The physical exam is a surprisingly uncommon critical piece in determining a diagnosis.

Most people without medical training drastically overestimate the sensitivity and specificity of the traditional physical exam as a means of diagnosis. In reality, for the vast majority of medical conditions, it is one minor data point that leads to a diagnosis. In fact, many, if not most, conditions have almost zero physical findings. Of course, some things are diagnosed exclusively by physical exam. But those are very few. And even fewer when that physical exam could be replaced by a photo, or occasionally, a video.

I would love to see, for each of the 1,500 common primary care diagnoses, a physician crowd-sourced weighting of how the diagnosis is made across the data all doctors use to diagnose a condition:

  • Asking specific, targeted questions
  • Physical exam
  • Lab tests
  • Imaging
  • Experimenting with a therapy and observing response over time
  • Observation the evolution of symptoms over time to eventually arrive at a diagnosis

This would allow all lay people to understand how likely a condition requires an office visit to diagnose.

Example: Ear infection

On a scale of 0 to 10, how do you weigh each as a contributor to an accurate diagnosis

Example: Diagnosing Diabetes

And for the few with a high physical exam weighting, ask if there’s a proxy. For example, acne. Traditionally it’s a “physical exam” in an office where a doctor visualizes your acne. But the obvious proxy here is a photo.

Here’s an example. Doctors, what do you think? Also, if you’re an academic and would like to work with me/Sherpaa on creating this data, let’s talk. As Virtual Primary Care evolves, patients need to wrap their head around when and why an office visit is needed based on how a doctor diagnoses conditions.