New birth center study does not show what its authors claim it shows

Isolated Aid Kit With Green Dollar Sign

A new study of birth centers, published yesterday, is being hailed by midwifery advocates. While it does demonstrate that giving birth in birth centers accredited by the American Association of Birth Centers is safe, it does NOT show that birth centers have a lower C-section rate than hospital care for comparable risk women, and therefore, it does NOT show that increasing birth center births would save millions of dollars.

The study is Outcomes of Care in Birth Centers: Demonstration of a Durable Model, by Stapleton, Osbourne and Illuzi.

The study found that birth in accredited birth centers was very safe:

There were no maternal deaths. The intrapartum fetal mortality rate for women admitted to the birth center in labor was 0.47/1000. The neonatal mortality rate was 0.40/1000 excluding anomalies.

This is comparable to death rates for low risk hospital birth.

So far, so good. Then the authors, in their efforts to promote birth centers, go far beyond the data. They claim:

The cesarean birth rate in this cohort was 6% versus the estimated rate of 25% for similarly low-risk women in a hospital setting.21 Had this same group of 15,574 low-risk women been cared for in a hospital, an additional 2934 cesarean births could be expected. The Medicare facility reimbursement for an uncomplicated cesarean birth in a hospital in 2011 was $4465.49 Given the increased payments for facility services for cesarean birth compared with vaginal birth in the hospital, the lower cesarean birth rate potentially saved an additional $4,487,524. In total, one could expect a potential savings in costs for facility services of more than $30 million for these 15,574 births.

But the C-section rate for comparable risk births in the hospital is NOT 25%. It is far lower, in the range of 4-8%. Indeed, I’m not sure where they even got their estimate of a 25% C-section rate for comparable risk births since does not appear in the reference specifically cited to support this number.

The first rule of scientific comparisons is to compare like to like. Therefore, when looking for the appropriate comparison group for women who give birth in accredited birth centers, we must restrict the group to low risk women, with single babies, at term, without intrauterine growth retardation. Furthermore, we must exclude from the comparison group any women who have pre-existing medical problems or pregnancy complications, since they would be excluded from delivering at the birth center.

In addition, we must take into account that women choose to deliver in a birth center are a self-selected group who differ markedly from the general population. They are more likely to be white, married and well educated and they are far less likely to smoke, drink alcohol or be obese.

What is the appropriate comparison group? It’s women who choose to deliver in the hospital with a CNM. There are a number of studies performed in the past 2 decades that look at outcomes for women who delivered with CNMs in a hospital. The C-section rate in that group ranges from 4-8%. Moreover, women who give birth in the hospital have access to pain relief, something that most women want.

There is no particular benefit to delivering in a birth center with a CNM as compared to delivering in a hospital with a CNM. There’s no decrease in C-section rate, and no savings from C-sections that were avoided. If the same group of 15,574 low-risk women had been cared for in a hospital by CNM, an additional 2934 cesarean births would NOT have been expected. Indeed, no additional C-sections would have been expected.

So while this paper makes an excellent argument for the safety of accredited birth centers that employ strict eligibility criteria, it does NOT show that birth centers reduce the C-section rate or save money by doing so.