Research

Measuring the effect of midwives

In November, the Journal of Midwifery and Women’s Health published this article:¬† Relationship Between Hospital-Level Percentage of Midwife-Attended Births and Obstetric Procedure Utilization.1 I find myself returning to this paper to ponder the implications.

When you read the abstract, you’ll notice that hospitals with more mid-wife attended births had lower odds of birth by cesarean and episiotomy; but no difference in odds of induction or severe maternal morbidity. You will also notice that hospitals were identified as more midwife births if they had more than 40% of births attended by midwives. So what am I pondering?

For starters, this data was taken from New York, after the change in the Nurse Practice Act — is this an improvement in midwifery care or has practice stayed the same?

And this data is take from hospital data because the question was about how having midwives changes the hospitals statistics — should we also be asking how the hospital changes the midwives statistics? Do midwives provide different care for low risk women when they attend the labor in a hospital?

And reading this article, it is easy to assume the strongest driver of variation in care of women is hospital level characteristics, but we are continually reminded that variation between women and between hospitals only explains a small part of the variation in the care women receive 2— would we get a different picture of the effect of midwives on hospital statistics if we had a way to account for the regional differences between the hospital locations?

These are big questions.

1.
Attanasio L, Kozhimannil KB. Relationship Between Hospital-Level Percentage of Midwife-Attended Births and Obstetric Procedure Utilization. J. November 2017. doi: 10.1111/jmwh12702
2.
Sebasti√£o YV, Womack L, Vamos CA, et al. Hospital variation in cesarean delivery rates: contribution of individual and hospital factors in Florida. A. 2016;214(1):123.e1-123.e18. doi: 10.1016/j.ajog.2015.08027
Jennifer (Author)