Kurzfassung

In this project we examine the impact of homebirths on perinatal mortality. Homebirths are becoming increasingly common in several developed countries and policy makers are increasingly propagating homebirths. Health care systems in most developed countries, however, are not fully geared toward homebirths. Only The Netherlands has an obstetric system that is fully geared toward homebirths and where homebirths are the standard for low-risk deliveries (currently, around 25% of all deliveries...In this project we examine the impact of homebirths on perinatal mortality. Homebirths are becoming increasingly common in several developed countries and policy makers are increasingly propagating homebirths. Health care systems in most developed countries, however, are not fully geared toward homebirths. Only The Netherlands has an obstetric system that is fully geared toward homebirths and where homebirths are the standard for low-risk deliveries (currently, around 25% of all deliveries take place at home). Studying homebirths in The Netherlands enables us to investigate whether homebirths can possibly be safe in a system that is arguably almost perfectly adapted to providing for safe homebirths. This will give valuable information for other countries considering whether and how to propagate homebirths.

As yet, there is no conclusive evidence on whether homebirths affect health risks. Partially, this is due to the fact that previous research usually simply compared child mortality among home and hospital births adjusting for a set of covariates. Since women choose where they give birth and have more knowledge on their own health and risks than researchers do, probably no set of covariates can possibly completely adjust for differences between women giving birth at home vs. in hospital. If so, then results from previous studies are likely to be biased. Other approaches are then needed.

We use data from the Netherlands Perinatal Registry (PRN), which contains detailed records of 99% of all births in The Netherlands between 2000 and 2008. Our analyses utilize an instrumental variables approach. I.e. we take a variable that affects where women deliver, but that itself does not affect perinatal mortality. We use this instrumental variable to predict where women deliver and then only use the variation in birthplace induced by the instrumental variable in our estimates of the effects of homebirths on perinatal mortality. This enables us to estimate this effect causally, provided that the assumptions that the instrumental variable affects birthplace, but not mortality hold.
» weiterlesen» einklappen