In this session from Techonomy 2011, in Tuscon, Ariz., Techonomy’s David Kirkpatrick talks to Craig Mundie, Chief Research and Strategy Officer at Microsoft, about how technology could reform the healthcare industry by giving healthcare providers access to huge amounts of data and statistics they currently don’t have.
Kirkpatrick: Given that you’ve spent so much time working in Washington and working with the President’s Council—and one of the things you worked on was healthcare, the big report that PCAST did on that, and I know you have very strong feelings about what would be possible to reform healthcare with technology, and yet you have some significant frustration that no matter what the will might be in government, in this case it’s just not going to happen. Tell me what you think.
Mundie: No, no, I think—in that particular report, we’re starting to actually see some progress.
Kirkpatrick: You think we are?
Mundie: Yes, but relative—but again, it’s sort of, as the last person was saying, it’s all about speed, right? When we wrote that report, Eric Schmidt from Google worked on it with us, and a number of other people from around the country. And in simple terms we said, “Hey, you should think about taking the technology of the scale internet as we know it,” and saying, “you should use those things and sort of repurpose them to solve the healthcare internet interoperability question.” And then we were quite prescriptive about how you could do that.
Kirkpatrick: Yes, you like to talk about internet scale solutions.
Mundie: Right.
Kirkpatrick: For something like healthcare.
Mundie: Right. And, you know, we say, “Hey, hard is it to do this adaptation?” And the answer is, well, if you really said that was strategic and you were just going to go get it done, our belief is that it can happen quite quickly.
Kirkpatrick: What could happen? How big of an impact?
Mundie: That you could take all the data that exists in the country in all the healthcare systems, and essentially make it completely exchangeable and interoperable and searchable, because in fact it’s really small compared to the amount of stuff we now index and search and find on the web every day, and we know now how to do metadata driven things. The financial institute, as a case in point—what is it, XBRL, the Expensible Business Reporting Language, was ginned up by the SEC to standard how the US would have public companies report in an electronic way, because every company had a different accounting system. And they went from a standing start to all US public companies were reporting that way in less than four years, and once they had done it, it turned out every other company followed suit, and now all the companies in the world can essentially electronically report. You can analyze an exchange, all of the detail financial—
Kirkpatrick: How big of an impact could it have on US healthcare, if this was done in healthcare?
Mundie: Well, I think it’s dramatic, and one of the reasons is one of the fundamental problems in US healthcare is we have a payment system that actually is perverse relative to making improvements.
Kirkpatrick: No, it was mentioned on the stage earlier today, yes.
Mundie: Right. But it turns out—
Kirkpatrick: But you don’t know what it costs, and you don’t care.
Mundie: Well, actually a different way to say it is you don’t care because it’s a third party paying on your behalf. And since you don’t have any visibility in that—and people have economic interests, both providing the service and providing the payments. It essentially is a system that has no braking component, nothing that holds it back. So costs just escalate forever. And of course the technology growth allows that to go farther and farther. So everybody knows the system is broke in the US, and everybody knows that you actually have to go to some kind of system where you want to pay based on outcomes.
But the thing that holds that back is to pay on outcomes you need to have two things. You have to have a total view of David, and your health condition and what you’re being treated for, and you have to be able to compare the provider of your current event with a lot of other people’s event. Because otherwise we don’t know whether we paid and got a good job or a bad job. So you have to have quality control. So it turns out that if you want to make this change to an alternative payment system, you can’t do it unless you can have that universal view of the data, both for comparative purposes, for pricing and quality control, and in terms of being able to say, “How efficacious was what we did to you?” So if you can’t get the data in a form that allows that, you can’t start to change the payment system. And so there’s all these little experiments that get done, but you can’t make any wholesale change because you don’t have the data to drive it. So it’s another one of these big data problems, but we are not really serious about aggregating the data.