Bulletproof Physical Therapy Decisions presents your decision as the ‘skilled’ component of the physical therapy encounter. And there are two types of decision you can make.
Physical therapy decisions are dichotomized into two domains to draw attention to insight decisions – decisions that occur in the blink of an eye and can’t always be explained by logic or supported by data. Physical therapists have, historically, based many treatment decisions on insight and many patients have benefited from these treatments.
Recent years have seen an explosion in physical therapy research into clinically relevant tests and measures like self-report, performance and classification measures.
Tests and measures that generate patient data by self-report, performance testing, classification and even ‘old school’ impairment-level measures are ‘obvious’ decision support tools and most are far better than what I had when I graduated with a 4-year Bachelors in Health Science in 1992.
Back then, you could explain your choice of a spinal traction treatment by pointing to the evidence from your examination:
- Signs of nerve root compression in the leg or arm such as
- altered sensation
- asymmetric reflexes
- segmental motor weakness
- unilateral muscle atrophy.
Insight decisions are difficult to explain, hard to teach and challenging to document – why not manipulate a lumbar patient with 4/5 positive predictor variables for manipulation? What is it about his presentation that gives me pause? Is the manipulation worth the risk?
This lumbar patient also had a positive sciatic tension test and a positive cross leg sciatic tension test – tests these are not predictor variables for any classification category, not even ‘contra-manipulation’. Yet, these tests are sensitive and specific for lumbar herniated nucleus pulposis. The patient also had a lateral shift – which is a predictor variable for a direction-specific classification.
So, I had some tests that clearly pointed out why I shouldn’t manipulate and some tests that indicated the patient belongs in a different classification but why did I bother to examine those items in the first place?
One of the promises of a classification approach to physical therapy is its reliance on a parsimonious data set of predictor variables – fewest data are better. Therapists can quickly gather data on complex patients and allocate their clinical resources (eg: time) to do the most good. Presumably ‘the most good’ means better patient satisfaction, lower costs or improved outcomes using standard self-report measures.
So, insight decisions are an important source of clinical wisdom that is difficult to quantify and write down. Nevertheless, we ought to listen to our insight and, more importantly, study how those decisions are made so we can improve our teaching and documentation.
Before I go on, I’d like to present an example of an insight puzzle.
This example is taken from Malcolm Gladwell’s bestselling book on decision making, Blink.
A man and his son are in a serious car accident.The doctor is the child’s mother.
The father is killed and the son is rushed to the emergency room.
Upon arrival, the emergency room doctor looks at the patient and gasps, “This child is my son!”
Who is the doctor?
Most of us assume that the emergency room doctor is a man. Not all doctors are men – of course. Since the emergency room doctors’ gender is not explicitly stated in the puzzle to get the right answer we need to use our insight.
The Problem with Insight
Bulletproof is about how physical therapists can make better decision, primarily with better data, but also through insight – and write down their decisions to show skill, progress and need.
However, the very act of writing down your thought impairs your insight into the problem!
Psychologist Jonathan Schooler studied people’s performance on insight puzzles by interrupting their work and asking them to describe their thought processes – how did they reach their conclusions? What steps did they take? Did they make mistakes and start over?
He then repeated the experiment but this time he asked the subject to stop and write how they arrived at their insights.
With both experiments Schooler found that his subjects were able to answer fewer insight problems (Schooler calls insight problems ‘non-reportable processes’) when they were interrupted and asked to describe, verbally or in writing, their thought processes.
Schooler’s control groups were also interrupted and asked to stop working on their insight problems for the exact amount of time the experimental groups were asked to describe their thought processes.
The experimental groups, both verbal and written, were able to solve 30% fewer insight problems than the control groups. Schooler call this effect verbal overshadowing 121.
Face recognition is another area of study by those, like Dr. Schooler, who study insight.
I’m going to ask you to picture the face of someone you know, but not personally, say a famous movie star like Brad Pitt. You probably have his face in your mind right now. You can see his eyes, his hair, his chin. But, how accurately could you describe him to another person? Would they recognize Brad Pitt from your description?
By verbalizing your insight into facial recognition you weaken your insight. Face recognition and insight word problems are just two areas where insight has been shown to be impaired by verbalizing and writing 121.
What Implications do Overshadowing Have for Physical Therapists?
If you are comfortable and conversant with the leading edge of physical therapy practice (as of the writing of this book) in areas like treatment-based classification (TBC), performance measurement, self-report measurement, electronic documentation, alternative payment systems and Medicare compliance then you should have no trouble generating a daily note and a chart that accurately describes your interventions, patient need, expected progress and your skill level, even your insight decisions.
In 2006, when I bought out my private practice partners I was only peripherally involved in these issues. I was very much dependant on clinical insight, pattern recognition, heuristics and ‘rules-of-thumb’ to make decisions. If you are like I was then you should take heed…
The future of physical therapy will become increasing dominated by the clinicians’ ability to generate meaningful numbers (data) that justify our skilled interventions.
What about insight?
So where does that leave insight?
What about the valuable contributions made by legions of clinicians who came before we had of all these new, powerful probabilistic and predictive tests?
What about the patient in the example I gave earlier who presented with 4/5 predictor variables for manipulation and negative imaging studies? Why didn’t I manipulate him?
