Maternity's Thin Line
What separates mothers who die from those who survive?
The handwritten notes in the hospital logbook are spare, but they tell the story of how a woman named Amina Seidu nearly died giving birth at one of the best hospitals in Kumasi, Ghana.
Seidu (whose name has been changed to protect her privacy) lives in a dusty settlement of cinderblock family compounds clustered beside a green-roofed mosque on the periphery of Ghana’s second-largest city. In one of those compounds, she shares two rooms with her husband and their daughters, 8 and 4.
On a late summer afternoon, she serves visitors water from a hand-cranked well and tells how her fourth pregnancy nearly killed her.
Labor pains woke her before dawn one morning last May, says Seidu, a 29-year-old woman with broad cheekbones and a turquoise floral headscarf. For this baby, she and her husband had decided that she would go straight to a referral hospital in Kumasi. The long taxi ride would be expensive for a couple that makes a living selling machetes and groundnut paste in open-air markets. But their third child had been born dead after a 15-hour labor in a local clinic, and this time they wanted expert care. Amina Seidu had gone to seven prenatal checkups and expected no problems.
When she and her mother-in-law arrived at the hospital at 6 a.m., the nurses told Seidu that her baby would take some time to arrive. By noon she had begun to gush so much blood that it pooled on the floor beside the bed. The nurse-midwives brought her pads. She soaked pad after pad. “I can’t count how many,” says Seidu, who remembers calling for the nurses many times. “They thought I was a complainer.” Seidu was hemorrhaging.
Eventually, she says, a doctor passed by, saw the blood and rushed her to the ultrasound room. She remembers nothing further until she woke to find that her baby had been delivered by cesarean section and that the baby was dead. Her husband and his father buried the child, a boy. She never got to see him.
“God gives and God takes away,” she says.
If Seidu had followed her son to the grave—and she came close—she would have been among the hundreds of thousands of women whose deaths cut a red swath across the world map in a recent Lancet assessment of maternal mortality. UN Millennium Development Goal 5 calls for a 75 percent reduction in maternal deaths between 1990 and 2015. The red on the Lancet map indicates countries that, at current rates, won’t reach that goal by 2015, or by 2025, but rather after 2040. Red covers most of Latin America and the Arabian Peninsula, much of Central and Southeast Asia, and nearly everywhere in Africa south of the Sahara. More than half the women who die from pregnancy-related complications worldwide are Africans, according to WHO.
In narrowly avoiding death, Seidu became what researchers call a “near miss”—a woman who nearly dies because of complications of pregnancy or birth. While the study of maternal death has long been a public health priority, researchers are increasingly studying near misses. They hope this research will help forestall severe complications, improve care when complications do occur and increase their understanding of what leads to deaths.
“Maternal mortality is ‘the tip of the iceberg,’” says Michelle Hindin, PhD ’98, MHS ’90, an associate professor in Population, Family and Reproductive Health (PFRH) at the Bloomberg School. “Women who nearly die but survive are much more common, and their needs are not being addressed.”
Near misses almost always occur in clinics and hospitals; medical interventions are what prevent crises from becoming deaths. However, the health care facility is a “black box,” says Özge Tunçalp, MD, PhD ’12, who studied near-miss cases globally and in Ghana for her doctoral work with Hindin.
Near-miss researchers look inside that box both to suggest improvements as well as learn about maternal mortality. The circumstances and events that lead to a near miss resemble those that end in a death, says Tunçalp, who now works for WHO in Switzerland. A woman in a life-threatening condition will become either a maternal near miss or a maternal death; the distance between the two is “a thin line,” says Tunçalp.
Numbers for near-miss cases also can serve as “proxy indicators” for maternal mortality, which is likely underestimated because one in three women in developing countries gives birth without medically trained attendants, and deaths at home often go unreported. Because near misses occur in clinics and hospitals, however, researchers capture them all.
