Saving Mothers and Newborns
Why are tens of thousands of mothers and newborns still dying every year from preventable causes related to pregnancy and childbirth? Recent research finds reason for hope in a comprehensive, systems-based approach that focuses on transport, communications, and making sure the right resources are in the right place at the right time.
Editors’ note: Listen to the full version of Atul Gawande and Katherine Semrau’s conversation in a special edition of PS Editors’ Podcast. Tune in to all episodes from your favorite podcast app, and subscribe via Apple Podcasts, SoundCloud, or RSS Feed.
Gawande: I was recently lucky enough to be a part of one of the largest studies of maternal and newborn health ever conducted. You led it. In a multiyear trial in Uttar Pradesh, India, you found that even with vastly improved care during the 48 hours around childbirth – when women and newborns face the greatest risk of death and complications – maternal and perinatal mortality did not fall. Given that 99% of maternal deaths from childbirth are considered avoidable, why are so many women still dying?
Semrau: The big killers of moms and babies around the time of childbirth tend to be the same worldwide, whether you are in the US or India. For women, the leading causes are hemorrhages, hypertension (also known as eclampsia), and sepsis (infection). And for newborns, they are asphyxia or breathing difficulties within the first few minutes of delivery, prematurity, and also sepsis.
AG: When you think about it, childbirth is really kind of a brutal thing. Let’s take the example of breathing difficulties, which affect 10% of newborns. How do we address it?
Subscribe to Project Syndicate
Enjoy unlimited access to the ideas and opinions of the world's leading thinkers, including weekly long reads, book reviews, topical collections, and interviews; The Year Ahead annual print magazine; the complete PS archive; and more. All for less than $9 a month.
KS: At the bedside, the birth attendant first has to recognize that the child is having difficulty breathing; then they have to ensure that the airway is clear. If it’s not, they have to use a neonatal bag and mask to help that baby start breathing.
So, a full chain of events has to happen. It starts with recognition by the staff that there’s a problem. The supplies and equipment must be available at the bedside, not somewhere else in the clinic or the facility; and the staff must have the right skills and capabilities to perform the appropriate intervention: newborn resuscitation.
AG: It seems like it should be simple to recognize that a baby is struggling to breath, but it actually isn’t always totally obvious. The very first thing a bedside attendant has to do is dry the baby vigorously with a clean cloth, which gets 90% of them going. And yet sometimes the sterile towel isn’t there. In the study, how did you ensure that all those things started to happen more commonly in these facilities?
KS: Our main focus was to instill a systems-based approach using the World Health Organization’s Safe Childbirth Checklist, which itemizes 28 evidence-based actions that should be taken for every mom and baby during labor and delivery. Then we paired that checklist with a peer-coaching model: nurses coached nurses, and physicians coached physicians, to improve bedside care. The point was to strengthen communication in the clinic to ensure that the necessary supplies and equipment were available, that the staff was trained, prepared, and adhered to best practices.
AG: Normally health-care practitioners would be told, “Hey, go get training, go take a course to learn to do this.” So, why did you decide to start at the bedside?
KS: We wanted to move the training process from those external locations to the room where nurses and birth attendants are actually delivering babies. That is the only way for trainees to see what they will have to deal with on a routine basis. Also, it’s important to remember that this isn’t just about individual birth attendants. Tackling this problem requires a much larger system of people communicating with one another, not just in facilities but also at the district and state level.
Lessons from India
AG: Let’s talk about the hard data from the study. How big was this trial, and how much did the care change?
KS: Over a three-year period, we worked in 120 facilities – 60 of which got the intervention, and 60 of which were control (standard of care) sites – and we followed 160,000 mother-infant pairs to find out whether we had made a difference.
All told, the quality of care changed dramatically. For example, in the control sites that did not receive the checklist or coaching, only 25% of mothers received the right medication to prevent hemorrhage, whereas in the intervention sites, 80% did.
Still, when we looked at the mortality rates for mothers and babies, we did not see a difference between the intervention and control groups, which suggests that we didn’t solve all the underlying problems. Even with improvements in the quality of care, there were still gaps in supplies and equipment, and the staff didn’t always have the skills needed to overcome challenges. And there wasn’t enough communication between frontline health facilities and the higher-level facilities that are equipped to manage complications.