I can’t say for sure (since I had to write down my findings in the chart) but I suspected that he had an acute, torn disc. I gave him a Straight Leg Raise Test and then, on the opposite leg, I gave him a Crossed Straight Leg Raise Test – both were positive. The combination of these two tests, in populations with suspected disc injuries, is both sensitive and specific for herniated nucleus pulposus. I called his doctor and asked for a Magnetic Resonance Image (MRI). The results were positive.
What if I had manipulated him? What if I had not listened to my insight?
Medicare Demands Data
The standard of care in physical therapy in the United States is increasingly being provided by doctors of physical therapy practicing evidence-based medicine using clinical algorithms, classification and all the other tools just mentioned.
The prudent physical therapist is wise to incorporate these tools into her clinical examination, daily notes and complete chart.
How then can we preserve our insight yet create complete, compliant charts?
Here are five strategies I use to create a Bulletproof decision for my daily note and my chart.
Take Measurements Daily
I was taught to take measurements weekly or at least monthly. Now, when a patient fails a performance test the test becomes a part of the plan of care. For example, the Sharpened Rhomberg Test is an evidence-based measure used to predict falls and future institutionalization in elderly people. The test is valid and reliable and responsive to your clinical interventions. Why not, instead of just doing the technique, write down the score daily? Then you can show need (based on future falls risk) and change (based on the Minimum Detectable Change (MDC)).
‘Trust but Verify’
Classification decision-making is based on a parsimonious data set – that is, no test is used unless it is predictive and only predictive tests are used. The danger of a parsimonious data set is that patients aren’t data points – patients can’t be fully described with five (or seven or ten) predictor variables. Treatment (or non-treatment) risk may not be fully encapsulated by predictor variables.
For example, lower back pain patients often suffer from complexities and co-morbid conditions such as obesity, a sedentary lifestyle, various diseases and fear of movement. These patient often benefit with interventions that emphasize ‘spine sparing’ techniques, abdominal bracing, spinal stabilization and activity modification to identify provocative and easing factors of the pain. Some clinicians might call this approach a ‘stabilization classification’ yet, anecdotally, I have noticed that many lower back pain sufferers are older, stiffer and present in various stages of acuity. In short, few older patients satisfy most of the predictor variables for the ‘stabilization classification’.
Just like the manipulation patient I described earlier I measure all the stabilization predictor variables and use the classification decision rule as one of my tools – I make the final decision.
Put Goals up Front
I used to write goals because I had to – as long as the patient felt better I felt like I had done a good job. I didn’t understand the motivating effect goals can have on patient satisfaction. I also didn’t understand how important meeting functional goals is for a Medicare audit.
Now, we put a paper copy of our goals list on the front of the chart to refer to every treatment session.
In my clinic we co-treat our patients. I routinely delegate treatment to Physical Therapy Assistants who implement my Plan of Care.
- Up-front goals are essential when flipping through a paper chart during a busy day.
- Up-front goals keep everyone focused on what’s important – function.
- Up front goals encourage the PTA to take daily measurements (see above).
- Up front goals send a message to the Medicare auditor when the chart is audited that goals are important.
- Insights happen when you least expect them – that’s what makes them insightful.
- Insights happen when you are talking or listening to your patient.
- Insights happen when you are reviewing patient data in the chart.
- Insights happen when you are laying in bed that night remembering what else you could have done for your patient.
Face time is good for your Medicare compliance program also – much time and talk have transpired at meetings, electronic list-serves and blogs about the one-on-one (constant attendance) requirements.
Here is an example of a recent answer to many of the questions about constant attendance asked in these forums:
“In the same 15-minute time period, one therapist may bill for more than one therapy service occurring in the same 15-minute time period where "supervised modalities" are defined by CPT as untimed and unattended -- not requiring the presence of the therapist (CPT codes 97010 - 97028). One or more supervised modalities may be billed in the same 15-minute time period with any other CPT code, timed or untimed, requiring constant attendance or direct one-on-one patient contact. However, any actual time the therapist uses to attend one-on-one to a patient receiving a supervised modality cannot be counted for any other service provided by the therapist.”Spending your clinical time face-to-face with your patient will improve your compliance and your insight.
Write for posterity
Assume you may come back to read your own notes – or worse, someone else will. I am amazed when I go back and read some of my notes from 10 years ago – not much data.
My notes from three years ago, however (when I started Bulletproof), contain valid, standardized tests and measures on impairments, performance and self-reports.
Many of the tests I use reveal persistent impairments in ROM, strength, physical function and important self-reported activities – I have monitored one of my running patients from initial presentation of ‘trochanteric bursitis’ to ‘I-T band syndrome’ to’ torn anterior hip labrum’ over eleven years – all the while coaching and treating him so he could continue to run his marathons. I also advise him to cross-train on the cycle and in the pool.
The point is that his numeric data are remarkably stable – his values for, say, hip internal rotation ROM don’t change all that much over eleven years. His stiff right hip from eleven years ago is still his stiff right hip today.
Physical therapy helps him maximize his activities and the data I generate help me make decisions.
The Bottom Line
There is still room for insight in physical therapy – actually, the need is even greater since patients are more complex, the workflow busier, the risk higher and the decisions harder.
Insight, however, needs to be the last piece of the puzzle, the bottom line of a page full of data.