Authored by Nate Felton and Brett Scott (DPT students at UMass Lowell taking my research methods and cardiopulmonary physical therapy courses in Spring 2015)
As entry level student clinicians, we have a large set of tools in our toolbox, yet we don’t have quite the experience to know which ones will work best. We are taught a great deal about individual diagnoses including the signs and symptoms that go along with them. However, people present with the signs and symptoms first, and we must abduct the cause of their problem. Many of the disease processes we know about also have overlap in their presentation. When creating an approach at this level, it is imperative to assess and evaluate using the knowledge given by the patient to think critically and come to a reasonable conclusion. Furthermore, recognising when the diagnosis you’re treating is getting better or worse should make you switch gears. It turns into a chess game, going back and forth with questions that help guide your evaluation and rule out extraneous tests, questions, and interventions. This implies knowing what tools to use from the information provided by the patient as well as your individual background knowledge of those tests. To become successful in diagnosis, it is necessary to be efficient with your evaluation and knowing what each test or question provides you with.
We’re in a bind: knowing all the different diagnoses we may see in a clinic is impossible, yet the way we’re taught is to learn a diagnosis and then the set of symptoms for each one. This is not the way patients present. Someone off the street will come in not knowing their diagnosis, and with physical therapy moving towards direct access, they may not have even seen a doctor before walking through our doors. They can only tell us about their experience, their symptoms. It’s our job to listen and make connections between their experience and the root of the problem. For example, someone comes in with chest pain. Based on this sole symptom, we cannot possibly make a diagnosis. Having a plan helps at the beginning certainly helps, but the plan will change throughout the conversation. We should be able to bounce between abductive and deductive inferences, thinking about the symptoms they describe, the signs we observe, and the outcomes of the select tests we perform to confirm and rule out possibilities.
It is no longer a matter of memorizing endless syndromes and throwing multiple diagnostic tests at the patient; the path forward lies in asking the right questions, listening to their responses, and knowing how certain tissue types (should) respond to testing and interventions. In the example mentioned above regarding chest pain, perhaps a good approach involves asking about the location and relationship of the signs and symptoms to activity, position, etc. to rule in or rule out potential causes. Once we have a good idea what anatomical/physiological system is/are in play, only then should we be selecting particular tests to examine further.
The most challenging aspect of practice is evaluating our treatments: are we just treating symptoms or the primary cause? How can we even know? In cardiopulmonary PT, we can see ECG, SpO2, MIP, HR and BP changes. In orthopedics, we can evaluate changes in ROM, MMT grades, and measurements of swelling. In neurological clinics, we see changes in balance, proprioception, and improvements in gait. Are any of these results from our interventions actually hitting the nail on the head? Are we getting to the cause? How can we tell? Short of cutting into a person to evaluate the tissue on a cellular level, we can’t really. What we can (and do) do is see how these changes we can monitor affect function. Using inference to the best explanation is the closest we have to finding causal relationships in PT. It is through these inferences that we guide our diagnosis, intervention, and prognosis. Understanding the basics and critically appraising research articles can give us more knowledge and tools with which to operate, and establishing a foundation of cause and effect relationships will guide us to a knowledge based practice of causal models.