Assessing the safety of freestanding birth centers with propensity score analysis

Abstract

Freestanding birth centers (FBCs) offer prenatal and childbirth services outside hospital settings where nurse-midwives are the primary providers for women with low risk pregnancies. Although previous research provides evidence supporting the safety of FBCs, critiques of methodological procedures call into question the validity of these claims because of potential effects of hidden bias common in observational studies. The purpose of this study is to assess safety by testing whether FBCs, compared to hospital-based services, produce comparable pregnancy outcomes. Data are obtained through Tennessee birth and death certificate records over twelve years (1990-2002). The FBC group includes women who gave birth at the center and those women who were transferred to a hospital during labor. Outcomes from FBCs are compared with two hospital groups: (1) women with low risk pregnancies, and (2) women with midwife-attended hospital births. The latter group controls for factors related to choice of provider. For the FBC versus midwife-attended comparison, pregnancy outcomes are controlled for pre-existing risk profiles by using both logistic regression and a propensity score analysis. A sub-classification analysis on the propensity score is used to determine if group differences are properly balanced on all measured pre-existing risks.

Results find that FBCs are significantly less likely than hospital groups to have occurrence of low birth weight, small for gestational age, preterm birth, fetal distress, and maternal complications during labor. Incidences of infant abnormal conditions and low Apgar scores are statistically equivalent in both groups. Giving birth to an infant who is large for its gestational age occurs more frequently in the FBC group than either hospital group; the outcome is only statistically significant when compared to the midwife-attended hospital group. Decisions to limit comparison groups to low risk pregnancy and midwife-attended hospital births, in addition to the use of propensity scores to balance pre-existing risk factors across groups, contribute to minimizing the bias found in observational studies. The inclusion of these techniques provides stronger evidence that FBCs can be a relatively safe option for low medical risk pregnant women safe option for eligible women.