The Checklist Manifesto Page 9
I searched anyway. I asked people around WHO for examples of public health interventions we could learn from. They came up with instances like the smallpox vaccination campaign that eradicated the scourge from the world in 1979 and the work of Dr. John Snow famously tracing a deadly 1854 London cholera outbreak to water in a public well. When the disease struck a London neighborhood that summer, two hundred people died in the first three days. Three-quarters of the area's residents fled in panic. Nonetheless, by the next week, some five hundred more died. The dominant belief was that diseases like cholera were caused by "miasmas"--putrefied air. But Snow, skeptical of the bad-air theory, made a map of where the deceased had lived and found them clustered around a single water source, a well in Soho's Broad Street. He interviewed the bereaved families about their habits. He made a careful statistical analysis of possible factors. And he concluded that contaminated water had caused the outbreak. (It was later discovered that the well had been dug next to a leaking cesspit.) Snow persuaded the local council to remove the water well's pump handle. This disabled the well, ended the spread of the disease, and also established the essential methods of outbreak investigation that infectious disease specialists follow to this day.
All the examples, I noticed, had a few attributes in common: They involved simple interventions--a vaccine, the removal of a pump handle. The effects were carefully measured. And the interventions proved to have widely transmissible benefits--what business types would term a large ROI (return on investment) or what Archimedes would have called, merely, leverage.
Thinking of these essential requirements--simple, measurable, transmissible--I recalled one of my favorite public health studies. It was a joint public health program conducted by the U.S. Centers for Disease Control and HOPE, a charitable organization in Pakistan, to address the perilous rates of premature death among children in the slums of Karachi. The squatter settlements surrounding the megacity contained more than four million people living under some of the most crowded and squalid conditions in the world. Sewage ran in the streets. Chronic poverty and food shortages left 30 to 40 percent of the children malnourished. Virtually all drinking water sources were contaminated. One child in ten died before age five--usually from diarrhea or acute respiratory infections.
The roots of these problems were deep and multifactorial. Besides inadequate water and sewage systems, illiteracy played a part, hampering the spread of basic health knowledge. Corruption, political instability, and bureaucracy discouraged investment in local industry that might provide jobs and money for families to improve their conditions. Low global agriculture prices made rural farming life impossible, causing hundreds of thousands to flock to the cities in search of work, which only increased the crowding. Under these circumstances, it seemed unlikely that any meaningful improvement in the health of children could be made without a top-to-bottom reinvention of government and society.
But a young public health worker had an idea. Stephen Luby had grown up in Omaha, Nebraska, where his father chaired the obstetrics and gynecology faculty at Creighton University. He attended medical school at the University of Texas Southwestern. But for some reason he was always drawn to public health work. He took a CDC job investigating infectious outbreaks in South Carolina, but when a position came open in the CDC's Pakistan office he jumped to take it. He arrived in Karachi with his schoolteacher wife and began publishing his first investigations of conditions there in the late nineties.
I had spoken to him once about how he thought through the difficulties. "If we had the kinds of water and sewage systems we've got in Omaha, we could solve these problems," he said. "But you have to wait decades for major infrastructure projects." So instead, he said, he looked for low-tech solutions. In this case, the solution he came up with was so humble it seemed laughable to his colleagues. It was soap.
Luby learned that Procter & Gamble, the consumer product conglomerate, was eager to prove the value of its new antibacterial Safeguard soap. So despite his colleagues' skepticism, he persuaded the company to provide a grant for a proper study and to supply cases of Safeguard both with and without triclocarban, an antibacterial agent. Once a week, field-workers from HOPE fanned out through twenty-five randomly chosen neighborhoods in the Karachi slums distributing the soap, some with the antibacterial agent and some without. They encouraged people to use it in six situations: to wash their bodies once daily and to wash their hands every time they defecated, wiped an infant, or were about to eat, prepare food, or feed it to others. The field-workers then collected information on illness rates among children in the test neighborhoods, as well as in eleven control neighborhoods, where no soap was distributed.
Luby and his team reported their results in a landmark paper published in the Lancet in 2005. Families in the test neighborhoods received an average of 3.3 bars of soap per week for one year. During this period, the incidence of diarrhea among children in these neighborhoods fell 52 percent compared to that in the control group, no matter which soap was used. The incidence of pneumonia fell 48 percent. And the incidence of impetigo, a bacterial skin infection, fell 35 percent. These were stunning results. And they were achieved despite the illiteracy, the poverty, the crowding, and even the fact that, however much soap they used, people were still drinking and washing with contaminated water.
Ironically, Luby said, Procter & Gamble considered the study something of a disappointment. His research team had found no added benefit from having the antibacterial agent in the soap. Plain soap proved just as effective. Against seemingly insuperable odds, it was more than good enough. Plain soap was leverage.
The secret, he pointed out to me, was that the soap was more than soap. It was a behavior-change delivery vehicle. The researchers hadn't just handed out Safeguard, after all. They also gave out instructions--on leaflets and in person--explaining the six situations in which people should use it. This was essential to the difference they made. When one looks closely at the details of the Karachi study, one finds a striking statistic about the house-holds in both the test and the control neighborhoods: At the start of the study, the average number of bars of soap house holds used was not zero. It was two bars per week. In other words, they already had soap.
