A Tale of Two Births - Comparing Hospitals to Hospitals

By Christine H. Morton, PhD

Today, Christine H. Morton, PhD, takes a moment to highlight a just released infographic and report by the California Healthcare Foundation that clearly shows the significance of birthing in a hospital that is “low performing.”  This is a great follow up post to “Practice Variation in Cesarean Rates: Not Due to Maternal Complications” that Pam Vireday wrote about last month. Where women choose to birth really matters and their choice has the potential to have profound impact on their birth outcomes.   – Sharon Muza, Science & Sensibility Community Manager.

An Internet search of “A Tale of Two Births” brings up several blog posts about disparities in experience and outcomes between one person’s hospital and subsequent birth center or home births. Sometimes the disparity is explained away by the fact that for many women, their second labor and birth is shorter and easier than their first. Or debate rages about the statistics on home birth or certified professional midwifery. Now we have a NEW Tale of Two Births to add to the mix. However, this one compares the experiences of two women, who are alike in every respect but one – the hospital where they give birth.

Screen Shot 2014-12-08 at 5.15.04 PM

The California HealthCare Foundation has created an infographic drawn from data reported on California’s healthcare public reporting website, CalQualityCare.org. In this infographic, we meet two women, Sara, and Maya who are identical in every respect – both are the same age, race, and having their first baby, which is head down, at term. However, Sara plans to have her baby at a “high-performing” hospital while Maya will give birth at a “low-performing” hospital. “High performing” is defined as three or more Superior or Above Average scores and no Average, Below Average, or Poor scores on the four maternity measures. “Low performing” is defined as three or more Below Average or Poor scores on the four maternity measures.

Based on the data from those hospitals, the infographic compares the likelihood of each woman experiencing four events: low-risk C-section, episiotomy, exclusive breastmilk before discharge, and VBAC (vaginal birth after C-section) rates (the latter one of course requires us to imagine that Sara and Maya had a prior C-section).

First-time mom Sara has a 19% chance of a C-section at her high-performing hospital, while Maya faces a 56% chance of having a C-section at her low-performing hospital. These percentages reflect the weighted average of all high- and low- performing hospitals.

Screen Shot 2014-12-08 at 5.15.22 PM

The readers of this blog will no doubt be familiar with these quality metrics and their trends over time. Two of these metrics (low risk C-section and exclusive breastmilk on discharge) are part of the Joint Commission’s Perinatal Care Measure Set. The other two – episiotomy and VBAC are important outcomes of interest to maternity care advocates and, of course, expectant mothers.

Hospitals with >1100 births annually have been required to report the five measures in the Joint Commission’s Perinatal Care Measure Set since January 2014, and these metrics will be publicly reported as of January 2015.

Childbirth educators can help expectant parents find their state’s quality measures and use this information in selecting a hospital for birth. In the event that changing providers or hospitals is not a viable option, childbirth educators can teach pregnant women what they can do to increase their chances of optimal birth outcomes by sharing the Six Healthy Practices with all students, but especially those giving birth in hospitals that are “low-performing.”

You can download the infographic in English and en Español tambien!

About Christine H. Morton

christine morton headshot

Christine H. Morton, PhD, is a medical sociologist. Her research and publications focus on women’s reproductive experiences, maternity care advocacy and maternal quality improvement. She is the founder of an online listserv for social scientists studying reproduction, ReproNetwork.org.  Since 2008, she has been at California Maternal Quality Care Collaborative at Stanford University, an organization working to improve maternal quality care and eliminate preventable maternal death and injury and associated racial disparities. She is the author, with Elayne Clift, of Birth Ambassadors: Doulas and the Re-emergence of Woman Supported Childbirth in the United States.  In October 2013, she was elected to the Lamaze International Board of Directors.  She lives in the San Francisco Bay Area with her husband, their two school age children and their two dogs.  She can be reached via her website.


A Tale of Two Births – Comparing Hospitals to Hospitals

December 9, 2014 07:00 AM by Allana Moore
I used this info graphic in my childbirth class this weekend and we discussed how rates vary from one facility to the next. One couple was deciding between two hospitals and found the information valuable in helping them decide.

A Tale of Two Births – Comparing Hospitals to Hospitals

December 9, 2014 07:00 AM by Scientist Mom
There's a critical logical fallacy here, which is that if you go to a hospital with a 54% c-section rate, you have a 54% chance of c-section. By that logic, I should try to avoid going grey by moving to a town with a lot of young people. Some women walk in the hospital door to a pre-scheduled c-section, they have a 100% chance of c-section. A VERY low risk woman, one who has previously given birth with no problems, no complications in current pregnancy, spontaneous term labor, has only a 2-3% chance of c-section, and that's not going to vary much no matter where she is.

A Tale of Two Births – Comparing Hospitals to Hospitals

December 10, 2014 07:00 AM by Katie Lorand, CNM
@Scientist Mom You misread the information about c-sections. They are specifically collecting information regarding c-sections in low-risk moms, removing planned c-section and high-risk moms from this data point. The issue is that in LOW RISK labors, some hospitals still have unacceptably high c-section rates without good medical justification.

A Tale of Two Births – Comparing Hospitals to Hospitals

December 10, 2014 07:00 AM by Koa
@Scientist Mom You don't think that differences in hospital policies and culture, differences between doctors and midwives, etc. contribute at all? Your example of moving to an area with lots of young people to avoid going grey is ridiculous (as you intended) because there's no plausible causal connection. Where an environmental factor plays a causal role, moving to a different environment may make a difference.