Although interest in near-miss research dates back to the 1990s, it wasn’t until 2009 that WHO defined a near-miss case as “a woman who nearly died but survived a complication that occurred during pregnancy, childbirth or within 42 days of termination of pregnancy.” Linda A. Bartlett, MD, MHSc, an associate scientist in International Health, served on the WHO working group that defined near miss. She also helped develop the WHO data collection form for near misses. And Tunçalp was one of the first scientists to use the tool in sub-Saharan Africa.
Tunçalp did her research at Korle Bu Teaching Hospital in Accra, the capital of Ghana. The relatively prosperous West African country reduced its maternal mortality ratio per 100,000 live births from 550 deaths to 350 deaths from 2000 to 2010. Still, Ghanaian women face a risk of maternal death 22 times that of women in developed countries.
“Even though the numbers are going down, we are still losing a number of women,” says Tunçalp’s Ghanaian colleague, Kwame Adu-Bonsaffoh, MD, a senior ob-gyn at Korle Bu. He, Tunçalp and colleagues prospectively analyzed medical records for 3,438 pregnant women who passed through Korle Bu during five consecutive months beginning in fall 2010. They found that 94 of those women were near misses (while another 37 died). Of the 94 women who nearly died, 29 suffered from severe pre-eclampsia (pregnancy-induced hypertension coupled with protein in the urine). All but one of the women who developed eclampsia received the recommended treatment, magnesium sulfate.
Care was not as good for women with life-threatening bleeding or infections. Of the 121 women with severe postpartum hemorrhage, Tunçalp found that only 67 percent got medications like oxytocin that can slow or stop bleeding. She also found that among 32 women with sepsis, a life-threatening response to infection, only 40 percent got intravenous antibiotics. Fewer than half survived. “Even the doctors at the facility didn’t realize that there was such a big issue with sepsis,” says Tunçalp. “If you look at cases one by one and then don’t look at the aggregate data and the trends, it’s really easy to miss even if it’s happening in front of you.”
When women endure a near miss, Hindin says, “it impacts their whole life: their family life and their ability to work. It may impact their future childbearing. Women who experience near misses often have serious complications that have long-lasting physical and mental health outcomes. These women need health care services, and monitoring near-miss rates will help governments and ministries of health plan better.”
Qualitative studies can help with this, particularly because women who survive life-threatening complications live to tell the tale. Tunçalp and Adu-Bonsaffoh interviewed 32 of the women from the larger study. “At first they are very happy and very grateful because they are alive,” recalls Tunçalp. “But then you probe them and ask them how they were treated, and all the staff-related issues or physical environment issues come out.”
The women’s testimony pointed out both interpersonal and structural problems not evident from morbidity and mortality data. Some women complained about disrespect from the hospital staff. A 36-year-old woman who nearly died described being ignored by the midwives: “They don’t attend, even when you are crying and the baby’s head is coming. They’d be sitting down and chatting and laughing at you.”
A 34-year-old woman who lost her baby and whose uterus ruptured after a 24-hour labor didn’t find out until 12 days later that she’d had a hysterectomy. As a physician told her, she began to cry, fearing her boyfriend would leave her because she could no longer have children.
Several women objected to recovering from surgery while lying on a mattress on the floor. (Patients often outnumber beds and end up on the floor.) One woman was incredulous that the ward had no water. (While Tunçalp was at Korle Bu, the water in the maternity block stopped working for several days.)
Tunçalp’s next step is to return to Ghana and share her results with the managers of the ob-gyn department at Korle Bu. She will recommend that they audit near-miss cases to assess the quality of care and solve problems. Whereas most U.S. hospitals schedule frequent morbidity and mortality conferences, the only systematic assessment of maternal care at Korle Bu is a monthly review of deaths. Discussions of near misses have the advantage of being less fraught than discussions of deaths, says Tunçalp. “It’s easier to talk about what went wrong when you know at the end of it the woman actually survives,” she says.
