Authors: Atul Gawande
In medicine, we have long faced a conflict between the imperative to give patients the best possible care and the need to provide novices with experience. Residencies attempt to mitigate potential harm through supervision and graduated responsibility. And there is reason to think patients actually benefit from teaching. Studies generally find teaching hospitals have better outcomes than non-teaching hospitals. Residents may be amateurs, but having them around checking on patients, asking questions, and keeping faculty on their toes seems to help. But there is still no getting around those first few unsteady times a young physician tries to put in a central line, remove a breast cancer, or sew together two segments of colon. No matter how many protections we put in place, on average these cases go less well with the novice than with someone experienced.
We have no illusions about this. When an attending physician brings a sick family member in for surgery, people at the hospital think hard about how much to let trainees participate. Even when the attending insists that they participate as usual, a resident scrubbing in knows that it will be far from a teaching case. And if a central line must be put in, a first-timer is certainly not going to do it. Conversely, the ward services and clinics where residents have the most responsibility are populated by the poor, the uninsured, the drunk, and the demented. Residents have few opportunities nowadays to operate independently, without the attending docs scrubbed in, but when we do—as we must before graduating and going out to operate on our own—it is generally on these, the humblest of patients.
This is the uncomfortable truth about teaching. By traditional ethics and public insistence (not to mention court rulings), a patient’s right to the best care possible must trump the objective of training novices. We want perfection without practice. Yet everyone is harmed if no one is trained for the future. So learning is hidden, behind drapes and anesthesia and the elisions of language. Nor does the dilemma apply just to residents, physicians in training. In
fact, the process of learning turns out to extend longer than most people know.
My sister and I grew up in the small town of Athens, Ohio, where our parents are both doctors. Long ago my mother chose to practice pediatrics part-time, only three half-days a week, and she was able to because my father’s urology practice became so busy and successful. He has now been at it for more than twenty-five years, and his office is cluttered with the evidence of it: an overflowing wall of patient files, gifts from people displayed everywhere (books, paintings, ceramics with biblical sayings, hand-painted paperweights, blown glass, and carved boxes, as well as a figurine of a boy who pees on you when you pull down his pants). In an acrylic case behind his oak desk there are a few dozen of the thousands of kidney stones he has removed from these patients.
Only now, as I get glimpses of the end of my training, have I begun to think hard about my father’s success. For most of residency, I thought of surgery as a more or less fixed body of knowledge and skill which is acquired in training and perfected in practice. There was, as I envisioned it, a smooth, upward-sloping arc of proficiency at some rarefied set of tasks (for me, taking out gallbladders, colon cancers, bullets, and appendices; for him, taking out kidney stones, testicular cancers, and swollen prostates). The are would peak at, say, ten or fifteen years, plateau for a long time, and perhaps tail off a little in the final five years before retirement. The reality, however, turns out to be far messier. You do get good at certain things, my father tells me, but no sooner than you do, you find what you know is outmoded. New technologies and operations emerge to supplant the old, and the learning curve starts all over again. “Three-quarters of what I do today I never learned in residency,” he says. On his own, fifty miles from his nearest colleague—let alone a doctor who could tell him anything like “You need to turn your wrist more when you do that”—he has had to learn to put in penile prostheses, to perform
microsurgery, to reverse vasectomies, to do nerve-sparing prostatectomies, to implant artificial urinary sphincters. He’s had to learn to use shock-wave lithotripters, electrohydraulic lithotripters, and laser lithotripters (all instruments for breaking up kidney stones); to deploy Double J ureteral stents and Silicone Figure Four Coil stents and Retro-Inject Multi-Length stents (don’t even ask); to maneuver fiber-optic ureteroscopes. All these technologies and techniques were introduced since he finished training. Some of the procedures built on previous skills. Many did not.
This is, in fact, the experience all surgeons have. The pace of medical innovation has been unceasing, and surgeons have no choice but to give the new new thing a try. To fail to adopt new techniques would mean denying patients meaningful medical advances. Yet the perils of the learning curve are inescapable—no less in practice than in residency.
For the established surgeon, inevitably, the opportunities for learning are far less structured than for a resident. When an important new device or procedure comes along, as they do every year, surgeons start out by taking a course about it—typically a day or two of lectures by some surgical grandees with a few film clips and step-by-step handouts. We take a video home to watch. Perhaps we pay a visit to observe a colleague perform the operation—my father often goes up to Ohio State or the Cleveland Clinic for this. But there’s not much by way of hands-on training. Unlike a resident, a visitor cannot scrub in on cases, and opportunities to practice on animals or cadavers are few and far between. (Britain, being Britain, actually bans surgeons from practicing on animals.) When the pulsed-dye laser came out, the manufacturer set up a lab in Columbus where urologists from the area could gain experience. But when my father went, the main experience provided was destroying kidney stones in test tubes filled with a urinelike liquid and trying to penetrate the shell of an egg without hitting the membrane underneath. My surgery department recently purchased a robotic surgery device—a staggeringly sophisticated nine-hundred-and-eighty-thousand-dollar robot,
with three arms, two wrists, and a camera, all millimeters in diameter, which, controlled from a console, allows a surgeon to do almost any operation with absolutely no hand tremor and with only tiny incisions. A team of two surgeons and two nurses flew out to the manufacturer’s headquarters in San Jose for a full day of training on the machine. And they did get to practice on a pig and on a human cadaver. (The company apparently buys the cadavers from the city of San Francisco.) But even this, which is far more practice than one usually gets, was hardly thorough training. They learned enough to grasp the principles for operating the robot, to start getting a feel for using it, and to understand how to plan an operation. That was about it. Sooner or later, one just has to go home and give the thing a try.
