The Trouble with Evidence-Based Medicine

Recently, the American College of Physicians released new guidelines for back pain. They weren’t guidelines most orthopedic surgeons are likely to be happy to read. If you have back pain, say the guidelines, you’re better off waiting it out with yoga or physical therapy than taking pills or, even worse, getting surgery.

While these guidelines might seem unconventional, they’re actually nothing new. In fact, the medical and public health fields have been trying (mostly unsuccessfully) for years to get doctors to stop prescribing opioid painkillers and performing complex, high-risk surgeries on people with chronic lower back pain. In fact, way back in 2013, Harvard researchers published a study in JAMA Internal Medicine based on 24,000 cases of spine problems between 1999 and 2010 showing that the vast majority of doctors ignored clinical guidelines on back pain treatment already in existence and continued to prescribe heavy-duty narcotics even though they really knew better.

So it’s not unreasonable to wonder what’s going on here. Is this a phenomenon that’s really specific to back pain and the doctors who treat it? Or are we looking at a more widespread, endemic problem in medicine?

The myriad ways in which doctors ignore medical guidelines

In some ways, that was really a leading question because we have a great deal of evidence that ignoring medical guidelines is a problem in medicine that goes far beyond treatment of chronic back pain. In fact, there are so many examples from so many medical fields of ignoring the guidelines that these days it seems like it’s almost more the exception to actually heed them. Let’s take a look at a few key examples.



By now you’ve probably all seen many headlines proclaiming the coming threat of completely ineffective antibiotics. And it’s true: antibiotic resistance is on the rise. There are a number of reasons for this, including extensive use of antibiotics in agricultural practice, limited development and availability of new antibiotics mainly due to lack of profit incentives for pharmaceutical companies, and poor regulation of antibiotic prescription and use in many countries around the world. But there is one cause of antibiotic resistance that in particular keeps us up at night mostly because it shouldn’t be happening, and certainly not at the rate it’s occurring: inappropriate prescribing of antibiotics by doctors.

The U.S. Centers for Disease Control and Prevention (CDC) estimates that roughly 30% of antibiotic prescriptions in the U.S. are unnecessary. That number is absolutely astounding – this means that for every 3 patients who walk out of the doctor’s office with a prescription for antibiotics, one of those prescriptions should not have been written. While there are certainly nuances and uncertainty involved in how a doctor determines whether an infection is viral (in which case it will not respond to antibiotics) or bacterial (in which case antibiotics are often called for), studies have shown that doctors are routinely prescribing antibiotics for symptoms and conditions for which the guidelines clearly state that antibiotics should not be prescribed under most circumstances.. These include ear infections, colds, viral sinusitis, and viral bronchitis. In all of these cases, there are clear, evidence-based medical guidelines indicating that doctors shouldn’t prescribe antibiotics,and yet they do it anyway. We will get to why this is in a moment, but for now it should be clear that not following medical guidelines in these cases is clearly wrong and also dangerous. Not only does it increase the risk of antibiotic-resistance infectious disease, it puts patients at risk for super-infections like the dreaded C. diff.

Screening for breast cancer

In 2009, the U.S. Preventive Services Task Force (USPSTF), an independent expert panel that advises the federal government on health, boldly flouted the recommendations of the American College of Radiology that all women over age 40 should undergo routine annual mammograms. The Task Force instead recommended that all women over the age of 50 undergo routine mammograms every other year. The Task Force faced swift and fierce criticism coming from a number of sources, including breast cancer patient advocacy groups and, notably, from physician organizations such as the American Cancer Society. The USPSTF guidelines were viewed as the latest in a broader government “scheme” to cut healthcare costs by allowing cold bureaucrats to make heartless decisions that affect people who have no say in the matter and presumably harm them.

But it turns out that aside from being costly and showing little evidence of making much of a difference in treatment outcomes for breast cancer, annual mammograms come with considerable risks as well. Over-diagnosis of breast cancer is rampant, frequently leading to unnecessary procedures and surgeries, many of which carry significant risk, high levels of anxiety for women and their families, and highly invasive treatment regimens such as chemotherapy and radiation, which involve serious and highly disruptive adverse effects. The American Cancer Society has since relaxed its guidelines slightly, but the American College of Radiology still recommends aggressive annual mammograms for all women over age 40 and many women continue to demand these screenings, often erroneously believing that frequent screenings will only help and not hurt them. And, just as is the case with inappropriate prescription of antibiotics, doctors tend to give in and comply with these demands. Ironically, this is all in the face of failing to urge women to take available preventive measures to defend against the condition that is most likely to kill them, coronary heart disease.


A recent excellent article in The Atlantic by David Epstein outlined some of the ways in which doctors fail to practice “evidence-based medicine.” One of the Epstein’s key examples isis the unnecessary use of stents, in some tragic cases resulting in a patient’s death. Indeed, as Epstein goes on to outline, doctors are often too quick to assume that a stent is the answer in cases where the evidence does not support that decision and may even contradict it. In recent years, it seems as though the problem may be getting slightly better, but a recent study of stent placements in a survey of about 1,600 hospitals across the country suggested that roughly 50% are not only unnecessary but also contraindicated.

