Diagnosis in Physical Therapy:

Clinical Thought Process

What’s the problem?

“My butt hurts.”

A significant proportion of my conversations start with this statement. She continued, “ I have regular active release along with dry needling. I stretch my hamstring ALL THE TIME and I still have pain.”

Her frustration was palpable.

 “I know I have a tight hamstring and there is scar tissue.  But it’s been 8 months.”


She assumes the pain must be the hamstring; we call this the pain generator.  The hamstring tendon attaches right where her pain starts. The tightness is over the hamstring muscle belly.  She describes the pain in a clear way that implicates the hamstring.  


She made a convincing argument that the hamstring is the issue and the diagnosis has been repeated by multiple medical providers including a physician and two physical therapists. The location matched.  Running increased the pain. Another match.Stretching and manual therapy provided temporary relief. 


But 8 months into the problem, the pain remains, limiting her activity and exercise while doing all the right treatment for a hamstring strain. In her mind, I needed to provide a different treatment for the hamstring to make the pain go away.  


I hate seeing people that feel like they lost months or even years of doing something they love.  The frustration she felt is due to an unclear process that is difficult for patients to understand with unclear timelines (see 3 Level Prognosis) and unclear diagnostic concepts.A poorly constructed diagnosis and a lack of testing to confirm the diagnosis with treatment and clear feedback loops are real, serious issues plaguing orthopedics. 


Let’s go back to the woman in my office.  Why didn’t the hamstring injury improve? The diagnosis was wrong. 


Why did it take 8 months to figure that out? The previous providers did not use a system to test their assumption that the hamstring was the pain generator.  This is a common diagnosis mistake in physical therapy. 


Any orthopedic problem requires digging into the history and developing a list of differentials (this means different potential pain generators including a chronic hamstring tendinopathy). The first phase of diagnosis requires a thoughtful inspection of the entire history. Snap judgements, uncontrolled biases, and poor heuristics derail the quality thinking needed to make a good diagnosis.  By following a simple series of steps, patients and providers can limit these issues.


What do we mean by diagnosis, why is so important in the physical therapy process, and why we think there should always be one competing with many?

Diagnosis, specifically medical diagnosis, ‘is the process of determining what pathology explains the signs and symptoms a patient exhibits.’  For our team, diagnosis is the first and most critical step of treatment. We create a mental model next. Mental models are recurring concepts that are used to improve our interactions and thinking in the real world. 

The diagnostic process provides numerous steps to mess up. In the example above, the original practitioner only considered one diagnosis then set up a treatment plan without feedback tools. 

I consider this a profound mistake because orthopedic diagnosis is difficult, prone to biases, and often wrong.  

The advantage of physical therapy practice in comparison to say an orthopedic physician is due to three reasons: the frequency of followup visits, the time during each visit, and the constrained diagnostic tools.  With the extra time, well timed follow up sessions, and the inability to ‘overtreat’, hypotheses can be tested. Literally, I test and retest up 20 or  30 times when building out my diagnostic conclusions.


The steps are simple.


1. What are the possible diagnoses from the subjective? 

2. What are the possible diagnoses from the objective?

3. What is my top diagnosis?

4. What is the physical therapy treatment that would most rapidly support the diagnosis?

5. Retest  – Did the treatment result in a 50% improvement in the most painful objective test?

Yes? Continue the same treatment and increase certainty in the top diagnosis

No? Attempt a different treatment for the top diagnosis and repeat steps 4 and 5

After 2-3 rapid treatments without improvement, the next diagnosis and treatment are attempted, repeating steps 3-5.



How does it help in our thought process?

I find it hard to write this, but I am often wrong with my original diagnosis.  I blunder on the first diagnosis because I am human. 

I mishear a portion of the story. 

I ask an unclear question.  

I do not interpret an exam test correctly. 

Patients tell me they feel tightness but they actually feel pain.

I look at all the questions asked and the exam performed, and none of them make a great picture

I hear a patient say they have pain while sitting.  Instead of asking a followup question to clarify between two possible diagnoses, I jump to the next question.


There are numerous ways to mess up. 


If you understand that diagnosis is complicated and mistake prone, then you never feel confident in your diagnosis. You become the person that constantly questions assumptions.  Diagnosis becomes a process within every physical therapy session instead of one time thing done at the start of a plan of care.  With all that can go wrong, it is a profound mistake to limit the differential diagnosis list to one pathology and compounded without repeated testing of the treatment followed by constant feedback.  Using these steps with a clear home plan sharpens the diagnosis until your differential list becomes one with the highest probability of explaining all the information you collected.



