Will AI scribe vendors actually replace EHRs?

The prevailing belief is that the end goal of ambient AI companies is to replace the EHR—otherwise, how could they justify billion dollar returns to VCs? Here’s where I think this will (and won’t) work:I suspect the core thesis stems from how everything else in the EHR is centred around the patient encounter.If you “own” the experience around how the patient encounter is captured, you can own everything downstream from that: billing, compliance, orders, etc.Just the other day a physician who told me: “my AI scribe captures all of my notes and does so much for me—I now spend almost no time using the rest of my EHR . . . so why am I paying so much for it?”Of course this was a physician speaking about his outpatient practice, and who hasn’t yet been able to convince his hospital to adopt an AI scribe.Read: Dr. Joshua Liu: Why Big Tech struggles to disrupt healthcareA physician (or a relatively small number of them) who own a practice can independently buy whatever software they want, which means not only buying an ambient AI solution in the first place . . . but eventually replacing the EHR too.In contrast, a large health system—centered around integrating inpatient and outpatient care—not only requires a much more sophisticated EHR, but also has far more layers of bureaucracy to prevent the incumbent EHR from being displaced.And let’s not be naive: the health system EHRs—Epic, Oracle, MEDITECH—already are or likely will be building their own AI assistants . . . starting with scribing and moving into everything else.This means AI assistants provided by the incumbent EHRs will continue to have a strong advantage with their distribution.Even if you build an incredible third party AI assistant that could integrate with the major health system EHRs . . . you can’t just turn it on and prove it. IT, governance and broader decision makers will go through their typical process (which means most will be denied entry).So what should we expect to see?The obvious (and likely) reality is the office-based EHRs are at the biggest risk of being displaced by the ambient AI vendors. If they aren’t aggressively building their own AI assistants . . . they are basically throwing in the towel but may not realize it yet.If you’re an ambient AI startup that wants the best shot at displacing some EHR . . . your best bet is to win the practice space.But the Epics, Oracles and MEDITECHs of the world have a much longer lead time at protecting their health system customer base because even if they are late to the game, many health systems will give their solutions a chance first.In fact, unlike the ambient AI vendors, the big EHRs can even afford to give their solution out “for free” (maybe not yet, but as cost of compute continues to decline, we may very well get there).Ultimately it’s a classic case of whether the startups will get distribution faster than the incumbent gets innovation.And by nature of the complexity and bureaucracy of health systems, they protect the distribution of the incumbent.What do you think?Dr. Joshua Liu is a Toronto physician and the co-founder/CEO of SeamlessMD. This item originally appeared on LinkedIN.

Will AI scribe vendors actually replace EHRs?

Register for free access or login.

Related Topics