The Doctor Who Revolutionized Hospital-Birth Safety

The newborn’s head arrived first. A shoulder should have followed, but it was lodged securely behind the mother’s pubic bone. The baby’s head made the “turtle sign” as it delivered, rocking back as if to retreat into his mother’s vagina.

“Shoulder dystocia!” the obstetrician called out.

The obstetric team at Boston’s Beth Israel Deaconess Medical Center jumped into action. Within seconds, additional nurses were in the room; in less than a minute, other physicians were standing by, too. Two nurses flexed the mother’s legs gently backward. Simultaneously, another nurse positioned a stool next to the labor bed, preparing to push the baby’s shoulder to the side to dislodge it.

“Two minutes,” one of the nurses called out, a reminder of how long the team had been at work. Brain injury, due to a lack of oxygen, can occur after approximately five.

Twenty seconds later, it was over. The baby slipped out, the umbilical cord was cut, and the infant was handed to two pediatricians. The infant was soon in his mother’s arms.

Complications like shoulder dystocia are rare in obstetrics. Even when they occur, mothers and babies are overwhelmingly likely to do well. This sounds like a blessing, but it actually has long been the field’s most dangerous curse. Obstetric medicine is like aviation: As even the worst airlines go years without a crash, so can even the most cavalier, careless hospitals go years without a mother or a baby dying or being serious injured. But these rare catastrophes add up: In 1990, the maternal mortality rate in the United States was just over one in 10,000—on par with many Eastern European countries, but about double the rate in the “safest” countries, such as Canada.

Largely because of this statistical problem, obstetrics joined the medical safety movement later than most other specialties. Procedural changes and training initiatives designed to reduce rare clinical errors are unlikely to lead to statistical improvements, so subscribing to principles of safety has often required a leap of faith.

If obstetrics has caught up in the past decade—and there is evidence it has—it’s due significantly to the work of a doctor who was willing to make the leap.

Physicians, by training, demand evidence that a new treatment is better before they are willing to abandon the old way. However, when “better” is difficult to measure, the evidence-based standard may deter progress, despite what common sense suggests is effective. A clinician who comes up with a new way to treat pneumonia may reasonably expect to be able to measure its success within a year or two. In obstetrics, it may take a decade to see the impact of an intervention designed to reduce a rare event like maternal death.

Steve Clark, an outdoors-loving doctor from Wisconsin, trained in obstetrics and gynecology at the University of Southern California in the late 1970s and early 1980s, an era of economically segregated medicine. Poor and uninsured pregnant women, many of whom were immigrants from Mexico and Central America, were shunted to USC, where Clark and his colleagues learned to care for patients with malaria, typhoid, and occasionally bubonic plague, as well as a host of other medical conditions rarely seen in the United States.

The trial-by-fire nature of this experience is widely recognized as unethical today, because poor patients received most of their hands-on care from unsupervised trainees. It did, however, turn Clark into an exceptionally competent obstetrician. He soon rose through a series of leadership roles at Intermountain Health System in Utah. There, he met Brent James, a leader of the patient-safety movement who was building a national reputation by cajoling colleagues to take care of patients with the same condition in the very same way—even if scientific evidence for the benefits of standardized care was lacking.

Clark took this strategy and applied it to obstetrics. He got his colleagues working on preventing those rare complications that were stubbornly persistent because they fell into the domain of statistical ambiguity. Early elective deliveries, for example, where obstetricians chose to induce labor or operate for convenience, instead of medical necessity, caused a nearly imperceptible increased risk in newborn complications. It was unclear, before Clark and others addressed this problem, that preventing unnecessary elective deliveries would improve newborn outcomes.

Where science left off, Clark applied his own common sense, derived from his prodigious experience, and then insisted that his colleagues follow the protocol.

Aaron Caughey, the chair of the department of obstetrics and gynecology at Oregon Health Sciences University, and a leading researcher on obstetric practice, recalls how Clark standardized the use of oxytocin.  The medication is commonly used during labor to help the uterus contract, but it can be dangerous if used at too high a dose. “He bothered to write the protocol down,” Caughey says, “and then he had the power to make everyone follow it.”

By the late ’90s, Clark’s tactics and results were getting noticed in the obstetric community, but what really accelerated adoption was the 1999 Institute of Medicine report that suggested as many as 98,000 Americans die each year due to preventable hospital errors. In 2004, the Hospital Corporation of America asked Clark to lead its obstetric safety program for all 115 of its obstetrics programs, representing approximately 220,000 deliveries annually, or approximately 5 percent of the deliveries in the United States. This gave Clark authority over enough deliveries that he could begin measuring the results of his policies, and over the subsequent years, the scientific validation of this leap of faith began to emerge.

Each month, HCA would send him data on its birth units, and Clark would pore over them looking for hospitals that were outliers. Then he would get on a plane and review the hospital charts for a sample of patients with a condition that had drawn his attention. “One third of the time, the problem was poor-quality data; one third of the time there was a perfectly good explanation for the problem; and one third of the time there was something that needed to be fixed,” says Clark, who is now on the faculty of Baylor College of Medicine in Houston.

Fixing problems involved several key safety principles that had become the playbook at Intermountain and elsewhere. First, quantify the problem and allow doctors and nurses to review their own data themselves. Second, involve local influencers—such as the department chair at the local hospital—in the process. Third, design the best intervention possible, without requiring that it be perfect, and then iterate when problems arise. Lastly, allow doctors to opt out in select cases, so they don’t feel like a safety program is being crammed down their throats.

For instance, it was well known by 2006 that the primary killer of pregnant women in the U.S. was pulmonary embolism, a blood clot that typically travels from the legs to the lungs. Yet women who were having a cesarean rarely wore the inflatable cuffs on their legs that most patients wear during surgery to prevent clots from forming. Clark had no published data demonstrating that the cuffs specifically prevented clots in women undergoing cesarean—clots are rare enough that such a study would be nearly impossible to conduct due to statistical limitations—but he thought it would be a reasonable approach. In the seven previous years, HCA had lost seven mothers to clots; the six years that followed his recommendation, HCA only lost one.

In the decade beginning in 2004, Clark’s approach led the standardization of a handful of other obstetric practices, including protocols for common, but occasionally dangerous, medications. He and his colleagues helped HCA cut its maternal death rate to 6.4 per 100,000 live births—half the national average—and reduce the number of lawsuits and the amount of litigation-related damages paid by HCA and HCA physicians. The HCA approach to clot prevention was rapidly adopted by national obstetrical organizations, and has now become the standard of care in the U.S.

“The work Steve did at HCA directly impacted an incredible number of women as well as their babies by reducing the number of complications,” says Gary Hankins, the chair of the department of obstetrics and gynecology at the University of Texas Medical Branch in Galveston.

The maternal mortality rate in the United States remains higher than most other wealthy countries, but examples like HCA demonstrate that the techniques exist to reduce preventable rare catastrophes.

Like the cuffs and the oxytocin checklist, the shoulder dystocia protocol used by the Boston obstetricians was implemented—in fact, became required—before scientific evidence demonstrated it helped at all. What ample evidence in multiple other medical specialties did show was that consistent practice leads to better outcomes. Years after Steve Clark insisted on consistency at Intermountain and HCA, research caught up with the hunch: Consistency helps mothers and babies, too.