Steps in Diagnosis at Physical Therapy – How can you take an active role in your treatment?

Step 1: What are the possible diagnoses from the subjective? 

I started questioning how the pain started, when it is felt, and how much.  


What is the quality of the pain?


Does she feel it when the leg is swinging through while walking?


Is sitting the most problematic issue? What is she leaning back while sitting?


The questions irritated her at first. In her mind, she had a hamstring tendinopathy, but the questions started to shift the probabilities.  I finished the interview and my differential list included the hamstring, but I put it as the lowest probability diagnosis.  


Top 3 Differentials:

Discogenic low back with somatic referral

SI joint Lesion

Hamstring Tendinopathy


Step 2: What are the possible diagnoses from the objective?

I tested the back, SI joint, hip, and hamstring. I recreated her buttock pain with a slump test, indicating neural tension, and active back motions. I did a posterior to anterior pressure on her L4 and L5 spinous process that recreated the buttock pain as well.  The probability of the low back causing the pain went much higher.


I resisted the hamstring muscle – no pain. 


I stress the SI joint – no pain.


Top Differentials:

L5-S1 Discogenic low back with somatic referral

L4-5 Discogenic low back pain with somatic referral

Clinical Lumbar Instability


Step 3: What is my top diagnosis?

At this point, I felt 90% certain that the injury was originating from her low back.  I was not sure which level was most responsible for the pain, but I could provide treatment at this point to confirm.


Step 4: What is the treatment that would most rapidly support the diagnosis?

I decided to start with gentle lumbar traction.


Retest  – Did the treatment result in a 50% improvement in the most painful objective test?

I tractioned her low back and her pain in the buttock reduced with retesting of the painful exam.  By the end of the visit, her buttock pain was 80% reduced with sitting.  I adjusted her home plan to stop hamstring stretching and to lie down 2-3 times per day for 10 -15 min.  At the second visit, she no longer had pain unless she sat for over 2 straight hours. We made sure she sat no longer than an hour forty-five. That ended up being the last visit.  She canceled the last visit because her pain resolved. 

Remember, diagnosis is complicated.  The problem requires interviewing, testing, assumptions, and feedback in a big loop. In the example, the butt pain mimicked a hamstring injury.  The treatment for 8 months never targeted the actual injury.  If you hear a clinician say, I know exactly what that is after a few seconds of conversation, you should be worried.  All clinicians develop biases in their practice and use heuristics, rule of thumb, to make decisions. Without a system, diagnosis is a crapshoot with terrible results. As a patient you can ask and use steps above to help with your treatment.


What if you do not understand your diagnosis and the other possible differentials?

When I go get my car fixed, the mechanic is basically speaking another language.  He could say anything is wrong and I would not know enough to disagree.  

The same power differential in knowledge occurs in the medical setting. The most important reason to follow the steps outlined above is to create a system that gets you back on more equal footing.  With a plan and timelines, months may pass with no progress with no accountability. 

If you use the SPC Diagnostic Process, it helps with the following issues:


Frustration at lack of results

Accountability for how long the problem would take

How you can improve your own outcome

Poor communication between you and the provider: Reduced compliance with the home plan and More visits and time to get better




We need to  think about musculoskeletal injuries as probabilities, not definite conclusions.  When our team explains this process, initially there is concern that we stink at our job. We disagree. The more you know about musculoskeletal diagnosis, the less confident you should be.   

Working through the steps  as seen above, lets us make clear decisions on what the problem is.  We circle through over and over again during a single session.  Every trip around the circle, the probabilities around a single diagnosis start to rise.

The actions during a session become tools increasing the probability of the main medical diagnosis.  The treatment becomes diagnostic, not just therapeutic.  We search for a 50% improvement in the first session because of this process.  A 50% improvement in the first visit means we have a good grasp on the diagnosis. 

 No improvement? Then our confidence in the diagnosis is low and we say that to the patient.

As a patient, you may want concrete, undeniable diagnosis with a clinician that seems confident.  The confident clinician that never makes a mistake in diagnosis.  That person does not exist.  If the clinician never waivers from any injury, then there is a high probability there is no learning occurring and you will end up with butt pain for 8 months.

There is another way to do it.  Think of diagnosis as a process with constant revisions.  Picture a circle that constantly moves with a probability counter increasing for one thing and reducing for others until the symptoms are gone.


1684 S Research Loop
Suite 508 and 518
Tucson, AZ 85710