The patient handed over the intake form, with the medication section mostly left blank. Just “a few things my doctor has me on.” The clinician — three years out of school, competent, thorough — asked the patient to list them verbally. Lisinopril. Metformin. Atorvastatin. Gabapentin. Omeprazole. A sleep aid, name not remembered. Something for anxiety, recently started. The clinician wrote them down, moved on to the shoulder examination, and began building a plan around the impingement findings that showed up on every test.
Fourteen visits later, the shoulder had not changed. The plan was defensible. The clinical reasoning was sound within the framework the clinician had been taught. The examination was textbook. What the clinician had not been taught — what the system had never asked anyone to teach — was that the medication list was not an administrative formality. It was a diagnostic artifact.
The atorvastatin was a statin. Statins cause myalgia and tendinopathy in a clinically significant subset of patients. The gabapentin was a gabapentinoid. Gabapentinoids cause cognitive dulling, fatigue, and dizziness — each of which alters a patient’s capacity to engage in the active rehabilitation that most shoulder protocols require. The sleep aid suggested a sleep problem that the intake form had not captured. The recently started anxiety medication suggested a psychological burden that the examination had not explored. Seven medications. Each one was a thread the clinician could have pulled. None of them pulled.
This is not a story about a clinician who failed. This is a story about a system that never built the expectation — in education, in clinical culture, in the structure of the evaluation itself — that the medication list would be read as clinical data rather than filed as paperwork.
The evidence underneath this story is not ambiguous.
The proportion of adults taking five or more medications has more than doubled in two decades — from 8.2% to 17.1% of all adults, with the figure among adults over sixty-five rising from 23.5% to 44.1% (NHANES data). This is not a niche geriatric concern. This is the caseload. The patients arriving in outpatient physical therapy clinics are more medicated than they have ever been, and the medications they take have direct, measurable effects on the tissues and systems clinicians are trained to evaluate.
Statins cause myalgia and tendinopathy. Fluoroquinolones cause tendon damage and neuropathy. Corticosteroids alter tissue healing at the cellular level. Gabapentinoids cause cognitive dulling that interferes with motor learning, balance, and the kind of active participation most rehabilitation protocols assume. These are not rare side effects in rare populations. These are common medications in the daily caseload, producing effects that land directly on the tissues and systems the clinician is trained to evaluate. The clinician who does not review the medication list is not cutting a corner. The clinician is missing information that directly shapes the clinical picture — and neither the education nor the system asked them to look.
But polypharmacy is only the pharmaceutical layer of a deeper shift. A 2025 CDC analysis of over 2.6 million respondents found that 93% of older adults, 78.4% of midlife adults, and 59.5% of young adults reported one or more chronic conditions. That last number is the one that should stop a clinical educator in their tracks. Young adults — the demographic that outpatient clinics have traditionally treated as straightforward. A seven-percentage-point increase among young adults over a single decade (CDC, 2025). The patients showing up at age twenty-eight with an ACL reconstruction are not the same patients who showed up at age twenty-eight a generation ago. They are more likely to carry a chronic condition, more likely to be medicated for it, and more likely to have a psychosocial burden that the standard post-operative protocol was never designed to accommodate.
And then the psychological layer. A 2025 JAMA Network Open meta-analysis of 376 studies across 50 countries found clinically significant depression among adults with chronic pain at approximately 39% and anxiety at 40% (JAMA Network Open, 2025). Not in a specialty pain clinic. Not in a behavioral health setting. Among adults with chronic pain, the population that fills outpatient rehabilitation schedules every morning. Four out of ten patients on the schedule carry a psychological burden that is invisible to the standard musculoskeletal examination and unaddressed by the standard musculoskeletal plan of care. The intake form does not screen for it. The evaluation template does not prompt for it. The plan of care has no mechanism to account for it. And the clinician — competent, earnest, working within the framework the system provided — has no reason to suspect it is there until the patient does not improve.
These are not separate problems. They converge. The patient with three chronic conditions is more likely to be on five medications. The patient on five medications is more likely to experience side effects that mimic or exacerbate the musculoskeletal complaint. The patient with chronic pain is more likely to carry depression and anxiety that alter pain processing, reduce engagement, and compress the window within which rehabilitation can produce meaningful change. The layers interact. They compound. And they do so within a reimbursement structure that allocates the same time and payment regardless of whether the patient carries one layer or five. The caseload is not getting simpler. It is getting more layered, and the layering is accelerating.
The system that trained the clinician in this story was built for a different caseload. It taught evaluation of the musculoskeletal system as though the musculoskeletal system existed in isolation — as though the shoulder could be examined without reading the medication list, as though the exercise prescription could be written without understanding sleep, as though the plan could be built without asking what else was happening in the patient’s life. That system was not wrong for the era in which it was designed. It is wrong now. And the distance between what it was designed for and what it is being asked to manage grows wider with every year of chronic disease data.
We do not yet have a standard clinical education model that teaches the medication list as a diagnostic artifact. The pharmacology courses in most DPT programs cover drug classifications and mechanisms of action. They do not teach clinicians to read the medication list as a window into the patient’s lived medical reality — to see the statin and ask about muscle pain, to see the gabapentinoid and ask about fatigue and sleep, to see the recent anxiolytic and ask what changed. We do not yet have a reimbursement system that recognizes the time required to integrate pharmaceutical, psychosocial, and comorbidity data into the clinical reasoning process. We do not yet have a triage mechanism that identifies, before the first visit, which patients carry this convergence and which do not.
The clinician in this story went home that evening, reviewed the chart, and saw fourteen visits of competent care that produced no change. The documentation was clean. The reasoning was defensible. Every visit note reflected what the system had taught the clinician to see — and none of it reflected what the system had never taught the clinician to ask. The patient was the one who absorbed the cost.
The medication list was on the first page.
Stay up to date on The Development Gap and subscribe to us on Susbtack!