So what did the study really change? Well, two things, Luby told me. First, "We removed the economic restraint on purchasing soap. People say soap is cheap and most house holds have soap. But we wanted people to wash a lot. And people are quite poor. So we removed that as a barrier." Second, and just as important, the project managed to make soap use more systematic.
Luby and his team had studied washing behavior in Pakistan, Bangladesh, and other locations around South Asia, and they found that almost everyone washes their hands after defecation. "There are strong ideas about purity in South Asia," he said. Even when the place to wash is far away, people go and clean their hands over 80 percent of the time, a rate that would put most denizens of airport bathrooms to shame. But the washing was not very effective, the researchers found. Often people did it too quickly. Or they cleaned just the "involved" hand. Or they used ash or mud rather than soap and water.
The soap experiment changed that. The field-workers gave specific instructions on hand-washing technique--on the need to wet both hands completely, to lather well, to rinse all the soap off, even if, out of necessity, as the published report noted, "hands were typically dried on participants' clothing." The instructions also got people used to washing at moments when they weren't used to doing so. "Before preparing food or feeding a child is not a time when people think about washing," Luby explained. The soap itself was also a factor. "It was really nice soap," he pointed out. It smelled good and lathered better than the usual soap people bought. People liked washing with it. "Global multinational corporations are really focused on having a good consumer experience, which sometimes public health people are not." Lastly, people liked receiving the soap. The public health field-workers were bringing them a gift rather than wagging a finger. And with the gift
came a few basic ideas that would improve their lives and massively reduce disease.
Thinking back on the experiment, I was fascinated to realize that it was as much a checklist study as a soap study. So I wondered: Could a checklist be our soap for surgical care--simple, cheap, effective, and transmissible? I still had a hard time grasping how to make a checklist that could be both simple and effective for the manifold problems posed by surgery on a global scale. I was uncertain that it was even possible. But several of my colleagues were more sanguine when the idea was raised at the Geneva meeting.
One brought up the experience of Columbus Children's Hospital, which had developed a checklist to reduce surgical infections. Infection is one of the most common complications of surgery in children. And the most effective way to prevent it, aside from using proper antiseptic technique, is to make sure you give an appropriate antibiotic during the sixty-minute window before the incision is made.
The timing is key. Once the incision is made, it is too late for the antibiotic. Give it more than sixty minutes before the procedure, and the antibiotic has worn off. But give it on time and studies show this single step reduces the infection risk by up to half. Even if the antibiotic is squeezed into the bloodstream only thirty seconds before the incision is made, researchers have found, the circulation time is fast enough for the drug to reach the tissue before the knife breaches the skin.
Yet the step is commonly missed. In 2005, Columbus Children's Hospital examined its records and found that more than one-third of its appendectomy patients failed to get the right antibiotic at the right time. Some got it too soon. Some got it too late. Some did not receive an antibiotic at all.
It seems dumb. How hard could this be? Even people in medicine assume we get this kind of simple task right 100 percent of the time. But in fact we don't. With all the flurry of things that go on when a patient is wheeled into an operating room, this is exactly the sort of step that can be neglected. The anesthesiologists are the ones who have to provide the antibiotic, but they are concentrating on getting the patient safely and calmly to sleep--and this is no small matter when that patient is a scared eight-year-old lying naked on a cold table in a room full of strangers getting ready to cut into her. Add in an equipment malfunction ("Is that red light supposed to be blinking like that?"), or the patient's asthma acting up, or a page for the surgeon to call the emergency room, and you begin to see how something as mundane as an antibiotic can slip past.
The hospital's director of surgical administration, who happened to be not only a pediatric cardiac surgeon but also a pilot, decided to take the aviation approach. He designed a preincision "Cleared for Takeoff " checklist that he put on a whiteboard in each of the operating rooms. It was really simple. There was a check box for the nurse to verbally confirm with the team that they had the correct patient and the correct side of the body planned for surgery--something teams are supposed to verify in any case. And there was a further check box to confirm that the antibiotics were given (or else judged unnecessary, which they can be for some operations).
There wasn't much more to it. But getting teams to stop and use the checklist--to make it their habit--was clearly tricky. A couple of check boxes weren't going to do much all by themselves. So the surgical director gave some lectures to the nurses, anesthesiologists, and surgeons explaining what this checklist thing was all about. He also did something curious: he designed a little metal tent stenciled with the phrase Cleared for Takeoff and arranged for it to be placed in the surgical instrument kits. The metal tent was six inches long, just long enough to cover a scalpel, and the nurses were asked to set it over the scalpel when laying out the instruments before a case. This served as a reminder to run the checklist before making the incision. Just as important, it also made clear that the surgeon could not start the operation until the nurse gave the okay and removed the tent, a subtle cultural shift. Even a modest checklist had the effect of distributing power.
The surgical director measured the effect on care. After three months, 89 percent of appendicitis patients got the right antibiotic at the right time. After ten months, 100 percent did. The checklist had become habitual--and it had also become clear that team members could hold up an operation until the necessary steps were completed.
