The Exam Was Never the Point
What practicing as a virtualist taught me about diagnostic reasoning — and why constraints reveal what comfort conceals.
Want to support Rewskidotcom Substack?
♻️ Share this post with friends
🫰 Become a paid subscriber
📢 Leave a comment belowThe Exam Was Never the Point
Practicing medicine without touching patients made me a better diagnostician. I wasn’t expecting to say that.
For a long time, I wouldn’t have said it out loud. The physical exam is one of those things in medicine you don’t question the way you don’t question morning rounds or the white coat. It signals effort. It signals rigor. It signals that you’re a real doctor doing real doctoring, not some corner-cutting algorithm in a button-down shirt.
But practicing at scale as a virtualist changed something in the way I think through a case. Not because virtual care is superior. Because it’s constrained. And constraints, if you sit with them long enough, start exposing what was actually holding things together.
What the History Does That the Exam Doesn’t
There’s a number clinicians know but rarely say in polite company: 70-80% of diagnoses are driven by the history alone. The physical exam shifts the picture less often than medical training implies. What it reliably does is confirm, document, and signal effort to everyone watching, including the patient, the attending, and the chart.
When I moved into virtual practice, I lost most of the exam — and was surprised how much survived the transition to video. General appearance, respiratory effort, affect, skin findings, the way someone moves when they think the visit is over. But what sharpened wasn’t what I could still observe. It was what I had to ask.
What I discovered, slowly and uncomfortably, is that I had been using the exam as punctuation more than as evidence. A period at the end of a clinical sentence I’d already written in my head.
Removing most of it didn’t make me guess more. It made me ask more. I had to get the history right because it was the primary tool I had. When you can’t default to “let me just take a listen,” you have to fill that space with something else. Usually, it’s a better question.
The Cover Story
Here’s the uncomfortable part for in-person practice: the physical exam can hide diagnostic weakness just as easily as it reveals it.
A physician who orders imaging after every ambiguous visit feels thorough. They examined the patient. They escalated appropriately. The rushed history, the narrow differential, the anchoring bias nobody caught because nobody asked: none of that is legible in the encounter note. The exam created a feeling of rigor that the reasoning didn’t actually earn.
Practicing as a virtualist strips that cover. When you can’t substitute physical presence for intellectual clarity, you either justify your reasoning in language or you have no reasoning. The diagnosis has to be narrated, not demonstrated. Over time, that narration becomes the discipline.
I’m not claiming every virtualist came out sharper. Plenty coasted on the same muscle memory, just over video. But the ones who leaned into the constraint, who treated the camera as a forcing function for better questions, were practicing more diagnostically honest medicine than they realized.
What COVID Actually Proved
Telehealth existed before 2020. The technology was not the barrier. The video infrastructure, the broadband, the platform capabilities: all of it was there. What changed in March 2020 wasn’t the tools. It was the will.
That’s worth sitting with. The restriction on virtual practice before COVID was a choice someone made, repeatedly, over years. Justified with language about clinical quality. Backed by the intuition that presence equals rigor. And then suspended in approximately two weeks when the alternative was seeing no patients at all.
What the pandemic proved isn’t that virtual care is safe. It’s that the previous restriction was a choice, not a necessity. And choices made to protect clinical quality often protect something else at the same time: geographic monopoly, billing justification, the institutional logic of the in-person visit as the atomic unit of care.
I’m not saying that’s why the rules existed. I’m saying it’s worth asking why they evaporated so quickly when pressure was applied from a direction that had nothing to do with clinical evidence.
The Pattern Library No Single Panel Builds
One more thing being a virtualist gave me: rare diagnoses become less rare when your denominator is national.
I’m licensed in all 50 states and DC practicing near full scope primary care at General Medicine. Seeing chief complaints across that entire geography — different health literacy levels, regional disease presentations, care access histories that vary wildly by zip code — builds a clinical intuition that a single patient panel, however deep, doesn’t produce. The virtual generalist catches signal the local specialist misses because selection bias has shaped everything the specialist has ever seen.
Scale is not the enemy of clinical depth. In the right architecture, it’s the accelerant.
And the discipline of going deeper into the history didn’t just make me a better clinician. It taught me how to think about designing AI decision support that actually works at the point of care. If the history is where diagnosis lives, that’s where the tool has to meet the clinician. Not after the visit in the documentation layer. During it, in the reasoning layer. That realization changed how I build.
The Question Worth Asking
If physical presence were truly essential to diagnostic accuracy, why do studies consistently show that adding five more minutes of history to an encounter improves diagnosis more than adding the exam to the history?
I practiced for years before I could ask that question without flinching. Now I can’t stop asking it.
The exam was never the point. The reasoning was always the point. Removing most of the exam just made that visible.
What’s your version of this? Did constraints in your practice ever reveal something your normal workflow was covering up?



