In May 2018, Yale University Medical School invited a select group of cardiologists, researchers, and health policy experts to discuss the controversial ORBITA trial in detail. Shannon Brownlee, senior vice president of the Lown Institute, and Dr. Vikas Saini, president of the Lown Institute, gave remarks about the historical context and implications of ORBITA; their edited remarks have now been published on the Health Affairs blog.
In Brownlee’s piece, posted below, she discusses the ORBITA trial in the broader context of medical reversal. ORBITA is just one of the many medical tests and procedures that has been proven to be no better than a placebo, like arthroscopic knee surgery, carotid endartectomy, vertebrosplasty, and many, many more. But once a procedure has been cemented into practice norms, it is hard to dislodge, even when evidence shows it is ineffective or unsafe. The response from some cardiologists to the ORBITA trial is a perfect example of this inertia, Brownlee argues.
Read the full piece below or on the Health Affairs blog!
This piece was originally published on the Health Affairs Blog.
By Shannon Brownlee
What can be done to encourage doctors and other clinicians to heed new evidence when it shows the tests or procedures they are delivering are in fact not helping their patients—and may even harm them? This question has become increasingly relevant in the face of a growing number of studies casting doubt on the safety and efficacy of commonly used treatments and tests. Even when new evidence is clearly negative, clinicians often resist abandoning a service they have believed in, and developed expertise in, especially when that service provides a significant portion of their livelihoods.
Take the PSA, or Prostate Specific Antigen test, widely promoted for more than three decades to screen for prostate cancer. Most patients and physicians assumed that catching cancer early leads to a lower risk of death. But results from several large randomized controlled clinical trials, which compared men who received regular PSA screening versus usual care, showed little to no difference between the two groups in terms of their risk of dying of prostate cancer. The PSA test, it turns out, leads to the diagnosis of unthreatening tumors in 20 to 40 percent of cases.
Other recently discredited procedures and tests include arthroscopic knee surgery, no better than placebo (or sham) surgery for relieving arthritis pain. Likewise veterbroplasty, which involves injecting a kind of superglue into the patient’s vertebrae and is also intended to relieve arthritis pain. And now, the mother of all medical reversals, the ORBITA trial, whose publication last year ignited a firestorm of controversy. ORBITA found that implanting a cardiac stent to relieve angina, or chest pain due to heart disease, is no better than a sham stent procedure.
One might think that such strong negative evidence would be enough to get doctors to give up a practice, or at least be more judicious in choosing their patients. After all, negative evidence says the procedure is more likely to harm their patients than help them.
Instead, clinicians often persist in delivering services long after the data say they should have scaled back or stopped. Eventually, truly useless interventions go the way of bloodletting, but, in the meantime, patients are exposed to the potential risk of harm with little or no chance of having their symptoms or conditions effectively treated. So the question of how to motivate clinicians to cease and desist should concern anyone who wants to reduce overuse and ensure that patients get the right care.
I came up against this problem in a very personal way about eight years ago, after one of my brothers called to tell me our then 80-year-old father, Mick Brownlee, had been hospitalized with a stroke. A vascular surgeon was recommending a carotid endarterectomy, Could I get on the phone to help decide what to do?
At the time, carotid endarterectomy had been a topic of intense debate in the neurology and vascular surgery communities for nearly a decade. In use since the 1950s, the procedure is a sort of Roto-Rooter for the arteries in the neck that supply the brain with blood. Built-up plaque inside the carotids can break free and move into smaller arteries in the brain, where it can block the flow of blood and cause a stroke. Endarterectomy removes some of that plaque in the carotids, reducing the risk of a stroke over the next five years.
Unfortunately, the procedure itself can actually cause a stroke or even death in about 4 percent of patients. A string of studies had shown that the surgery is worth doing only for some patients; for many others, the potential benefit is marginal. For most of the patients who were getting it, the risk of stroke was greater any potential protective effect the surgery might provide. In 2005, as many as 92 percent of the 135,701 carotid endarterectomies in the United States were still being performed on patients who stood a greater chance of being harmed by the procedure, not helped.
By the time my father had his stroke, the total number of endarterectomies had declined, but there was not a lot of information about whether or not surgeons were paying attention to the data and offering the procedure only to the right patients. I was pretty sure Mick was not an appropriate patient. For one thing, he was growing increasingly frail. For another, he’d fallen several times in the last year. He had lost weight, and his short-term memory was shot—all signs that he was unlikely to live five years more and thus was not a good candidate.
When I got on the phone with brothers, stepmother, father, and the surgeon, my family shared the impression that the endarterectomy might somehow, magically, bring back the Mick of just a couple of years before, when he was still sculpting, and struggled less with memory and balance. My saying otherwise would have no effect. The doctor had to say it.
The vascular surgeon began by pronouncing Mick an “appropriate candidate” for this “highly effective surgery.” Then he asked if we had any questions. Eventually, I asked the question that hit home: “What will this surgery do for his quality of life?”
“Nothing,” said the surgeon. “It’s all about preventing another stroke.”
There was murmuring on the line. “Are you saying the surgery won’t help him with his memory? Or walking?” asked my stepmother. Eventually my family understood that the surgery would not change Mick’s current situation and decided against it.
Maybe this surgeon had not been reading his medical journals, and was recommending an endarterectomy because he was unaware of the evidence. I think the real reason has less to do with knowledge than psychology. Most clinicians believe in their treatments and take pride in what they do. They don’t like to imagine that a service they’ve been delivering might be useless or harmful. Add to that the fact that physicians get paid to do what they do, and new evidence threatens both their sense of being good doctors and their livelihoods.