A Tale of Two Births – Comparing Hospitals to Hospitals

December 10, 2014 07:00 AM by Christine H Morton, PhD
Thanks to both commenters. I'd love to hear from more CBEs like Allana, who use this information in their classes. For Scientist Mom, please note that these data reflect rates of C-section among women who are nulliparous (first birth), carrying a single baby who is head down at term (also known as NTSV C-Section). The NTSV C-section rate among this population varies tremendously in California hospitals. I should have mentioned that the website, http://www.chcf.org/publications/2014/11/tale-two-births, also maps the high- and low-performing hospitals on which these likelihood statistics are based. For example, if you click on the Greater El Monte Community Hospital, you'll see their rate of this "NTSV C-Section" is 43.6%, way above the 27.8% state average. In contrast, among the high-performing hospitals, the NTSV CS rate for Ventura County is 22%. You're right that if a woman has had a prior vaginal birth, her chance of C-section in subsequent births is quite low, which is why focusing on rates of the NTSV C-section is so critical. We want to minimize the number of C-sections to these first time birthing women, and the problem with comparing overall C-section rate is that it includes women with scheduled or medically necessary C-sections, regardless of parity. That is one reason the NTSV C-section measure was created - so that hospitals could be compared on a metric involving the same group of women - low risk women having their first (single, head-down) baby at term, with no confounding medical diagnoses. Thanks!

A Tale of Two Births – Comparing Hospitals to Hospitals

December 11, 2014 07:00 AM by Scientist Mom
As I said, those are just some of the risk factors. Age, overall health status, and quality of prenatal care matter, too, and those can vary considerably from one hospital to the next, even when those hospitals are close together. For example, a 40-year-old woman having her first baby is three times more likely to need a c-section than a 16-year-old! NTSV is NOT sufficient to ensure you are comparing apples to apples. Hospital policies may play a role, but you're putting a hospital in East LA up against one in Ventura County. I suspect there might be some variation in patient population, there. Also note, Greater El Monte does only 254 births a year, which is quite low. The infographic in this post literally claims that the c-section rate of your hospital is the probability that you will experience a c-section, and that's at best a serious exaggeration.

A Tale of Two Births – Comparing Hospitals to Hospitals

December 12, 2014 07:00 AM by Bruce Spurlock
Hello, I am the executive director of CHART and wanted to respond to the dialogue above about population vs. individual likelihood. First, I'm thrilled to see this dialogue occurring because it will drive improved care, which is one of our main objectives. Thank you. It is true, as with any study using population based statistical analysis that the results of the population might not apply to the individual. This is true for chemotherapy, antibiotics, surgery, mammograms, etc. The only way to evaluate for an individual is with a N=1 study but these are not generalizable to the rest of the population. The unit of analysis for our work was the hospital, which meant the population of woman with NTSV deliveries. This is NOT further risk-adjusted by patient characteristics, some of which have a relationship to C-section rates (although causality of these may not be linked). So it is possible that other unmeasured patient factors explains some of the variation in hospital performance. We also suspect many hospital/physician characteristics/practices that would also explain the variation we see in the data. But the limited research, and certainly our analysis, does not include them. What is safe to say based on our long experience with reporting hospital performance in a number of areas, the level of variation in maternity care performance is very, very large and much larger than we've ever published before. Although we can't prove it with data, since we don't collect it, we strongly suspect that hospital/physician factors explain such a large amount of the variation that there is significant opportunity for improvement. That means consistently high performing hospitals likely really are different than consistently low performing hospitals. And this should be a call to action to improve maternity care in California. Something the folks at CMQCC are dedicated and committed to accomplishing. It is also important to note that patient characteristics, hospital/physician characteristics do not describe all of the events of the delivery process, many of which are unpredictable and would strongly indicate the need for an appropriate c-section such as deep, prolonged stress on the baby, so that under any circumstance, we would never advocate for 0% c-sections and would love to publish "balancing measures" so that no delivery/baby has an untoward outcome because of the desire to publish and reduce avoidable c-sections. With respect to the language we chose to indicate a mother's likelihood of having a procedure, we had many reviewers looking at the document from a communication standpoint, technical statistical standpoint, graphical display perspective and others. If we would have used the technically more correct language "in a population of woman with NTSV deliveries and no other risk-factors measured, XX% had this procedure/outcome" would be more accurate but less accessible for the average reader. We often sacrifice this level of verbal, technical precision to be able to convey ideas to a broad group of diverse audiences. The likelihood that any one patient will have any one of the procedures/outcomes after the fact is either 1 or 0. Moreover, women that are armed with this data and use it to influence their care are applying the Hawthorne effect which influences the results/measures simply by studying it. But to share these nuances in the infographic misses the major point - we desperately need to reduce the variation in maternity performance in CA, we have some star hospitals who can lead the way and the moms and babies of CA deserve it.

To leave a comment, click on the Comment icon on the left side of the screen.  You must login to submit a comment.  

Recent Stories
LamazeLIVE! 2018 - An Event Like None Other

Looking for New Teaching Activities - Check Out These Four Websites for Ideas

How Low Can You Go? An Educator Activity Challenge