AG: It seems like context matters a lot here. Give us a picture of childbirth in Uttar Pradesh, where, during the years of the study, 5% of babies were either stillborn or died in delivery or within the first week. What do the health-care facilities there look like?
KS: Typically, what we’re talking about are concrete-block buildings with 3-5 rooms, only one of which is dedicated to labor and delivery. There are usually one or two metal beds, which are raised off the ground to help the birth attendant catch the baby. There are often no curtains between the delivery beds, and there may not even be constant electricity, heat, or running water.
Also, it is not uncommon to have several women delivering at once. When women come in, they are often quite advanced in labor, which leaves little time to manage a complication. And after delivery, women usually stay for less than a day, because the facilities are overcrowded. There may not even be a recovery room available. Unfortunately, the first 24 hours is also the highest-risk period for the baby, so that is another missed opportunity to catch complications.
I should also mention that respectful maternity care is lacking in many parts of the world. Recent research shows that up to a third or a quarter of women are disrespected or abused. They may be hit or slapped, or even have an episiotomy performed on them without consent.
That said, most birth attendants I have seen really do want good outcomes for moms and babies. They are working with what they have, which often isn’t enough.
AG: Having visited these places with you during the study, I’ve come to think of them as primary-care clinics, not hospitals. The doctor is busy seeing people around the clock for all primary-health-care issues, and the nurse is typically the one who performs the delivery. So, give us a picture of the nurses. How experienced are they, and what do they do if the woman needs a C-section?
KS: In the trial, the birth attendants were typically staff nurses with 2-3 years of classroom training, who were then trained at the facility to be a general nurse. Around half had also received what’s known as Skilled Birth Attendant training, which teaches the most appropriate ways to provide labor and delivery care.
Of the staff nurses working in labor and delivery rooms, we found that most had around nine years of experience, on average, but hadn’t received additional training in about four years. As for performing C-sections, the one physician at the facility may or may not have that capability. Generally, C-sections cannot be performed at these frontline facilities, because there is no operating theater, so women have to be referred to a higher-level facility.
AG: We were both part of a separate study that found that about 19% of all deliveries should be C-sections. In other words, when C-sections are performed at that rate, mortality declines, because, without an intervention, babies who are stuck can suffer from asphyxiation and put the mother at risk of hemorrhage or an obstruction leading to sepsis. In your trial, how many of the 160,000 mothers couldn’t deliver at the facility and had to move to another one?
KS: Our C-section rate was 2%, which was much lower than what we had expected. One inference is that women who may have been laboring for quite some time were not being referred to a higher-level facility for a C-section.
The System, Not the Service
AG: When the study was published in the New England Journal of Medicine this past December, an accompanying editorial praised its scientific rigor, and concluded that these frontline facilities simply cannot be improved enough to get high-quality care. Women need to be delivering in hospitals where they can have a C-section, if needed, and where they can receive a blood transfusion in the case of a severe hemorrhage, the likelihood of which is about 3%. What do you think – is that right?
KS: It’s not that rural primary-care facilities can never provide labor and delivery care. What really matters is the functionality of the system as a whole. As we know from other global studies, health facilities that deliver fewer than 500 babies per year probably should not be providing labor and delivery services, because the staff will not keep up the skills needed to manage complications, which affect about 15% of women.
So, it’s really about preparing that frontline facility to handle complications while also connecting it to a higher-level facility that can provide a C-section quickly. I don’t think we’re ever going to get to a point where we can simply build hospitals everywhere. We need to focus on the transportation and patient-referral systems between lower-level facilities and hospitals.
AG: All right, I want to dig into this, because it raises a really interesting fundamental question. My parents are both from India, and my father is from a rural village like the ones you’re trying to help with the checklist-and-coaching program. My parents also both became doctors, and they settled down in the Appalachian foothills, in Athens County, Ohio, the poorest county in the state.
Growing up there in the 1970s, I remember that our first obstetrician arrived in 1968, and that we did not have a formally trained anesthesiologist until 1983. If the obstetrician was away, you had to rely on transportation to the next town, which was 45 miles away. Needless to say, this was a rural community with fewer than 500 deliveries per year. It wasn’t that different from the village my father had come from.
