Where is the safest place to have a baby?

Two big studies have been published in the last few weeks, both of which have confirmed previous data: home birth is not as safe as hospital birth. These studies show that having a baby at home increases the risk of your baby dying by about 4 times. That really is a big increased risk — especially considering that most home births are supposedly low-risk pregnancies. Those babies should be less likely to die.

From the January 2014 issue of the American Journal of Obstetrics and Gynecology comes a study of over 10 million babies born in 2007-2009. This study looked only at term deliveries, excluding small babies and twins and babies with congenital anomalies (that’s now the preferred term. We don’t really say “birth defects” any more.) The babies were then divided into groups by the setting of their delivery: in a hospital, in a free-standing birth center, or at home. The results are stark. The neonatal mortality among babies delivered by a midwife in a hospital was 3.1 per 10,000 births. For midwives delivering at home, the death rate was 13.2 per 10,000 (about four times the hospital risk.) There were far more babies born in the hospital than at home, but plenty of home births were analyzed, including over 48,000 by midwives. This was a large study with a reliable data set cross referenced from CDC data, and if anything it underestimates home birth mortality because babies transferred to the hospital because of complications during home birth counted as hospital babies.

The second January 2014 study came out in the Journal of Midwifery and Women’s Health. This study did not have a built-in comparison group — it collected data only from women intending to have a home delivery by midwife from 2004 to 2009. The authors looked at many outcomes, including whether the babies successfully delivered at home, Apgar scores, and their use of medical interventions. The overall intrapartum death rate was 13 out of 10,000 — and that includes only deaths during labor itself (not including babies who died shortly after birth.) Note that the death rate, 13 out of 10,000, just about matches the death rate for home midwife births from the ACOG study, which was 13.2 per 10,000. Though this study had no built-in comparison group, the rate is much higher than the hospital death rate from the ACOG study. And, again, the four-fold increased death rate is very likely an underestimate – this number does not include babies who barely survived delivery and died shortly afterwards. Also, data submission was entirely voluntary, capturing only 20-30% of home births. I’m thinking that midwives who delivered dead babies may have been somewhat less motivated to submit their data.

Though the total number of deaths was not large — the vast majority of deliveries in either setting were successful — a four-fold or more increase in death risk is not something I think most families would consider acceptable. The fact of the matter is that obstetric complications are not always predictable, and that hospitals are the place where medical interventions can be done quickly. These studies concerned deaths, but keep in mind that for every dead baby, there are many more that suffer brain damage with lasting handicaps.

Based on these and other good, large studies, a hospital birth dramatically improves the safety of delivery compared to having a baby at home. Further studies could improve the safety of home births — by developing stricter criteria to limit home births to the lowest risk pregnancies, and by making sure that home birth midwives are qualified to handle complications. But even in an optimal home birth situation, with a very competent midwife, some mothers and babies will suffer complications like massive bleeding or strokes or placental separations or umbilical cord catastrophes that will require near-instant hospital assistance to help mom and baby survive. Sometimes, there just isn’t time to wait for an ambulance.

If you want the safest choice for your delivery, choose a hospital. And then bring your healthy baby home.

Roy Benaroch is a pediatrician who blogs at The Pediatric Insider. He is also the author of Solving Health and Behavioral Problems from Birth through Preschool: A Parent’s Guide and A Guide to Getting the Best Health Care for Your Child.

Comments are moderated before they are published. Please read the comment policy.

  • JR

    Isn’t the first study the one where they looked at birth certificates only, which means all the women who had accidents (didn’t make it to the hospital in time, didn’t know they were pregnant, etc) were considered “home births”?

    Pst: Trying to scare women out of home births just won’t work. It’s the scare tactics that are used by hospitals that get women to have a home birth in the first place. More fear = more home births.

    Instead, look to start providing trauma-informed care inside the hospital. Here’s an interview with an OB on just how to do that:


    That will get women back in the hospitals. Not fear mongering.

    • http://www.pediatricinsider.com/ Roy Benaroch MD

      JR, I referenced two studies. The relevant data from both studies looked only at midwife-delivered babies at home– those could not have been “accidents”, as you say, unless there happened to accidentally be a midwife in the house. These were planned home deliveries.

