Why do we use ineffective treatments?
There are many treatments used by physical therapists which do not have a robust (or even small) evidence base, yet practitioners swear by them. How do we explain this? Both therapists and patients are clearly perceiving a positive change, so what’s going on here? Why do we believe in ineffective treatments?
One explanation is yes, the existing literature has not yet adequately explored the intervention in question. This is certainly a possibility, and we should therefore continue our research efforts. But there are other explanations behind these treatments worth considering.
In this article we will discuss why many physical therapy patients tend to get better regardless of the treatment we provide (and sometimes despite our ineffective treatments), and why we tend to think it is our treatment of choice that made all the difference.
Why our patients get better
There are two fairly straightforward reasons why our patients get better independent of our treatment. One can be explained through biology, and the other through statistics. These two concepts aren’t unique to physical therapy, or even to medicine as a whole. The first refers to the natural history of disease and the second refers to the concept of regression towards the mean.
Simply put, natural history of disease refers to the progression and “natural” course of disease in a person over time in the absence of treatment. This process ends in one of three ways: recovery, disability, or death. This is an important area of study and is one of the reasons why we include a natural history group when conducting experiments.
In physical therapy, we often see patients with specific pathologies that are known, through extensive and well thought out studies, to have a poor natural history (it is highly unlikely that one is returning to sport without taking care of a ruptured ACL). We also have plenty of evidence supporting what we are able to do for these patients. There is little debate or confusion in these cases.
“If you treat a cold it will disappear in a week, but if you leave it alone it will last for seven days.”
But we also see quite a few patients without any specific identifiable pathology. Sometimes things hurt, and we can only really speculate why. Some of these patients will benefit from some sort of temporary load or lifestyle modification that will influence tissue homeostasis. However, for many other musculoskeletal aches and pains, the natural history is generally favorable and reverts towards baseline. That is, the patients will get better no matter what you do.
Regression towards the mean
Regression to the mean is a statistical concept which was first described in the late 19th century by Francis Galton when looking at the heights of parents and their children. The concept was later expanded upon by, among others, Daniel Kahneman when helping train Israeli fighter pilot instructors. For the purpose of this article, regression to the mean simply means that there will be natural up and down fluctuations in measurements around the mean of a group.
So what does natural history and regression to the mean look like in the clinic? Patients rarely present to physical therapy when they have a minor ache. It is only when symptoms have reached a certain severity that patients seek care. That is, when their symptoms have reached, or are close to, the ‘peak’ (B). For many musculoskeletal pains, symptoms come (A) and go (C) and naturally trend towards baseline after this peak. Recovery is coincident with treatment, not because of it. This belief that symptoms improve because of treatment is known as the regression fallacy.
Why do we think this way?
As mentioned previously, this problem is not unique to physical therapy, or to medicine as a whole. It is simply a feature of the human mind. Numerous authors have noticed and written about these exact phenomena in their respective fields, including psychiatry and chiropractic medicine.
It is natural to see cause and effect, especially when it is us delivering the treatment and when there are real objective changes in our patients. But it is important to recognize when we might be fooling ourselves and to understand why we tend to think this way. The list below is nowhere near exhaustive, but it does help shed some light on how our minds work.
1. We misperceive change when it does not occur and we misinterpret change when it does occur.
Our sensations and perceptions are extremely limited and don’t neatly reflect the world around us. We have a tendency to focus on what is most obvious in our environments and ignore more subtle background information. This characteristic (in a simplified sense) is known as naive realism, and leads us to confirming thoughts such as “seeing is believing” and “I saw the change with my own eyes.”
It seems reasonable to attribute changes in our patients to our sometimes ineffective treatments, especially when we are directly interacting with visible tissues and structures. But change following therapy is not the same as change because of therapy.
2. We forget about the times that the treatment did not work.
It is completely natural for us to have “favorites” when it comes to treatments, especially when we see our patients improving. However, this becomes a problem if we are presented with conflicting evidence regarding its use, and instead of engaging with it, we reject it and dig our heels in.
Confirmation bias leads us to seek out only evidence consistent with our beliefs while denying or ignoring evidence that is not. It also leads us to remember the times the treatment seemed to work, while forgetting about all of the times it did not. This leads to further overconfidence in an ineffective treatment.
3. We overestimate our ability to influence events.
We are more likely to believe we have control over a situation when we are 1) personally involved in the treatment, 2) familiar with the treatment, and 3) when we have a history of previous success with the treatment. We often believe we have more causal power than we really do, especially when interventions are consistently followed by improvement.
Why does this matter?
No one wants to acknowledge the idea that what we do doesn’t work, but there are both ethical and practical reasons why this discussion is necessary. With dwindling healthcare resources, we have an obligation to provide more efficient and effective care, and sometimes that means simply offering our patients reassurance and education about the course of their condition. This doesn’t mean simply saying “don't worry, you’ll be fine,” but it does mean that we need to be honest with our patients and ourselves about when treatment is warranted.