There is a physical therapist practicing in a small rural town — the kind of town with one stoplight and a four-hour drive to the nearest academic medical center. This clinician is talented. They have been seeing the same population for seven years. They have watched patients cycle through the system — primary care, imaging, specialist referral, physical therapy, repeat — and they have started to notice a pattern they cannot fully articulate. Some of these patients are not difficult because their conditions are complex. They are difficult because the framework being applied to them is insufficient.
This clinician wants fellowship training. Not the certificate. The development. They want a mentor who can sit inside a case with them, ask the question they have not thought to ask, and help them understand what they are looking at when the presentation does not fit the pattern.
There is no fellowship mentor within three hundred miles. The only pathway available requires relocating for a year or leaving the community they have spent seven years building. So they stay. And they keep practicing at the edge of their framework, without anyone to help them see past it.
The profession has a structural assumption embedded so deeply that it rarely gets named directly: that proximity produces mentorship quality. That the mentor who sits across the room is doing something categorically different — and categorically better — than the mentor who appears on a screen. That the geography of the relationship is a proxy for its depth.
This assumption has consequences. It concentrates fellowship-level development infrastructure in urban academic centers, in regions with enough training density to make co-location feasible. It makes the developmental continuum — the residency-to-fellowship pathway that the evidence says is the only reliable route to adaptive expertise — a resource available primarily to clinicians who happen to live near a hub. And it quietly tells the clinician in the rural town that the development they are asking for is not something the profession has designed a pathway for.
The assumption deserves more scrutiny than it has received.
A 2025 randomized controlled trial in health professions education found no significant difference between remote and in-person mentorship conditions on trainee competency outcomes — when supervision quality was controlled. That last clause matters. The study did not find that geography is irrelevant. It found that geography is not the operative variable. Quality is. Structure is.
A qualitative case study examining synchronous e-mentoring in musculoskeletal physiotherapy found something counterintuitive: trainees in the remote condition reported deeper engagement in reasoning articulation than those in direct observation settings. The authors’ explanation was straightforward. When the mentor cannot see the patient, the trainee has to speak the reasoning out loud — completely, precisely, without the shorthand that shared physical space permits. The required explicitness turns out to be a feature, not a deficiency.
VA health system data comparing tele-rehabilitation and in-person clinical outcomes across a range of musculoskeletal conditions found equivalent patient outcomes, with the largest gains in access concentrated in rural and geographically isolated populations. When the researchers turned to the mentorship infrastructure implications, the argument was direct: treating geographic distribution as a quality compromise, rather than as a structural design question, reproduces access inequities that the developmental continuum is supposed to address.
What the research consistently shows is that poor outcomes in remote mentorship are not caused by remoteness. They are caused by the absence of structure — unscheduled contact, no standardized reasoning documentation, no shared entrustment language, mentors who were not trained for the specific demands of the remote context. Those failures are program design failures. They are solvable.
The model that the evidence supports has a name, even if the profession has not settled on one yet. Hybrid mentorship: synchronous live remote clinical reasoning sessions as the primary mentorship modality, structured reasoning documentation as the connective tissue between sessions, and in-person intensives for the domains that genuinely require physical presence.
Because some domains do require physical presence. Real-time observation of a trainee with a patient — watching the hesitations, the false starts, the moments when the examination sequence changes — is something that video cannot fully replicate. Psychomotor skill integration, the development of manual technique as an expression of clinical reasoning rather than a separate procedure, requires hands on a patient and a mentor who can see the whole interaction. Multi-rater faculty calibration of entrustment decisions benefits from everyone being in the same room at least some of the time.
The hybrid model is not designed to approximate in-person training with an inferior substitute. It is designed to deploy each modality where it performs best. Synchronous remote sessions for the cognitive and reasoning work — the case discussions, the hypothesis probing, the assumption exposure that is the heart of fellowship-level mentoring. In-person intensives for the embodied work that proximity uniquely enables. Not a compromise. A deliberate architecture.
The distinction matters because the profession’s instinct has been to treat the hybrid model as a concession — something you do when you cannot afford the real thing, or when geography leaves no alternative. That framing has it backward. The question is not whether the hybrid model is good enough. The question is whether programs are designing it with the rigor the evidence requires: structured session schedules, standardized documentation, shared entrustment language, mentor preparation for the specific cognitive demands of verbal-only case reasoning, and in-person intensives integrated at the points where they are actually necessary.
The physical therapist in the rural town is still there. Still seeing the cases that sit at the edge of their framework. Still without a mentor.
The pathway that could change that is not science fiction. The competency-based infrastructure, the EPA frameworks, the entrustment-calibrated mentoring architecture — these exist. The evidence on remote delivery, on equivalence outcomes, on what structure produces and what its absence costs — this exists too.
What does not exist is a profession-wide commitment to distributing the developmental infrastructure to the clinicians who need it, wherever they practice. The hybrid model is how that commitment becomes something more than a principle. It is how the clinician in the rural town gets the mentor they have earned the right to have.
Brett Windsor, PT, MPA, PhD Executive Director, NAIOMT