      If anything, both studies understate home-birth deaths, because babies transferred to a hospital because of birth complications were counted as hospital births.

      This has nothing to do with fear. It’s sharing reliable information so families can make an informed choice.

      • JR

        The first study you linked says it is a “retrospective study using CDC data…” Where does the CDC get its data? Right. The data is considered to be inaccurate and probably biased.

        The second study mentions several different countries in it, so it seems it’s an international study, not a US based study, but I’m not familiar with it.

        Then we have another study that shows 3.5/10,000 for midwife attended home deliveres in the United States. It’s also considered biased.


        The truth is probably somewhere in between.

        Regardless – if you want to stop home births, it’s important to understand why women are choosing home births and change the system.

        I once again recommend the article about preventing trauma in patients I linked before.

        • http://www.pediatricinsider.com/ Roy Benaroch MD

          The “second study” was all data from the US, and is the exact same study you bring up and then link in your third paragraph. It almost certainly underestimates deaths, as I’ve already said, because only a small percentage of midwives submitted data, and because hospital transfers were considered hospital births. Even if you ignore those factors, their overall death rate is about four times the expected death rate for low risk deliveries in a hospital.

          The CDC data is solid and publicly available; they’ve correlated it well with other records. And the CDC-data study and the MANA-based study came to nearly identical conclusions regarding the death rate (though the authors of the MANA study spun their result differently, their numbers– the death rate– is nearly identical to the death rate among planned midwife-attended birth from the ACOG study.)

          I have not called for stopping home births. But I do think families need clear information on the increased overall risk of death.

          • JR

            I have statistics that are very different from what you have. It shows that low risk births are comparable to hospital births, and high risk women are safer in a hospital.

            For the MANA study:

            Midwife Attendened, Singleton term births:
            (Midwife is defined by specific educational criteria)

            All hospital births:

            CDC study:
            Midwifes 12.6/10,000
            (Midwife is defined by any one who reported they had a midwife – meaning I could call my sister a midwife when registering my baby’s birth.).

            All hospital births:

            Properly educated and certified midwifes are essential as they have fetal heart monitors, access to drugs like picotin, etc.

          • http://www.pediatricinsider.com/ Roy Benaroch MD

            JR, the death statistic I quoted is directly from the text of the paper, above figure 5: “The rate of intrapartum fetal death (occurring after the onset of labor, but prior to birth) was 1.30 per 1000.” (I changed the demoninator to express this as 13 per 10,000 for easier comparison with the 1st paper.) This is the overall intrapartum death rate, and it is about 4 times as high as comparable hospital births.

            You are right to highlight that there are different credentials for midwives. “CNMs” have stringent educational requirements and often work alongside obstetricians to provide care. Another credential, the CPM, requires minimal education and training– it is a unique “qualifier” that doesn’t exist in any other first-world country. But as the MANA database shows, almost all of the participating home birth midwives are CPMs, not CNMs (table 1 of the study, about 320 of the 400 midwives). One way to improve the safety of homebirth would be to insist on properly trained and experienced CNMs, but that is not what is occurring. The statistics presented are what’s happening in the US now: homebirth overall is resulting in a disproportionate number of deaths, many of which would have been prevented in a hospital.

      • JR

        So the ACOG just released this data on vaginal vs c-cection birth:

        Maternal Mortality:
        Vaginal 3.6/100,000
        C-Section 13.3/100,000

        I don’t think women being offered a C-Section are routinely told their chances of dying increases by 400% if they get a C-section?


  • Lisa

    I had my son (now 33) at home because I wanted to avoid any uncessary interventions. I think the continuing interest in home births is motivated by similar concerns on the part of women.

    The MANA study includes the the outcomes of mothers and babies who were transferred to the hospital during a planned home birth. In addition, the data included intrapartum deaths, early neonatal and late neonatal deaths. 5.2% of the deliveries were in the MANA study were by cesarean, as compared to the US national average of about 30%. Fewer episiotomies, less pitocin, and few epidurals. I think the Mana study indicates that low-risk women who are concerned about intervention in labor can safely give birth at home.

    • http://www.pediatricinsider.com/ Roy Benaroch MD

      I agree that if your primary concern is avoiding interventions, then the MANA study supports home birth. Women who delivered at home were far less likely to have the interventions Lisa listed.

