KK: What about that other technology that you were demonstrating that was almost like a temporal artery blood pressure monitor?

ET: We’re testing that right now at Scripps. It’s called Scanadu Scout, and it’s about the size of a half dollar. You hold it up to your temple, and you get blood pressure, heart rate, and oxygen concentration in the blood. You could carry it in your pocket or your purse, and if you need to get all of your vital signs intermittently, it takes about 10 seconds. Who would’ve ever thought that that would be possible?

KK: Do you think this technology is better in the hands of a physician guiding care with their patient?

ET: At the end of the day, the consumers should be able to make that call. What I’ve learned is that with most patients who are worried about their heart rhythm, this electrocardiogram technology gives them a reading that is normal, and it’s very reassuring. It saves a lot of emergency care visits and urgent care visits. It should be the choice of the patient, but obviously, they have the ability to consult with their doctor and ask them.

KK: Let’s say someone has chest pain, and they decide to use this device to decide whether they should go to the emergency department. Let’s make it even better—say they have palpitations or they feel a rapid heartbeat, but it’s paroxysmal. They look at the rhythm strip, see nothing wrong, and think, “I don’t have to go to the emergency department.” Maybe they didn’t define the abnormal rhythm they had. Or let’s say it wasn’t the rhythm that was the primary problem. They have a pulmonary embolism, which is why they feel that sense of tachycardia. The strip read as normal, but their resting heart rate is 60 and their current heart rate is 90. From a relative standpoint, they’re tachycardic, but they just self-triaged themselves out of an emergency department. That worries me a little bit.

ET: Misdiagnosis is a big deal, but Kevin, I think you’re well aware we’ve got a little problem with that right now anyway. Twelve million Americans or more each year are getting a serious misdiagnosis. We’re working toward this ability to integrate multiscaled, multilayered information so that it wouldn’t just be one metric. You’d see the oxygen concentration, the SpO2, drop with the tachycardia, and you’d say, “I suspect a pulmonary embolism.” The point is that things are getting datafied, more objective, and there are these machines that can actually do a good job of processing the data and learning. I’m predicting over time that this could work pretty well. It isn’t the same as a doctor, but it’s processing a lot of objective data in a meaningful way. Someday it could be a kind of medical Turing test [test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human] where you have a computer doing a pretty darn good job in terms of accuracy. Diagnosis is different from treatment; that’s where the doctor has to come in, and the doctor would provide oversight that the diagnosis was correct.