We appreciate the comments made by Vallee-Pouliot et al concerning our publication “Early and total neonatal mortality in relation to birth setting in the United States, 2006-2009.”

The aim of our study was to evaluate neonatal death in relation to birth settings in the United States, and we showed that planned midwife births at home were associated with unacceptably high neonatal mortality rates. As the authors acknowledge, there is considerable variation in requirements for licensure of midwives in the states, especially when compared with Canada, England, and the Netherlands. Further evaluation of outcomes based on midwife credentials was beyond the scope of this study and may be addressed in future studies.

The authors correctly state that transfers from midwives within the hospital-to-hospital physicians decrease midwife neonatal deaths and potentially increase physicians’ neonatal mortality rates because these transfers usually are attributed to physicians in the Centers for Disease Control and Prevention database. Vallee-Poulio et al neglect to mention that this principle also applies to transfers from attempted planned midwife home births to the hospital. Neonatal mortality rates among supposedly low-risk planned midwife home births were >2-fold higher (neonatal mortality rates in planned midwife home births when compared with physician hospital births: 1.26 per 1000 births vs 0.55 per 1000 births for hospital physicians). If neonatal deaths after transfers from planned home births to the hospital were included in planned home births, the already unacceptably high neonatal mortality rates among homebirths would likely be even worse than the Centers for Disease Control and Prevention data show.

The authors’ reference to Dutch midwives is inappropriate because Dutch midwives have strict uniform selection criteria for homebirths. but US midwives do not. In a recently accepted study, we show that planned midwife home births in the United States have an unacceptably high number of risks.

As the authors state, differences of patient characteristics between home births and hospital deliveries are stated clearly in our patient characteristics Table. This is an accepted scientific method to account for differences among patient populations.

We believe that the major contributor to the increase in neonatal deaths in planned midwife homebirths is a function of its location (with delayed access to hospital care and cesarean deliveries) and the unavailability of essential services, medication, equipment, and well-trained personnel. Inadequate midwife education and training and the absence of patient selection criteria among midwives who perform home births in the United States may contribute further to the increased neonatal mortality rates in home births.

We believe the unacceptably high neonatal mortality rates among US planned midwife home birth shown in our study must be disclosed as an essential component part of the informed consent process with all patients who consider a planned home birth.