      OTOH, if your primary concern is having a safe, healthy, living baby, then hospital birth is a better option.

      • Lisa

        I think you can have both a safe delivery and avoid interventions.

        Clearly, the rate of cesarean deliveries in the US is much higher than in European countries, but our infant mortality rate is higher.

        • http://www.pediatricinsider.com/ Roy Benaroch MD

          Infant mortality is a complex topic, in large part because the definition is not consistent between countries. Some places count any baby born before XX weeks a “stillbirth” (ie, not included in infant mortality), where in the US even a baby born well before viability is “alive” if any heartbeat is present. There are also valid questions about the reliability of statistics from some countries.

          In the US, our infant mortality is almost all driven by our relatively high rate of premature and multiple births. Again, a complex issue, but contributing to these are our high rate of assisted fertility and obesity and children being born to women of older age.

          More about infant mortality statistics:

          • Lisa

            You don’t give up, do you?

            I’ll read the article, but I doubt it will change my mind.

            I still inclined to think it is a reasonable choice for a low risk woman to choose a home delivery. When I had my son I was aware of the risks of a home birth; I was also aware of the benefits. I decided the benefits outweighed the risks.

        • querywoman

          Brazil has an even higher C-section rate.

  • JRM

    You can go on but if you are trying to accurately inform pregnant women of their risks then it should be made clear that having a home birth increases the chances of having your child die.

    The studies weren’t looking at maternal morbidity and mortality.

    • Rebecca Coelius

      My point is that hospitals have driven women into the arms of home birth with all of the above. I shouldn’t have to weigh a traumatic birth experience for me and the baby in the majority of cases (not to mention exceptionally costly) vs an extremely small chance that the outcomes will be poor for the baby. It’s an unnecessary choice.

      Id also be interested in the outcomes when there is readily available hospital backup vs home births an hour from a hospital.

      • http://www.pediatricinsider.com/ Roy Benaroch MD

        Rebecca, it’s regrettable that you and your baby had a traumatic experience. We ought to do better.

        But I do not accept your unreferenced assertion that the majority of hospital births are traumatic. I’ve been seeing newborns and new moms and dads professionally for 15 years. Most of them are not disappointed.

        RE: choosing a higher chance of a poor outcome at homebirth, sorry, you don’t get to decide whether that decision is “necessary” or not. All decisions have consequences. We can honestly look at the outcomes and try to figure out why these babies are dying, or we can make excuses and pretend the numbers are wrong. My goal here was to share what these two studies both show: there is a big increased risk of death among low risk babies at homebirth versus hospital birth. Moms and dads should decide for themselves what to do with that information.

        • JR

          Read the stories from women who choose to home birth. You’ll see a very strong consistent story of women who had a traumatic birth and choose to birth at home for their next child, to avoid being traumatized again.

          Then there are women who are already having traumatic reactions, and pregnancy generally brings it out.

          “… recent data suggest that nearly 1 in 12 pregnant women in the U.S. are suffering from chronic post-traumatic stress during pregnancy” http://kennedyinstitute.georgetown.edu/about/news/critical-role-of-trauma-informed-care.cfm#sthash.EC2jpDYI.dpuf

          So much care in hospitals is focused on physical well being that mental health becomes a secondary concern.

        • Rebecca Coelius

          The parents who end up traumatized are usually the ones who know the facts. Most of those new Moms and Dads are not aware that they ended up with a c section or some other intervention for no reason. For example there is very good evidence that continuous fetal monitoring increases the c section rate with no improvement in outcomes for the baby. But I saw at least six families convinced we “saved” their baby when we went to c section, even though after the baby was born it was clear it was tolerating labor just fine. Do you think we shared that with the mom recovering from major abdominal surgery, or let them assume we were baby savers?

          It IS an unnecessary choice to the extent that women in other parts of the world with excellent health outcomes do not have to make it. They have midwives, birth centers, and low intervention approaches to obstetrics. The American maternity care system can and should get its act together. Being part of that movement would be a far better use of everyone’s time than this continued war between home birth and hospital birth, as if those are the only possible models available to us.

  • querywoman

    Women who are already seriously ill are also more risky surgery subjects.

Most Popular