Why do doctors ignore the guidelines?

So now that we’ve established that medical guidelines and evidence-based medical practices are ignored in a wide variety of medical situations, from relatively minor contexts such as the case of the patient who goes to the doctor with a cold and gets a prescription for antibiotics all the way to the very serious context in which a patient arrives at the emergency room with chest pain and trouble breathing and gets a stent implanted, we have to ask the obvious question: Why? Why are doctors ignoring the fruits of many carefully designed, meticulous, and expensive studies in favor of treatment options based on reasoning such as “because that’s the way I’ve always done it” or “because my patient really wanted it that way?”

There are some examinations of this question, but in our view, still probably not quite enough. Of course doctors are human and many of the reasons they make mistakes are the same reasons why the rest of us make mistakes. For example, doctors are of course prone to confirmation bias like the rest of us. Confirmation bias refers to our tendency to look for and pay attention only to information that confirms what we already believe. In that sense, we have an “attentional” bias, meaning that all pieces of information we receive are not treated in the same fashion. The pieces of information that fit our pre-existing views get magnified and the ones that don’t are ignored.

We can easily see confirmation bias at work in medicine. Vinay Prasad likes to talk about how doctors are quick to form an opinion but slow to change it. This is the basis for his excellent book, Ending Medical Reversals, which delves into the phenomenon of “medical reversals” and how when the overwhelming evidence forces medical opinion to change, most doctors are unable to change along with it. This notion of “sticking to your guns” is actually quite familiar for most of us. There are even brain imaging data showing that it’s pleasurable to maintain an opinion and fear-inducing to change it. In this sense, doctors are right there with the rest of us, using cognitive shortcuts and responding to neurobiological impulses.

There are other known reasons that doctors ignore evidence, like financial incentives  to do more tests and procedures in fee-for-service models or conflicts of interest involving pharmaceutical and medical device company payments that influence practice But these reasons, while all valid, are only part of the story. The bigger story, we believe, lies in an underlying system that has essentially completely neglected the human side of things and has therefore failed in a big way.

Allow us to elaborate. Medicine is a field of science, right? It follows a certain set of methods and makes decisions and recommends courses of action based on scientific evidence. And there’s a whole evidence-based movement in medicine right now, as if it was ever okay to practice “non-evidence” based medicine. 

While all of this is true (medicine is science-based, it follows a certain set of methods, and it’s focused on evidence), in emphasizing the objectivity and science base of medicine, we’ve left out to the critical human side. And by “human side,” we really mean the humanity of the healthcare provider, not the patient. In an effort to make medicine as “objective” as possible, we’ve left the human judgment, emotions, and needs of the physicians out of the equation. And this has really backfired. Medical students are trained to “be objective” and put their emotions aside. Of course, being a physician is difficult and you can’t be distracted or swayed by excessive emotions when you’re trying to make a clinical decision.

One of us remembers having to do spinal taps (lumbar punctures) on infants while a pediatric intern. The babies, who presented with high fever and possible meningitis, howled at the pain of the procedure and their parents winced. Emotion mandated forgoing sticking in the needle, but evidence mandated that not doing put the infant’s life in imminent danger. One had to learn to ignore the emotional reaction to a crying baby and do the tap.

On the other hand, when you tell people to put their emotions aside and continuously harp on how objective and evidence-based they are expected to be, a funny thing happens: they themselves are much less able to notice when their emotions and biases are making decisions for them.

The culture of medicine has also perpetuated this “physician-as-machine” myth. When do medical students or doctors ever have time to slow down and really think more deeply about the decisions they are making? The reality is that physicians are placed under an incredible amount of stress by the structural circumstances surrounding them, including hospital administration, insurance companies, and of course the legal system, and by the predominant culture of medical practice, which tends to be focused on extremely hard work, extremely long hours, and an incredible degree of personal sacrifice. The demands on doctors are almost unimaginable sometimes and to think that we can then expect them to be able to take a step back and think more carefully about their decisions when they haven’t slept in 36 hours is not reasonable. Yet not to do so makes the doctor prone to ignore his or her emotions and not realize that some decisions are based on biases that do not fit the evidence.

So what do we do now? Our first step of course is to recognize that, even with all the research cited above, most of us still operate under the assumption that physicians are objective machines who religiously consume and follow medical guidelines and scientific data. This is, of course, wrong and must be recognized. The next step will be to implement systemic changes that can to change the culture of medicine from one that relies upon people to be perfectly objective and unbiased in all circumstances to one that recognizes, appreciates, and makes concessions to the human side of healthcare delivery. That does not mean we want to encourage doctors to make decisions that are contrary to scientific evidence. Rather, it means we want them and us to acknowledge that doctors too are driven by emotions, emotions that must be recognized and sorted out in the process of deciding what a patient needs.

In the next piece in this series, we will lay out a number of solutions to this problem and begin to chart a way forward. In the meantime, we’d love to hear your thoughts as always. Are you a physician? Have you ever been reluctant to change your practice based on new guidelines? Do you notice this reluctance in your colleagues? Have you ever felt pressured by a patient to do something you didn’t think was medically indicated? What did you do? Comment below or contact us here.