The Trouble with Evidence-Based Medicine, Again
How to get doctors to follow guidelines
In Part I of this mini-series on medical guidelines, we discussed some of the challenges physicians face in following evidence-based medical guidelines. Now it’s time to face the much more difficult question: So what are we going to do about it?
But first, a seemingly counterintuitive question: Should we do anything about it? As a public health professional, I was trained to always ask this question. We are often so quick to take action that we forget one of the most powerful options available to all of us: doing nothing. In this case, the answer is “we should do something, but we have to be careful not to do too much.”
What do we mean by this? Critics of these guidelines and of the wider movement of evidence-based medicine contend that medicine is more of an art than a science and that strict adherence to guidelines takes all of the idiosyncrasy out of medicine, which is where a lot of the skill actually lies. In this view, overburdening physicians with guidelines takes away from their ability to use independent clinical judgment to make decisions in complex situations.
It is certainly true that medicine is not a pure “science” and that there is absolutely “art” involved. If it were that straightforward, there probably wouldn’t be a need for these guidelines in the first place and we certainly would not be writing this series of articles. The right course of action would be obvious to everyone 100% of the time and it would also work 100% of the time.
No one, not even the most vehement supporters of evidence-based medicine, actually believes this. This criticism of evidence-based medicine falls a bit flat because it’s not true that proponents of evidence-based medicine and guidelines actually believe that independent judgment should be taken out of medical practice. It is, however, absolutely the case that doctors make a lot of decisions that result in expensive and ineffective care for patients. So there is clearly a problem with eliminating standards and the way we are doing things now is clearly not working either.
That much is clear. What’s less clear is whether flooding the field with extensive guidelines is actually helpful. That’s what we mean by “do something but don’t do too much.” Simply handing out new guidelines will not correct some of the non-evidence-based decision-making we see in medicine today. So what will? Here are a few ideas.
Change incentive structures
Any economist will tell you, if you want to get someone to change his or her behavior, there has to be money involved. While this of course isn’t always true, money is certainly a powerful incentive and can be used effectively to change behaviors in certain situations. In the case of physicians making poor treatment decisions, what we often see in medicine is over-treatment. Whether it’s insisting on doing a mammogram every year from the age of 25, prescribing opioids, or doing back surgery, a lot of the decision-making problems we see in medicine involve unnecessary treatment that’s often quite expensive and can carry serious risks as well.
Is this pattern a coincidence? It doesn’t seem that way. In fact, most physicians are paid according to a structure called “fee-for-service.” This means that physicians get paid for each service they provide. The more “services” they provide, the more they get paid. So if a doctor just examines a patient and says take two aspirin and call me in the morning, that doctor will get paid significantly less than if he or she, say, orders an MRI. So there’s a clear incentive built in for doctors to perform procedures and order tests. In many cases, doctors may be influenced by this incentive without even realizing it.
What’s the solution to this problem? Well, we need to pay doctors in a different way obviously. Sounds easy, right? It turns out to be very difficult to come up with a good alternative to fee-for-service payments. More recently, some insurance providers have been trying out a “pay-for-performance” model, in which doctors are rewarded based on the quality of the care they provide rather than the volume of services. This of course opens a whole other can of worms about what constitutes “high-quality” medical care and who decides on that definition. So the jury is really still out on how to resolve this problem, but it is certainly the case that the fee-for-service model can lead to serious flaws in medical decision-making and that good alternatives are needed.
Training & Support
For many years, people thought that the problem of doctors not following evidence-based medical guidelines was a problem of lack of information. As a result, it is common to find new guidelines being accompanied by educational trainings to teach physicians about the new recommendations.
Not surprisingly, more recent findings have suggested that these educational trainings are not that effective at increasing physician uptake of new guidelines. The finding that educational strategies don’t always result in changed behaviors is not new. We know, for example, from recent experience with denial of scientific evidence among members of the general public that simply throwing more facts and information at people doesn’t work. And in fact, it sometimes results in a backfire effect in which people become even more entrenched in their original views. This is probably because people don’t like changing their minds and our brains are wired to help us hold on to our beliefs. As Tali Sharot, among others, has pointed out, this steadfastness is helpful overall but gets us into trouble when major changes in our environments or evolving knowledge of a subject should prompt us to be more flexible.
Some have also advocated for “technical solutions” that prompt doctors to follow new guidelines by allowing doctors to enter and share data to try to convince them of the efficacy of the new guidelines. This method is flawed for similar reasons cited above. In addition, if these data sets are sufficiently small or are inconsistently used or, worse, biased, then once again they will only come to confirm pre-existing beliefs rather than pushing physicians toward new ones, let alone changing their behaviors.
Now that we’ve determined that educational strategies might not work in this arena and that changing the incentive structure in medicine is necessary but not easy, where does that leave us? How do we ensure that more physicians are willing to follow new guidelines?
As you may know by now, we are developing a methodology to improve conversations about difficult topics related to denial of scientific evidence as part of our work here at Critica, stemming from some of the recommendations in Denying to the Grave.
One of the examples of our methodology in action you’ll find on our site details a scenario in which a patient asks her doctor for a C-section, even though in most cases C-sections are not medically indicated, increase the chances of other kinds of complications, are expensive, and have been shown to be over-utilized in this country. This example is relevant to the topic at hand here, because a significant part of the reason why doctors sometimes fail to follow updated medical guidelines is because patients put pressure on them to do something different.
For example, because of the very large breast cancer consumer movement, many American women have been extensively exposed to the message that mammograms save lives. While this is true, some of these consumer-facing organizations continue to aggressively spread the message that annual mammograms for women ages 40 and older is absolutely essential and life-saving. But rigorous population-level research has demonstrated that this is not in fact the case (to learn more about this debate, see part 1 of this series here). Nevertheless, for a variety of complex psychological reasons, people pay more attention to the consumer organization’s message advocating for frequent mammograms than to the official public health message that yearly mammograms are not only medically unnecessary but may even cause harm by resulting in false positives, which can result in unnecessary treatments that are both stressful and potentially risky.
The problem is that when these patients go to the doctor, they often insist on getting the screening. In many cases, the doctor feels pressured into doing the test when face-to-face with an individual who is distressed and insistent. Social norms and social psychology dictate compliance in these cases, especially when combined with the fact that many doctors are used to performing annual mammograms and it’s very difficult to change people’s minds, even when there are rigorous, persuasive data. We can communicate all these data to the doctors, but if we don’t equip them with the capabilities to have that difficult conversation with the patient and to empathically and gently lead the patient away from her emotional biases and toward evidence-based medical practice, the pressure to comply with the patient’s pleas will very often win out over the doctor’s desire to follow the evidence.
We are continuing to refine our methodology around these conversations and also around helping doctors to be more aware of their own biases and pitfalls and to prevent those biases from making decisions for them. We would love your help as we do this. Are you a healthcare provider? Have you ever felt pressured by patients to make treatment decisions that don’t align with the guidelines? Have you ever been reluctant to practice in accordance with new guidelines? We welcome all of you to share your stories and thoughts by commenting below or contacting us here.