Sramana Mitra: Interesting. You prefaced the comparison with the existing EMRs versus yourself by saying that the old school EMRs were primarily billing systems whereas you are focusing more on the experience. One of the things that is viable in an electronic setting that is not viable in an in-person setting is recording of the session and then transcribing that session automatically. Is that part of what you do?
Bret Larsen: It’s an area we’re looking at and developing actively. When you look at what eVisit is, we’re a care platform. While EMRs are trying to compete in the virtual video element, we see them as partners. When we go in, integration with the EMR is a large component of the use case to make sure that we can push and pull that data back.
To your last question of being able to transcribe and automatically chart a visit, that is a big component of where we’re headed.
Sramana Mitra: There are probably two directions that we should pursue. If there is a lot more recording, then there’s a lot more data. There are opportunities for applying more artificial intelligence to the process. That’s one direction. The second direction is billing. We are part of the Stanford Healthcare System. Stanford has a whole bunch of facilities here.
One of the things that we often comment on is that the caregivers in our system all want to make appointments so that they can charge for that. Let’s say we go for a blood test. The results are coming out and the physicians cannot get paid unless they make an appointment to go over those test results.
Do I really need an appointment to go over those test results? AI can do the analysis and tell me what I need to know. There seems to be a bunch of incentive misalignment here.
Bret Larsen: There is. There always have been. Your point is something we’ve discovered and looked through. We believe that a geographically relevant provider-patient relationship is very important. The work that our clinical providers do is an integral part of the process. As you look at machine learning and artificial intelligence, the question we’re trying to answer is how do we use that data and the efficiencies that can be gained to support clinical decisions.
Providers are highly-trained and highly-educated statisticians. When you walk into your doctor’s office, they’re trying to narrow down, based on a number of data points, what it is that you’re experiencing. To be able to help supplement and support that decision-making process with data that we have is a key component for the future of care.
Sramana Mitra: There’s a lot more that can be done through e-visits. There’s a lot more that can be done automatically. The more interactions are recorded and processed by technology and machine learning, the more opportunities for inserting highly-competent algorithms to add value without depending on the doctor to make those decisions.
Bret Larsen: It’s about keeping the physician quite involved. There’s more to it that we need to explore. To be honest, most of them didn’t get into medicine to treat a lot of the things that can be taken care of through some of that data and algorithm work. They didn’t spend eight years in training to diagnose a sinus infection or a UTI; they got into it to help improve people’s lives. Being able to help them practice at the top of their license by bringing the right technology is key.
If you look at the thermometer, I believe it was discovered a hundred years before it became common practice. The physicians were still taking a patient’s temperature using the back of their hands. In healthcare, one of the interesting dynamics that we deal with is the mantra of, “Do no harm.” When you take unnecessary risks in care, people get hurt. It is about making sure we programmatically advance the next phases of technology in care to support the right patient outcomes and also the right clinical decision-making.