a truly rigorous) peer review, the process by which other experts fact-check the study and ensure that at least a minimum level of trustworthiness and quality is met. An unpublished dissertation, on the other hand, has literally no known standard for accuracy. The author could write falsehoods, misinterpret data, make grievous factual errors, and misquote sources, and there is no system of accountability.
To place an unpublished dissertation alongside published peer-reviewed studies seems naively disingenuous at best and deliberately deceptive at worst. If this dissertation, dated 1999, is some of the best evidence for the safety of home birth, why hasn’t it at least been picked up by one of Lamaze International’s sponsored journals?