Entrepreneurs: There is Still an Opportunity in Medical Automation!

I read a fascinating article this week on the coronavirus and the challenges in the U.S. response to it. But one aspect that really surprised me was the difficulty researchers have in getting access to medical records to help track the virus.

“Our medical records are electronic in the sense that my grandfather’s radio was electronic” says a colleague of the author.

The typical clinician “spends an hour feed documentation into a computer for every hour they spend with patients” and the billion-dollar investment in electronic medical records pales before “the energized improvisatory medical role of Twitter.”

Before you decide to accept this challenge, though, remember the obstacles you will face:

HIPAA: medical privacy law that stops much ad hoc research in its tracks

Current sunk costs and techno-weariness: “we already spent the millions of dollars on this system and now you want to sell us something else?”

The inherent conservative nature of the medical profession. It took decades to get them changed over to the current clunky state of pseudo-electronic unsearchable medical records, but prying their fingers off the current system to try something new will be a challenge.

What would EMR 2.0 look like?

First, it would have to be more transparent than the current system. It should allow a physician’s spoken or written (scribbled, even) notes to be recognized, digitized and anonymized for research purposes in one near-real-time step.

It should feed AI learning systems that should be great at pulling out correlations of symptoms … this is the machine learning problem so often applied to commercial interests like figuring out what gets you to buy more coffee (or whatever).

It has to seamlessly work as a front end to the current market leader, EPIC. Will EPIC and company cooperate? I’d be surprised …

But please: let’s get going on this before the next pandemic!