June 2003 Bulletin

Putting backbone into safer surgery efforts:
A roadmap for change

By Michael L. Millenson

This article is based on remarks made by Mr. Millenson on April 10 during the AAOS Board of Directors Workshop on Patient Safety in Naples, Fla.

Seventeen years ago, when I started doing research about quality of care, I quickly encountered a thoughtful and dynamic orthopedic surgeon named Bob Keller. The founder of the Maine Medical Assessment Foundation, Bob was a pioneer in using the power of professionalism to address practice variation. Given the outright hostility of many of his colleagues at the time, it wasn’t easy.

The American Academy of Orthopaedic Surgeons (AAOS) was also a pioneer. When other professional societies responded to the medical literature about practice variation with skepticism and stonewalling, the AAOS took a very different tack. Concerned about cost-effective, high-quality care, the Academy issued a soft-cover reference guide for surgeons that included reprints of key research articles, as well as information on AAOS efforts related to the problem.

Because of the great respect I have for your tradition of leadership, I am both pleased and honored to keynote this important meeting. And because of that tradition, I want to challenge you to live up to your potential as change agents — both as individuals and as a professional society — and to put some real backbone into the patient safety movement. The key question facing the profession relates to the interplay of evidence, ethics and effective actions.

Make a commitment

Forget position papers and proclamations. Are your actions and those of your colleagues in regard to ensuring the safety of each patient consonant not only with the medical literature, but also with your medical ideals? If not, what specific changes must you undertake—both as individual physicians and as a professional society?

But before we discuss the how and why of professional change, let’s listen to another voice, the voice of the patient. These stories do not necessarily involve orthopedic patients, but they share a common thread. Here’s the first story:

"The day after my father’s third joint replacement may have been the luckiest day of his life. That morning, he received a morphine injection. He was allergic to morphine. Dad was struggling with post-operative pain … Because of his [various] allergies, as well as potential drug interactions from his arthritis medication, he was taking only acetaminophen for post-op pain. So the doctor ordered what was described as ‘half of a child’s dose of morphine.’ He was given a standard adult dose by injection. My mother walked into the hospital room the instant after the injection, with the nurse still there and, to hear Mom tell it, my father’s eyes rolling back into his head. The events were astonishingly rapid: A conversation with the nurse. A code called. A crash team. A second wave of medical residents who were on rounds nearby. A positive resolution. No harm done—except perhaps to our faith in medical care."

Those words come from a column entitled, "A Son’s Plea," by Alden Solovy, the associate publisher of Hospitals & Health Networks, the official publication of the American Hospital Association.

Shrug of the shoulders

Medical errors. Of course, nobody is in favor of them. Nonetheless, I’d like to raise the question of whether it’s really worth taking the time and effort needed to address this problem. By talking candidly about the continuing justifications for inaction in regard to patient safety, I hope to provide ammunition for you to refute those arguments when you hear them from others.

Over the years, physicians have responded to medical errors with the equivalent of a regretful shrug of the shoulders. Stuff happens. Medical mistakes are part of the "hazards of modern diagnosis and therapy—the price we pay."

Those words, which sound so contemporary, were written in an issue of the Journal of the American Medical Association (JAMA) published in 1955. The fatalistic attitude toward patient fatalities has not been isolated. In early 1976, in the middle of a malpractice "crisis," the New York Times published a series of page-one articles about quality of care problems, including an examination of medication errors. Even the American Medical Association admitted the death rate was too high, and the profession as a whole promised to make things better. But … nothing happened.

In 1986, a grand jury report resulting from the death of Libby Zion, daughter of New York Times reporter Sidney Zion, recommended that computerized drug-drug interaction systems be instituted at all major teaching hospitals in New York State. And, for those who remember, it was the death of Libby Zion that started the debate about residents’ work hours–a problem that the medical profession is finally addressing in an official way a mere 17 years later. Also around 1986, the People’s Medical Society, a consumer group, recommended that doctors undertake a "sign your site" program to eliminate wrong-site surgery. Nothing happened to the recommendation about doctors changing their use of computers. Nothing happened to the recommendation about doctors changing their use of pens.

To Err is Human

Finally, you get the Institute of Medicine (IOM) report in late 1999, To Err is Human. Many were shocked at the reported level of medical errors.

Leaders of medical and hospital organizations were also shocked. What shocked them, though, was that the report was such a big deal. After all, the basic data had been published in 1991, based on a study of medical charts from 1984. In private, many leaders expected the issue to go away, as it had so many times in the past. Except this time, it was different. This time, the individual consumer was paying attention.