Patients do eventually benefit—often enormously—but the first few patients may not and may even be harmed. Consider the experience reported by the pediatric-surgery unit of the renowned Great Ormond Street Hospital in London, as detailed in the
British Medical Journal
in the spring of 2000. The doctors described their results in operating on three hundred and twenty-five consecutive babies with a severe heart defect, known as transposition of the great arteries, over a period (from 1978 to 1998) when its surgeons changed from doing one operation for the condition to another. Such children are born with their heart’s outflow vessels transposed: the aorta emerges from the right side of the heart instead of the left and the artery to the lungs emerges from the left instead of the right. As a result, blood coming in is pumped right back out to the body instead of first to the lungs, where it can be oxygenated. This is unsurvivable. The babies died blue, fatigued, never knowing what it was to get enough breath. For years, switching the vessels to their proper positions wasn’t technically feasible. Instead, surgeons did something known as the Senning procedure: they created a passage inside the heart to let blood from the lungs cross backward to the right heart. The Senning procedure allowed children to live into adulthood. The weaker right heart, however, cannot sustain the body’s entire blood flow as long as the left. Eventually, these patients’ hearts failed,
and although most made it to adulthood, few lived to old age. Then, by the 1980s, a series of technological advancements made it possible to do a switch operation safely. It rapidly became the favored procedure. In 1986, the Great Ormond Street surgeons made the changeover, and their report shows that it was unquestionably a change for the better. The annual death rate after a successful switch procedure was less than a quarter that after the Senning, resulting in a life expectancy of sixty-three years instead of forty-seven. But the price of learning to do it was appalling. In their first seventy switch operations, the doctors had a 25 percent surgical death rate, compared with just 6 percent with the Senning procedure. (Eighteen babies died, more than twice the number of the entire Senning era.) Only with time did they master it: in their next hundred switch operations, just five babies died.
As patients, we want both expertise and progress. What nobody wants to face is that these are contradictory desires. In the words of one British public report, “There should be no learning curve as far as patient safety is concerned.” But that is entirely wishful thinking.
Recently, a group of Harvard Business School researchers who have made a specialty of studying learning curves in industry—in making semiconductors, building airplanes, and such—decided to examine learning curves among surgeons. They followed eighteen cardiac surgeons and their teams as they took on the new technique of minimally invasive cardiac surgery. This study, I was surprised to discover, is the first of its kind. Learning is ubiquitous in medicine, and yet no one had ever compared how well different clinicians actually do it.
The new heart operation—involving a small incision between ribs instead of a chest split open down the middle—proved substantially more difficult than the conventional one. Because the incision is too small to admit the usual tubes and clamps for rerouting blood to the heart-bypass machine, surgeons had to learn a trickier method, which involved balloons and catheters placed through groin vessels.
They had to learn how to operate in a much reduced space. And the nurses, anesthesiologists, and perfusionists all had new roles to master, too. Everyone had new tasks, new instruments, new ways that things could go wrong, and new ways to fix them. As you’d expect, everyone was found to experience a substantial learning curve. Whereas a fully proficient team takes three to six hours for such operations, these teams took an average of three times longer for their early cases. The researchers could not track rates of morbidity in detail, but it would be foolish to imagine that these rates were not affected.
What’s more interesting is that researchers found striking disparities in the speed with which different teams learned. All teams received the same three-day training session and came from highly respected institutions with experience in adopting innovations. Yet, in the course of fifty cases, some teams managed to halve their operating time while others failed to improve at all. Practice, it turned out, did not necessarily make perfect. Whether it did, the researchers found, depended on
how
the surgeons and their teams practiced.
Richard Bohmer, the one physician among the Harvard researchers, made several visits to observe one of the quickest-learning teams and one of the slowest, and he was startled by the contrast. The surgeon on the fast-learning team was actually quite inexperienced compared with the one on the slow-learning team—he was only a couple of years out of training. But he made sure to pick team members with whom he had worked well before and to keep them together through the first fifteen cases before allowing any new members. He had the team go through a dry run before the first case, then deliberately scheduled six operations in the first week, so little would be forgotten in between. He convened the team before each case to discuss it in detail and afterward to debrief. He made sure results were tracked carefully. And as a person, Bohmer noticed, the surgeon was not the stereotypical Napoleon with a knife. Unbidden, he told Bohmer, “The surgeon needs to be willing to allow himself
to
become a partner [with the rest of the team] so he can accept
input.” It sounded perhaps a little cliched; but then again, whatever he was doing worked. At the other hospital, the surgeon chose his operating team almost randomly and did not keep it together. In his first seven cases, the team had different members every time, which is to say that it was no team at all. And he had no pre-briefings, no debriefings, no tracking of ongoing results.
The Harvard Business School study offered some hopeful news. We can do things that have a dramatic effect on the learning curve—like being more deliberate about how we train, and about tracking progress, whether with students and residents or senior surgeons and nurses. But the study’s other findings are less reassuring. No matter how accomplished, surgeons trying something new got worse before they got better, and the learning curve proved longer, and affected by a far more complicated range of factors, than anyone had realized. It’s all stark confirmation that you can’t train novices without compromising patient care.
This, I suspect, is the reason for the physician’s dodge: the “I just assist” rap; the “We have a new procedure for this that you are perfect for” speech; the “You need a central line” without the “I am still learning how to do this.” Sometimes we do feel obliged to admit when we’re doing something for the first time, but even then we tend to quote the published success rates—which are virtually always from experienced surgeons. Do we ever tell patients that because we are still new at something, their risks will inevitably be higher, and that they’d likely do better with others who are more experienced? Do we ever say that we need them to agree to it anyway? I’ve never seen it. Given the stakes, who in their right mind would agree to be practiced upon?