As a student and clinician, I’ve often heard about “evidence based practice”. I even had a class on the topic in PT school. Yet during clinical rotations, I almost never heard discussion of research evidence among clinicians. In fact, some clinicians disparagingly said “Yeah I don’t have time to read research”. Some clinics had monthly journal clubs where clinicians would meet, chat about an interesting paper over lunch, and then go back to treating patients. There appears to be a disconnect between the supposed importance of evidence and how PT’s actually practice.
To help address that disconnect I’m sharing Steven Kamper’s 2018 series (https://pubmed.ncbi.nlm.nih.gov/29852833/) on how to practically apply evidence to the clinic. I also include my own commentary in [brackets].
Why should PT’s use evidence?
-The public expects medical care to be based on science [Ideally. There will likely always be a market for “body workers”, faith-healing, and related practitioners of pseudoscience.]
-If PT wants to be considered a scientific, it must use evidence
[I would argue that being scientific is the ONLY way PT can differentiate itself from the pseudoscientific professions mentioned previously. If we aren’t scientific, we’ll have to compete by developing increasingly strange, “cutting edge” treatment methods.]
-Agencies like insurers and government are trending towards only paying for evidence based treatments
But why can’t we just rely on clinicians using experience?
As clinicians we are subject to biases. First, the confirmation bias. We overvalue information that supports our views, ignore/forget what contradicts our views, and interpret ambiguous information to support our views. Second, the recall bias. We tend to remember extreme events (like a patient who did really well or really badly). [For example, if a patient raved about how they were “cured” by dry needling, we will tend to remember that patient rather than those who had a mediocre response. A related issue is that some patients stop attending PT, but we tend to forget those or assume that they improved.] Third, in clinical observation there is a fundamental difference between change in outcome and treatment effect. That is a treatment’s theoretical effect may not be solely responsible for the patient’s improvement in symptoms.
Why might patients improve regardless of the treatment effect?
-Natural history. Many conditions resolve completely or mostly on their own, especially acute conditions. For example, after a mild ankle sprain a patient will see improvement simply due to the passage of time.
-Regression to the mean. Many conditions are episodic or fluctuating. Patients may visit us only at the peak of an episode, after which point the patient’s symptoms will mostly improve regardless of intervention, since the condition is episodic. For example, some types of back pain are episodic and have exacerbations that will abate, regardless of the intervention given.
-Placebo. Placebo effects are any effects that occur due to manipulation of patient expectations and/or conditioning. [For example, if a patient expects that a chiropractic adjustment will “fix” their back pain, they are likely to feel less pain simply because they expect the adjustment to work. The point is not to disregard any treatment that has a placebo element, but we must examine the research to try to determine how a treatment compares to placebo. This has real practical implications. For example, suppose research shows that a highly “scientific”, intricate kinesio-taping pattern is found to work no better than sham, placebo taping to reduce knee pain. This means that simply applying tape to a patient’s knee will likely give the same effect as using a highly “scientific”, intricate taping pattern. If that’s the case, is it really worth it to invest time and money into a course on taping? Or should we look for treatments that have more robust evidence? Now with the COVID-19 pandemic, is it ethical to tell a patient they “need” a treatment that is known to mostly be a placebo (putting their health at risk)? Is it ethical to knowingly sell a patient a placebo with lots of “scientific” explanations?].
-“Polite patients”. As clinicians we build relationships with patients and sometimes they don’t want to “disappoint” us, so they say they feel better. [I’ve noticed this in the clinic. As a patient, it is very awkward to tell your friendly PT that all their hard work didn’t help you. Even on questionnaires, many patients will state more improvement than has actually occurred.]
The next post will discuss how to identify evidence relevant for a real patient case. As always there are exceptions and nuance to this discussion. All constructive, good faith comments are welcome!