Tunçalp will also recommend that the hospital work to improve the system for referring and transporting women to Korle Bu from other facilities. Ghanaian public health researcher Easmon Otupiri, PhD, MPH, MSc, agrees that delays in referrals constitute the biggest barrier to good emergency obstetric care in Ghana.
Several roadblocks (besides the rough roads themselves) stymie efforts to move a woman with a complication from a local health post to a referral hospital with specialists and a blood bank. The most obvious problem is a lack of ambulances staffed by skilled technicians, says Otupiri, a professor of community health at the medical school at Kwame Nkrumah University of Science and Technology in Kumasi (and a former Gates Visiting Scholar at the Bloomberg School). Another challenge: communication. When a patient with complications is moved to a higher-level hospital, doctors and nurses must use their personal phones to call an acquaintance at the destination hospital. If the colleague’s phone is out of credit, the call doesn’t go through.
And women facing life-threatening emergencies sometimes wait on gurneys in the hallway while less-urgent cases enter the OR, Otupiri says, adding “we don’t seem to have a well-organized system of who goes in when.”
Complicating triage is “a mal-distribution of cases,” says Roderick Larsen-Reindorf, MD, an ob-gyn at Kumasi’s main referral hospital, Komfo Anokye Teaching Hospital. “We have a huge mix of normal labors that a midwife could have handled in an outside hospital, combined with a huge number of complicated cases. Our midwives can be overwhelmed with numbers,” says Larsen-Reindorf, adding that five midwives sometimes must manage 25 women in labor.
PFRH associate professor Cynthia Stanton, PhD ’96, MPH, points out that the “brain drain” contributes to a shortage of midwives. “Particularly in Anglophone countries [which include Ghana], the U.S. and the U.K. have been poaching nurses for many years,” says Stanton. “It’s really a crisis in many countries.”
Better family planning could reduce maternal mortality. As Larsen-Reindorf puts it, “If a woman doesn’t get pregnant, she doesn’t die of it.” Tunçalp reports that although most women she met wanted contraceptives, many went home without them. That’s what happened to Amina Seidu in Kumasi. Not only did she leave the hospital without contraception, but she was told that she could try to conceive again in six months.
Seidu’s situation disturbs Amy Ong Tsui, PhD, MA, director of the Bill & Melinda Gates Institute of Population and Reproductive Health, on two counts. First of all, “she should have left with some kind of contraception, that’s for sure,” says Tsui, a PFRH professor. The story does not surprise her, however; only 23 percent of Ghanaian women use modern contraceptives, including condoms.
Second, Tsui would recommend that Seidu wait four times longer before conceiving again: A woman can improve her odds of a healthy pregnancy if she waits at least 27 months between pregnancies. If a new pregnancy follows too quickly, says Tsui, the mother may be nutritionally compromised and therefore more vulnerable to infections and to anemia.
Tsui says that in much of sub-Saharan Africa, couples once abstained from sex for several months after a birth. “You weren’t supposed to resume sex until the baby reached a certain milestone of growth, usually when it could sit up or walk. Sitting up is six months, walking is 12 months.” Nowadays, she says, fewer Africans follow the custom of postpartum abstinence.
What Tunçalp has initiated in Ghana is what WHO advocates worldwide: that clinics and hospitals track and analyze maternal near misses, determine what might prevent them, make changes and then repeat the cycle.
Tunçalp says that the near-miss approach is gaining traction. “More and more articles are being published in journals, and the majority of these are coming from low- and middle-income countries. At the most recent [international ob-gyn] conference I attended, there were many oral presentations and sessions on near-miss morbidity and near-miss audits.” Paradoxically, she says, “when you start improving maternal health, you might actually start getting more near misses because you are preventing deaths.”
Meanwhile, the nurse-midwives at Komfo Anokye hospital in Kumasi keep patient records by hand, entered into the bound logbooks. They record vital signs in the left column and progress notes on the right. In each book, more than once, a midwife has drawn a cross in red ink at the end of a patient narrative, with a woman’s name and this inscription: “May her soul rest in perfect peace.”