Neonatal Mortality Rates Part 2: Who has the best/worst infant morality rates?

This is the second part of our series on understanding neonatal mortality rates, in which we try to answer two questions: “Which is riskier, hospital or home birth?” and "How does place of birth and birth attendant impact outcomes?"  This post will specifically address mortality rates based on data from Wisconsin’s WISH initiative.*

As out-of-hospital birth advocates are quick to point out when a baby dies on their watch, babies die in hospitals, too. This is true. But does this mean that hospitals are less safe? No! One important reason is that the patient populations for home births and hospital births are vastly different. A hospital takes mothers of every kind: healthy mothers, obese mothers, drug addicts, mothers pregnant with multiples, breech babies, preemies, mothers with heart conditions, mothers with high blood pressure, and mothers with every other kind of complication you can imagine. On the other hand, a midwife practicing outside of the hospital is supposed to take on only normal, low-risk pregnancies. Thus, if the safety of the care itself is the same in both places, we would naturally expect the hospital to have higher neonatal mortality rates, given the high-risk population it serves.

You might be surprised to know that is not the case - at least not in Wisconsin.

The data collected in the WISH database can be stratified in many ways.  For our purposes, we have stratified the data by the type of birth attendant. We’ll try to understand how the specific mortality rates for physicians working in a hospital setting (MDs, in this case) compare to direct entry midwives largely responsible for home birth (DEMs, otherwise known as CPMs). We’ll also consider CNMs and their role, since they deliver in both settings.     

The data collected in Wisconsin and presented in the chart below show that the mortality rate for home births attended by DEMs was 4-5 times higher on average during 2003-2008 despite the low-risk client population they serve! 

When you consider that these data include all comers, with high-risk moms tending to choose hospitals and home birth midwives only taking on low-risk pregnancies, things look pretty bad for the DEMs. There seems to be little doubt that care provided by DEMs is much riskier.  The CNM data, however, are more difficult to assess because CNMs can deliver babies in both settings -- hospital and home.  (Note: CPMs would be listed as DEMs in this data set because they are part of the home birth sub group statistics.) 

So let’s look more closely at the role of CNMs and their mortality rates.  The chart above shows that CNMs have mortality rates on par with or even better than MDs. How is this possible when we read that out-of-hospital birth is 3-4 times more risky? Or when the ACOG (American College of Obstetrics and Gynecology) states that perinatal mortality is higher in out-of-hospital births?

To better understand this seemingly conflicting evidence, we asked Deb O’Connell, CNM and home-birth midwife, to answer a few questions: 

Safer Midwifery: Deb, ACOG states that perinatal mortality rates are higher in out-of-hospital births. Is this true? 

Deb: Not exactly. When CNMs are managing out-of-hospital births we have the LOWEST perinatal mortality and morbidity rate in the country. That is actually true regardless of our practice setting, home or hospital.

Safer Midwifery: OK, but you have also said that home birth is not as safe for a baby as being born in a hospital. So I'm a little confused. Can you please explain? I think this can be misread as thinking midwife outcomes are the same or better, when there is so much more to be explained. 

Deb: Sure. The reason that out-of-hospital birth is not as safe for baby (and this not limited to home-birth -- it includes freestanding birth centers) is due to delayed response time for intubation if required.  A CNM who is practicing in an out-of-hospital setting and who is practicing within a strict set of protocols will have roughly the same perinatal outcomes as her counterparts practicing in a hospital setting, and if they are NOT the same or BETTER than she needs to revisit her risk-out criteria.

Safer Midwifery:  Okay, I think I am starting to get it. If I understand correctly, there are really two separate but related issues. For two identical pregnant moms, the risk for out-of-hospital birth will be higher than in the hospital. But high-risk births will tend to go to the hospital, so the patient pool for midwives will be mainly low-risk births. Thus, if we see that outcomes are worse for a midwife practice, we can infer that they are doing something very wrong: either they are delivering poor care to low-risk births, or they are taking on too many high-risk births. Is that right?

Deb: Yes!

Thanks to Deb for helping us think that through! You can see now why things are so confusing. As parents, the number we really want to know when choosing between out-of-hospital and hospital birth is the first one: What's the difference in safety across the two settings for two identical, low-risk pregnant mom? The problem is that we don't see this comparison reflected in the Wisconsin dataset, or in any dataset for that matter. We only see final outcomes that reflect the two confounding effects: higher-risk practice versus lower-risk population.

It appears then that CNMs have comparable mortality rates to doctors when they function under a strict set of risking out criteria.  That being said, for two identical mothers, one giving birth at a hospital and one outside of the hospital, the risk is 3-4 times greater outside the hospital --  and even higher if your CNM is not adhering to  strict risking out criteria. Of course, with poorly trained DEMs or CPMs, the risk of out-of-hospital birth are also likely higher. 

Evaluating the efficiency and relative safety by place of birth & type of birth attendant in Michigan is nearly impossible.  You may have wondered why we used data from Wisconsin for this post.  We were not able to use data from Michigan because as a state we are lacking in our data collection efforts  and reporting outcomes as related to birth.  The type of birth attendant is not included on MI birth certificates, thereby making it impossible to adequately collect data stratified by place and care giver type.   In addition, when midwives aren't required to report their outcomes, data becomes even more vague and consequentially so does proper assessment of outcomes, measures of safety, and relative efficiency.  Michigan needs much revision when it comes to collecting data about birth in order to utilize that data to ultimately improve outcomes for mothers and babies.

*(Wisconsin's Department of Health offers a website/database which gives information about health indicators [measures of health] in Wisconsin. WISH allows policy makers, health professionals, and the public to submit questions [requests for data] and receive answers [tables] over the Internet.)