With To Err is Human, patient safety stopped being a subject for medical journals and physician conferences and started being a topic of everyday conversation. Put differently: the medical literature hadn’t changed. Professional ethics and the obligation to patients hadn’t changed. And even technology hadn’t changed that much.

But something was different. Patient safety was suddenly front and center on the professional and public agenda. What happened to cause that change? How can we support and even accelerate the process of change? Where, in other words, is the "roadmap for change?"

Before we explore that topic, let’s pause again to hear the patient’s voice. This anecdote comes from a consultant discussing what happened to her 80-year-old father after he suffered a stroke.

"The afternoon shift nurse came in and asked if the morning shift nurse had given my father his meds, because it was not documented in the chart. She had no idea, given all the confusion. And it was serious meds: blood thinners. They had no idea if it was a double-dose or not. It wasn’t on the chart. It turns out they didn’t give it to him."

Some years ago, the sociologist Everett Rogers wrote: "Getting a new idea adopted, even when it has obvious advantages, is often very difficult." This rather mundane-seeming insight turned out to be anything but. People don’t change their behavior just because it’s "good for them." Appeals to rationality are only one part of the picture.

Rogers’ book, Diffusion of Innovations, demonstrates that the speed with which any innovation is embraced depends on five characteristics. The first characteristic is relative advantage over what currently exists; the second, compatibility with existing values and behaviors; third, lack of complexity; fourth, the ability to be subjected to experiment ("trialability"); and fifth, producing results everyone can see ("observability"). Let’s see how these criteria apply to the patient safety movement.

Relative advantage

That first requirement—that an innovation produce "relative advantage"—is critical. Demonstrating relative advantage can be surprisingly difficult to achieve because the innovator’s advantage must not only be real, it must be perceived as real. Applied to the patient safety question, that means that errors first must be perceived by the profession as an urgent problem for the profession to work urgently on a solution. That has not been true and still isn’t a universal perception.

For instance, many hard-working, caring physicians believe that the overall magnitude of deaths and injuries due to preventable medical error is overstated, no matter what the IOM or the medical literature might say. Or, they believe that regardless of what the overall error rate might be nationally, it’s not their problem—"Errors aren’t happening at my institution." This perception is not surprising. Many categories of errors might not be understood as errors by the clinicians involved, or those errors that are clear-cut may not come to the attention of the clinicians who were not personally involved.

Another barrier to perceiving error reduction as producing relative advantage is the belief that the intensity of the harm done to patients is exaggerated by attorneys looking for big jury awards in malpractice cases. As Troyen Brennan, a lawyer and physician, wrote in the April 13, 2000 issue of the New England Journal of Medicine: "To address the problem of iatrogenic injuries seriously, we must reform the system of malpractice litigation."

Let me tell you what that sounds like to a patient: "If Congress agrees to cap pain-and-suffering damages at $250,000 per patient, doctors will stop cutting off the wrong limb and giving drug overdoses." Or, perhaps: "First we kill the lawyers, then we stop killing the patients." So much for "first, do no harm."

During this period when malpractice reform is a topic of great debate, whatever the connection may be between defensive medicine and health care costs—and the medical literature shows that very few medical errors result in malpractice suits, much less payments —there is no evidence that I know of to suggest that defensive medicine is a disproportionately large contributor to iatrogenic injuries. To suggest otherwise, in my view, panders to physician emotionalism at the expense of removing the incentive to change.

How do we change perceptions about errors? One place to start is to change the stories we tell. Rather than telling stories about greedy lawyers or unfair patient expectations or the shortage of good nurses, we need to tell stories that reflect a different reality. Stories are important; they make a difference in how we perceive what goes on around us.

When it comes to wrong-site surgery, one story made a very big difference. It didn’t appear first in a medical journal but, on TV and radio and in newspapers all over the country. It was the 1995 story of a diabetic named Willie King who had the wrong foot amputated.

Between January 1985 and December 1995, there were 225 wrong-site orthopaedic surgery claims, according to the Physician Insurance Association of America. That was more than twice the 106 claims in all other surgical specialties combined. But it wasn’t until a victim became a clearly identifiable individual that the medical profession and various watchdog groups finally started to take action.

The Canadian Orthopaedic Association began its own wrong-site surgery prevention program in 1994. But the AAOS began to build off that effort, starting its "Sign Your Site" program in 1997, only after the highly public Willie King scandal sparked lots of news coverage, lots of soul-searching by orthopedists…and a new sense of regulatory fervor by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO).

Human-interest stories about medical errors—including the death of Boston Globe columnist Betsy Lehman from a cancer drug overdose — both helped inspire To Err is Human and greatly increased the impact of the IOM report’s impact.