Not surprisingly, when strong evidence is published showing a cherished procedure is ineffective, many physicians who perform that procedure initially express shock and disbelief. Then they come up with numerous, often unsubstantiated reasons to cast doubt on the results and continue business as usual.
When the ORBITA trial was published in December of last year, surgeons who perform stent procedures expressed their doubts in the journals and on social media. The study was too small, they claimed. The patients in it weren’t suffering from serious disease; there was a subgroup in the study that benefited. Some said their personal experience showed their patients benefit from stents. Others claimed to already know everything the trial showed.
Some surgeons argued that sham-controlled trials are unethical because we just know stenting works. At one point, one of the principal investigators of the trial, Darrell Francis, a British interventional cardiologist, wrote on Twitter, “I would like to thank the patients for participating in #ORBITA without realizing what a low level of intellect would be put into criticizing their efforts.”
The arguments clinicians raise against results they don’t like are often similar to those raised against ORBITA, regardless of the procedure. A prime example is the response spine surgeons and radiologists had to two high-quality, randomized trials published in 2009, showing that vertebroplasty is no better than sham surgery. Some faulted the trials for enrolling patients too early in the course of their condition. Others said patients were enrolled too late; patients were not in as much pain as their patients; the studies were too small, too big, and, unbelievably, “too rigorous.” In an editorial, The Spine Journal editor-in-chief, Eugene J. Carragee, called the reactions, “imperious.”
Imperious or not, what’s more astonishing is how willingly physicians take up new procedures and tests on the basis of the slimmest of evidence that they work. In the case of vertebroplasty, early reports claiming “immediate” and “complete pain relief” in the vast majority of patients should have raised at least some suspicion that the results were too good to be true. Instead, physicians made perfunctory remarks about the need for better evidence and then proceeded to perform hundreds of thousands of vertebroplasties, even as increasingly rigorous studies showed a decreasing size of the procedure’s effect. In 2008, there were more than 100,000 vertebroplasties performed in the US. Five years after the negative trials were published, that number had dropped only 25 percent.
This should be depressing news for anybody who thinks simply publishing new evidence is enough to ensure that patients get the right care. If we want physicians to heed evidence, we should probably consider multiple other approaches.
First, we should be doing the research before the procedure or test is widely used—and paid for, which tends to reinforce physicians’ attachment to it. It took a decade for vertebroplasty to be put to a real test, and that’s fast compared to how long it has taken to mount rigorous studies of other procedures.
Stents have been around for 40 years and we’re only now getting around to the definitive studies. If the ORBITA trial is confirmed by other studies, we will have spent hundreds of billions of dollars over the last four decades on a treatment that at best adds only the slimmest of additional benefit to optimal medical therapy for heart disease (drugs, lifestyle changes) and at worst is no better than a placebo. Carotid endarterectomy was performed for nearly 40 years before real studies were launched. The need to do pivotal studies early applies even more to procedures that relieve pain—like vertebroplasty and stenting—and psychiatric illnesses; these conditions are particularly susceptible to placebo effect, which can fool patients and doctors alike.
Second, isn’t it time to stop paying for inappropriate care? Physicians don’t like to contemplate this solution, and neither do insurers, because they know that doctors and patients will wave the rationing flag. But this is one incentive that actually reduces useless care. In 2014, both Canada and Australia stopped paying for Vitamin D testing, which was leading to high rates of prescriptions for Vitamin D for patients who did not need it. Australia’s Medicare program was spending $12 million a year on Vitamin D testing. In Canada, more than 25,000 tests a year were being performed. Once the authority stopped paying for it, testing in Canada dropped by 80 percent within six months.
Third, let’s stop using marketing language. Calling patients “consumers” who need to be “engaged” reinforces an attitude that says being a person who is sick, frightened, in pain, and in need of care is no different from a customer buying a used car. That means the doctor is merely a “provider” of a medical service and has a diminished duty to serve the patient’s needs first. One day, patients won’t trust doctors any more than they trust used car salesmen.
Finally, patients and families need unbiased, understandable information in order to make choices about treatments. We can’t leave the transmission of that information to their doctors, because many of them are not very good at making sure patients and families understand the tradeoffs involved in different treatment choices. Several studies have found that patients giving consent are largely uninformed about a wide variety of medical decisions. The reason is not that they can’t understand what’s at stake; it’s that doctors don’t convey information effectively.
Making sure patients understand their treatment choices has turned out to be a difficult problem to solve. For 30 years, proponents of shared decision making have been advocating the use of decision aids—booklets, video, and other forms of clear, simple information—to help patients understand treatment options, and then for clinicians to share the decision based in part on what the patient cares about. In my father’s case, he cared about not being harmed by a surgery that offered only the slimmest chance of benefit more than preventing a stroke that he might never have. But maybe shared decision making is a little like Brazil (which has been famously called “the country of the future” for last several decades): It would improve care dramatically, but its time has not yet come. Perhaps the only way to get shared decision making into common practice is for patients to start asking uncomfortable questions.
Failing to explain tradeoffs allows doctors to avoid thinking hard about the potential negative effect of their procedure on the patient and family. On paper, my father might have been an appropriate candidate for carotid endarterectomy, if marginally so. Until the surgeon was confronted with the unrealistic expectations of a worried family, he did not have to consider the increasing frailty of the man in front of him. Unfortunately, most families don’t have as a member someone who studies overuse of medical procedures for a living, and who can help ensure that the conversation includes the negatives of a treatment as well as the positives.