Dr. Joseph Kvedar, the vice president of Connected Health at Partners HealthCare, thinks the future of healthcare is in ever increasing levels of personalization and automation -- to the point where in just a few years, we might all have little automated health coaches in our smartphones, operating similarly to Siri, Cortana, or Google Now.
"This is quite doable," Kvedar said during his opening keynote at the Partners Connected Health Symposium in Boston. "It’s as if the puzzle pieces are there and we haven’t put the jigsaw puzzle together yet."
Those puzzle pieces are advanced sensors to deliver up-to-the-minute information about our health; systems for aggregating, analyzing, and normalizing that data; and a way to integrate it with the EHR and other information systems. Finally, this will all fail without a focus on engagement to make sure the patient actually uses and wants to use the services being offered to them.
Normalization is an important, missing piece to Kvedar because it allows a consumer to get a more complete picture of their health by meaningfully combining multiple data streams, something that by and large isn't happening yet.
"I right now have a Withings scale in my bathroom, I’m wearing a Fitbit, and if I’m meticulous using MyFitnessPal I can get a reasonable estimation of my calorie intake," he said. "Those three numbers are on my dashboard and they don’t relate to each other at all. They should tell a story. Calories in, calories out, weight: should all relate to each other. That’s what I mean by normalization."
Also important: getting the data capture to where it's really easy to use. Kvedar used the example of Proteus Digital Health, a company whose ingestible-sensor-based adherence system he recently tested.
"Frictionless data capture is a big deal," he said. "As much as I love my friends from Proteus, I spent a half hour onboarding that system the other night. They’ll never scale it if you have to have a half-hour for onboarding for every patient. We have to get it so you take it out of the box, slap it on, and the data just flows. That’s really important -- pairing, unpairing, re-pairing, that’s all got to go away. Get that right cause it’s not right now."
Kvedar compared healthcare to other industries, where it's been sufficient to use big data to make predictions about people's behaviors and needs based on the needs of similar people. He thinks that healthcare needs to focus more on things like sensors and personal data analytics, because people's health is too important to trust that it will follow generalized trends. He pointed out the example of pharma companies, per a 2013 Wall Street Journal article, using big data to recruit for obesity clinical trials.
"They’re drawing analogies like 'people who have a big cable package are more likely to be overweight, so we might want to recruit them. People who have a minivan but no children are more likely to be overweight.' I mean, how offensive is that stuff? ... That’s not going to fly. It’s pretty coarse. We have to get above the level of 'If people like Joe do X, Joe will do X' and we have to get it down to the individual level. Again, the things in the Internet of Things will help us do that."