Error statistics attached to a name and a human face provided reporters and politicians with real people to dramatize the 44,000 to 98,000 preventable-deaths-per-year dry statistics. The average American stopped thinking of medical errors as something that could happen to "the other guy," and was struck by this thought: "Hey, that could be me."

It was the public uproar that made the problem into an urgent one. The "relative advantage" for a hospital and for physicians became clear: NOT to be shamed on page one of the local newspaper. NOT to be cited by JCAHO. NOT to have to confront an uncomfortable gap between your ideals and your actual practices.

More reinforcement has come from state and federal legislators. More than 15 states now have mandatory adverse-event reporting. JCAHO requires telling patients about errors that harmed them or a loved one. And the federal government is considering a "voluntary" national error reporting initiative.

Then there’s the economic aspect of relative advantage.

Economic advantage

The Leapfrog Group, a coalition of Fortune 500 companies launched in November 2000, now has 135 public and private member organizations representing some 33 million covered lives in all 50 states. Leapfrog member companies spend, literally, billions of dollars on health care.

What do Leapfrog members promise to do? "Purchasers will aggregate available validated performance information on their major providers of health care into comparative value ratings for their employees." More recently, they began a pilot project to reward hospitals with a payment premium for high quality.

One last point on relative advantage—which is really the critical starting point of any innovation adoption. The AAOS has recognized—in my view, very shrewdly—that the relative advantage of safety initiatives needs to be personalized. And so, your leadership has begun noting that an orthopedic surgeon who practices for 25 years has a one-in-four chance of committing a wrong-site surgery.

In other words, the anguish and pain you prevent may well be your own.


The AAOS [patient safety] initiative has another very important impact. An emphasis on improving patient safety has become consistent "with existing values and behaviors" – a second condition Rogers says must be present for an innovation to catch on.

NOT making mistakes has always been consistent with professional values. But tackling medical errors as a systems problem was not. Just a few months before the Willie King incident, a commentary in a December 1994 issue of JAMA by David Blumenthal, MD, bemoaned the profession’s "ostrich-like" attitude towards errors. The profession both pretended errors were uncommon–when they were not–and ignored the literature on error prevention.

Blumenthal’s commentary accompanied an article on errors by his colleague, Lucian Leape, MD. And what was the reaction of doctors to those two articles? JAMA editor George Lundberg, MD, later admitted, "Hate mail began pouring in [from AMA members]. I was accused of being on the side of lawyers, a damned turncoat and a traitor to the cause."

Today, I don’t think that kind of reaction would happen. But it’s not the medical literature that’s changed, it’s the culture of professional values.

Culture change

The AAOS has helped lead culture change, for which you deserve credit. Now, what will you do to accelerate effective actions by orthopaedic surgeons as a result of that culture change? And what will you do to make sure that the culture never changes back to the "bad old days"? What those who seek lasting organizational change have learned is that, while tools are important, so, too, is the culture in which those tools are applied. As Donald Berwick, MD, has written:

"Scientific approaches…are not sufficient. To be effective, they must be applied within organizations and cultures capable of nurturing and sustaining them, much as scientific clinical work can thrive only in cultures where it is valued."

Malcolm Gladwell expressed a variation of the same idea in "The Tipping Point," referring to how "trends" become mainstream. Ideas are like a virus. They are, in a sense, contagious.

When an idea fails to make it out of the "trendsetter" community, it is because it doesn’t root itself broadly enough in the culture. As one expert said, "There aren’t enough cues. Usually, if something’s going to make it, you’ll see that thread running throughout everything." To start an idea "epidemic," Gladwell concludes, you have to really concentrate your resources in a few areas.

Will AAOS concentrate its resources on patient safety? This meeting is a good sign. But let me also sound a cautionary note. I began by praising Bob Keller’s work on practice variation. He, and his work, was path breaking. But, ultimately, he and the AAOS were not able to penetrate the immune system of the profession.

Another quick patient story from Norbert Goldfield, a physician and co-author of a well-known handbook for physicians on quality. But here he’s talking as an individual.

"My mother had significantly elevated high blood pressure when she was initially admitted [for a stroke]. Unfortunately, she was given the wrong medication. The medical care was clearly inadequate during her first few days. There is absolutely no question in my mind that my mother would not be alive today if I were not a physician and watching over every step of the care."

We’ve been talking about perceptions, about believing that change is necessary. If Goldfield hadn’t written a letter as a physician to the hospital’s CEO about this incident, would his mom’s doctors have perceived that their institution needed to change?