I was intrigued. But I remained doubtful. Yes, using a checklist, this one hospital got one aspect of care to go consistently right for surgical patients. I was even willing to believe their surgical infection rates had fallen significantly as a result. But to take a serious bite out of overall complication rates, I argued, we needed an approach that would help across the much wider range of ways in which surgery can go wrong.
Then Richard Reznick, the chairman of surgery at the University of Toronto, spoke up. He explained that his hospital had completed a feasibility trial using a much broader, twenty-one-item surgical checklist. They had tried to design it, he said, to catch a whole span of potential errors in surgical care. Their checklist had staff verbally confirm with one another that antibiotics had been given, that blood was available if required, that critical scans and test results needed for the operation were on hand, that any special instruments required were ready, and so on.
The checklist also included what they called a "team briefing." The team members were supposed to stop and take a moment simply to talk with one another before proceeding--about how long the surgeon expected the operation to take, how much blood loss everyone should be prepared for, whether the patient had any risks or concerns the team should know about.
Reznick had never heard about the demise of Master Builders, but he had gravitated intuitively toward the skyscraper solution--a mix of task and communication checks to manage the problem of proliferating complexity--and so had others, it turned out. A Johns Hopkins pancreatic surgeon named Martin Makary showed us an eighteen-item checklist that he'd tested with eleven surgeons for five months at his hospital. Likewise, a group of Southern California hospitals within the Kaiser health care system had studied a thirty-item "surgery preflight checklist" that actually predated the Toronto and Hopkins innovations. All of them followed the same basic design.
Surgery has, essentially, four big killers wherever it is done in the world: infection, bleeding, unsafe anesthesia, and what can only be called the unexpected. For the first three, science and experience have given us some straightforward and valuable preventive measures we think we consistently follow but don't. These misses are simple failures--perfect for a classic checklist. And as a result, all the researchers' checklists included precisely specified steps to catch them.
But the fourth killer--the unexpected--is an entirely different kind of failure, one that stems from the fundamentally complex risks entailed by opening up a person's body and trying to tinker with it. Independently, each of the researchers seemed to have realized that no one checklist could anticipate all the pitfalls a team must guard against. So they had determined that the most promising thing to do was just to have people stop and talk through the case together--to be ready as a team to identify and address each patient's unique, potentially critical dangers.
Perhaps all this seems kind of obvious. But it represents a significant departure from the way operations are usually conducted. Traditionally, surgery has been regarded as an individual performance--the surgeon as virtuoso, like a concert pianist. There's a reason that much of the world uses the phrase operating theater. The OR is the surgeon's stage. The surgeon strides under the lights and expects to start, everyone in their places, the patient laid out asleep and ready to go.
We surgeons want to believe that we've evolved along with the complexity of surgery, that we work more as teams now. But however embarrassing it may be for us to admit, researchers have observed that team members are commonly not all aware of a given patient's risks, or the problems they need to be ready for, or why the surgeon is doing the operation. In one survey of three hundred staff members as they exited the operating room following a case, one out of eight reported that t
hey were not even sure about where the incision would be until the operation started.
Brian Sexton, a pioneering Johns Hopkins psychologist, has conducted a number of studies that provide a stark measure of how far we are from really performing as teams in surgery. In one, he surveyed more than a thousand operating room staff members from hospitals in five countries--the United States, Germany, Israel, Italy, and Switzerland--and found that although 64 percent of the surgeons rated their operations as having high levels of teamwork, just 39 percent of anesthesiologists, 28 percent of nurses, and 10 percent of anesthesia residents did. Not coincidentally, Sexton also found that one in four surgeons believed that junior team members should not question the decisions of a senior practitioner.
The most common obstacle to effective teams, it turns out, is not the occasional fire-breathing, scalpel-flinging, terror-inducing surgeon, though some do exist. (One favorite example: Several years ago, when I was in training, a senior surgeon grew incensed with one of my fellow residents for questioning the operative plan and commanded him to leave the table and stand in the corner until he was sorry. When he refused, the surgeon threw him out of the room and tried to get him suspended for insubordination.) No, the more familiar and widely dangerous issue is a kind of silent disengagement, the consequence of specialized technicians sticking narrowly to their domains. "That's not my problem" is possibly the worst thing people can think, whether they are starting an operation, taxiing an airplane full of passengers down a runway, or building a thousand-foot-tall skyscraper. But in medicine, we see it all the time. I've seen it in my own operating room.
Teamwork may just be hard in certain lines of work. Under conditions of extreme complexity, we inevitably rely on a division of tasks and expertise--in the operating room, for example, there is the surgeon, the surgical assistant, the scrub nurse, the circulating nurse, the anesthesiologist, and so on. They can each be technical masters at what they do. That's what we train them to be, and that alone can take years. But the evidence suggests we need them to see their job not just as performing their isolated set of tasks well but also as helping the group get the best possible results. This requires finding a way to ensure that the group lets nothing fall between the cracks and also adapts as a team to what ever problems might arise.