What’s striking to me is that in the 1920s, the United States’ per capita GDP was lower than that of China today, but we still said, “We have to build that hospital.” And in 1946, the US Congress passed the Hill-Burton Act, which funded the construction of hospitals across the country. In fact, that is part of the reason why my parents ended up where they did. The government was paying for greater capacity everywhere.
If that is what it takes to save moms and babies, then don’t you have to go in that direction?
KS: I think giving all women immediate access to tertiary-care facilities that can perform C-sections is right in the long term. But we also need a process to make labor and delivery safer now – not in ten, 15, or 20 years. And it needs to work in places that may not have the economic resources to build hospitals rapidly. That means ensuring that nurses have the right level of training, emphasizing the importance of midwifery for labor and delivery, and making necessary supplies and equipment available at the bedside.
AG: One thing from the study that really struck me was the variance in newborn mortality rates between facilities. Maybe we can start by getting everyone down to the same level as the lowest-mortality facilities.
KS: Exactly. Our data show that perinatal mortality reached as high as 104 deaths per 1,000 births at one facility, which means that 10% of babies were stillborn or dying before seven days post-partum. At the same time, another facility had just 14 deaths per 1,000 births, which is much closer to a middle-income country’s rate of perinatal mortality. Clearly, there is an opportunity to make progress in the short term by looking at what the low-mortality facilities were doing differently.
AG: Broadening the scope, I want to talk about what we saw in Namibia, where we partnered with a district hospital that adopted the checklist-and-coaching model. They started with 70% adherence to basic practices, and then what happened?
KS: Over a two-year period, they had achieved 90% adherence. But this was in a hospital setting. They had midwives providing labor and delivery care, and also a leader who was committed to improving quality. His staff had a skills lab, and they conducted maternal-death audits to provide feedback at all levels of the system.
AG: It sounds like there are some building blocks there for others to use. I was also struck by the fact that they reported not just a 50% reduction in stillbirths, but also a decline in maternal deaths, from three to four per year to none within two years. It’s astonishing.
KS: It really is quite a success story, and it provides a good model for us to think about.
AG: Since the study in Uttar Pradesh, we’ve seen people across 30 different countries saying they want to roll out the checklist-and-coaching approach and test it for themselves. What does that tell you?
KS: Globally, the maternal- and newborn-health community is looking for more than one-off interventions. The WHO Safe Childbirth Checklist offers guidance on key practices as well as a framework for a systems-based approach.
In those 30 countries, the checklist is already being used in myriad ways, sometimes as a training tool, and sometimes at the bedside during labor and delivery. It is also helping to identify gaps in health systems, such as insufficient supplies or skills challenges people may be facing. I’m excited to see all the different implementation strategies.
AG: And it’s being used not just in developing countries such as Indonesia, Mexico, India, and in parts of Africa, but also in Italy and the United States, which has the highest maternal mortality rate among major developed countries, according to the OECD. Specifically, some US facilities are adopting the checklist to tackle two of the biggest challenges: managing blood pressure and hypertension. It will be interesting to see if it brings about a reduction in mortality in these settings.
Let’s end with a wish-list question. If you had unlimited funds to reduce maternal and perinatal mortality right now, how would you spend it? Where does it need to go?
KS: Bearing in mind the long-term goal of providing universal access to well-equipped, well-staffed facilities, I would start by ensuring that there are enough providers with the proper skills available at existing facilities to meet current needs.
With that, labor and delivery staff could manage complications and refer women to higher-level facilities when C-sections are required. It’s important to remember that this is not just about sending attendants to training. Reducing mortality requires continuous medical education over time. And I think that fact needs to be emphasized globally.
Second, we could do more to connect frontline and higher-level facilities through better transportation and communication. And, finally, I think we really need to focus on community relationships. We have pushed women to deliver in facilities, yet those facilities are not always delivering on the quality of care that we have promised. We know that women vote with their feet. They know where the high quality is. They know whether they have a trusted relationship with providers.
We should understand more about what women expect when they go to a facility, and to ensure that they know their rights. That starts with building relationships of trust and respect within communities.