Back to Diffusion of Innovations. Relative advantage and a fit with existing values are just the first two parts of the puzzle. The next step is "trialability;" that is, persuading clinicians there are specific actions they can take to address the problem.

Organizations like the Institute for Healthcare Improvement, Kaiser Permanente, the Veterans Administration and others have all begun to work in the trenches to show that real change is possible. On a macro level, success stories have been publicized by organizations as diverse as the IOM and USA Today.

Coalitions of hospitals, physicians and employers are starting to address safety cooperatively to show that this change model can be successful. One example is the Pittsburgh Regional Healthcare Initiative, which has endorsed a "zero tolerance" goal for preventable medication errors and preventable nosocomial infections. If, prodded by former Alcoa chairman Paul O’Neill, Pittsburgh hospitals can embrace those ambitious goals, can orthopaedic surgeons nationally do anything less than endorse the goal of "zero tolerance" for wrong-site surgery?

AAOS’s ongoing efforts to encourage "trialability" provide crucial credibility to those who would make patient care safer. Most visibly, speakers at this conference can provide valuable suggestions based on their own successes and failures. Less visibly, physicians identify with their specialty society. What AAOS provides is not just advice, but peer-reinforcement that the advice is worthwhile.

Your activities not only make error-reduction "trialable"—and underscore its compatibility with existing health-care values—they also reduce the complexity of innovation, the fourth requirement on Rogers’ list. Here’s an additional example of reduced complexity: the Joint Commission has introduced standardized error-reporting rules that all hospitals can use to make improvements based on a "root cause analysis" of mistakes.

Observable results

Take "trialability," combine it with the lessening of complexity and you get "observable results." Some of these results are publicized informally, in the trade press or at meetings. Other results are published in the medical literature.

All five factors necessary for an innovation to take hold–relative advantage; compatibility with existing values and behaviors; lack of complexity; the ability to be subjected to experiment ("trialability"); and producing results everyone can see ("observability")–are now present in regard to error reduction in health care and, in particular, to the efforts to make orthopaedic surgery safer.

Moreover, I believe the pressure to address this problem will increase as Baby Boomers and their parents age, and as consumers become more informed. All the patient anecdotes I have used came from savvy, well-informed children talking about the care of their parents. Each of those children was someone who worked full-time in health care. In other words, people just like you and me and our friends and family.

Rest of the roadmap

Accountability, transparency of results and evidence-based standards–including adherence to the highest-quality safety practices–are becoming the new cultural norm in medicine. Many of the components involved in improving patient safety are systematic in nature. There are organizational factors, such as an institutions policies and procedures. There are environmental factors, such as equipment, staffing and resources. There are human factors, such as clinical competency or communication skills.

And yet we should not delude ourselves by talking euphemistically: system changes depend upon decisions made by individuals. Individuals decide whether to spend the time and money needed to fix failings in safety systems. Individuals decide to support computerized physician order entry that produces safer care or insist, instead, that the money be spent on a computerized diagnostic machine that produces billable scans.

Individuals commit to clear performance expectations in patient safety–or cling, instead, to the belief that the inability to achieve perfection justifies inaction on improvement.

Ultimately, individuals—like you—must stand up to those who believe the status quo is an option and tell them that it is not. The problem is NOT that individual doctors and nurses don’t care about individual patients. They DO care: deeply and passionately. But caring is not enough; evidence and ethics demand that we take effective action. Now.

To emphasize that point, I’d like to share a last patient story. These words, by the mother of Boston Globe reporter Betsy Lehman can serve as inspiration to all of us. Mrs. Lehman wrote:

"May I appeal to you to pause for a moment, if you will, in your important task. For in the wings outside your busy meeting rooms may be heard the murmurings of patients gone now due to fatal medical error, or harmed by a medical system they trusted. They are the ones absent … among them is my young, brilliant daughter. Patient safety must be utmost and constant, both ingrained into the system you seek to strengthen, and into caring hearts."

Memory, tradition and myth

In many ways, professional societies are the repositories of the memories, traditions and myths of the medical specialties that they represent. You at the AAOS have a great deal to be proud of in your tradition. But for all your achievements, you have a great deal further to go. After all, 20 to 40 percent of your members still do not sign their site.

The challenge that lies ahead for the AAOS is to build a new tradition that integrates the best of the medical evidence, medical ethics and effective actions. You have already started to blaze that trail. I look forward to doing whatever I can to assist you.

Michael L. Millenson is a principal in the health care and group benefits consulting practice of William M. Mercer, Inc., a visiting scholar at Northwestern University and an award-winning journalist. He is author of the book "Demanding Medical Excellence: Doctors and Accountability in the Information Age."

Home Previous Page