Why Big Data in Healthcare is Failing, and What Can Be Done About It
A few days ago a colleague of mine was inching south through the mother of all traffic jams: 60 straight miles of construction work on I-95 just south Washington DC. The three-lane highway was jammed. Route 1, which runs parallel to I-95 was also jammed. Cars were stalled in the middle of the highway having run out of gas from waiting so long.
He looked at the map on his phone and there seemed to be no escape, just miles and miles of thick, dark red lines symbolizing Google’s traffic data along the only two routes south. In roughly three and a half hours, he went roughly 50 miles. The jumbled, wasteful, clogged highway wasn’t altogether different from the state of America’s healthcare system.
Then, something sort of amazing happened. He’d exited the highway to Route 1 (for the second time), when Google Maps recalculated the route. If he continued straight, Google told him, it would take another 55 minutes to get through the traffic. But if he took the next left, it would be only 10 minutes.
Google wasn’t wrong. He weaved the car down a narrow Virginia country road, down a steep, wooded hill, then up a steep hill and through a homeowners association development, speed bumps and all, then around another corner to meet up with Route 1 again. He’d bypassed about 5 miles of road and about 45 minutes of traffic.
It’s too bad healthcare doesn’t work like this.
It’s not just the elegant merging of smart direction-finding software with satellite and crowd-sourced data on the roads and traffic that was impressive. It was that a single company, Google, could bring it all together, and then get it onto the user’s phone with a native, user-friendly app.
Healthcare doesn’t just need this sort of innovation, it needs a company like Google or Apple to bring it all together. As Imran Qureshi recently wrote in Wired, getting “Big Data” into the hands of physicians isn’t enough. It’s not nearly enough. Qureshi wrote that tomorrow’s innovation will merge Big Data with usable workflows, the kind that are actually usable in the real world of practicing medicine.
What’s needed is the app that automatically and seamlessly helps you avoid 45 minutes of traffic in the palm of your hand.
It’s no secret that the use of information technology in healthcare is years behind, maybe even a decade behind what other industries have been able to do. One reason is silos.
In healthcare right now information is siloed almost beyond belief. First you have the EHRs. About the best they’ve managed to do is throw dozens or even hundreds of pages of data at doctors, little of it useful, much less integrated into meaningful workflows.
Hospitals silo their own data aside from the EHR. States also have their own data silos, in the form of health information exchanges. These you have to log in to separately on a case by case basis – again not a good workflow for patient care.
Another silo of huge troves of data is genomic, and the ability to quickly search, track genetic markers, and build risk profiles for people. Mastering genomics and integrating it into healthcare delivery opens up doors that are almost beyond comprehension.
Then you’ve got the consumer industry: the wearables, the gadgets, the fitbits, and the hundreds of startups that are cashing in on the power of mobile technology. These flashy toys collect their own proprietary data, which is rarely integrated or shared with the larger care delivery system.
Finally, you have the data that is already being carried around by nearly every smartphone user: geo-location data about where you go, how you get there, and other information that could be of great use to healthcare providers. Did you go to McDonald’s three times today? Did you go to the pharmacy to pick up your prescription? Do you walk to work every day? Your smartphone knows these things, but the data isn’t being meaningfully used in a way that truly helps people.
The other problem with information technology in healthcare, aside from silos, is the companies themselves. Either their proprietary systems are designed specifically not to be interoperable, or they are designed to be interoperable but the company is too small for that to matter.
There are a series of large incumbents which have dominated information technology in healthcare for a long time. These huge companies are almost universally reviled by healthcare providers for their lack of user friendliness, their lack of interoperability, and their general failure to provide healthcare providers with meaningful ways to actually improve care.
Meanwhile, many smaller startups actually are doing very smart thing with mobile technology and healthcare, but they are too small to be able to integrate with the larger systems.
Companies like Google and Apple could make a huge difference in bridging this divide. They know how to own the market, they know how to leverage useful partnerships with other large incumbents, and they know how to create software that actually works for people. Actually, it could be argued that only these two companies can bring together the unique characteristics needed to finally, once and for all transform the use of information technology in healthcare. (Ok, maybe we could add Microsoft to that very small list).
This Summer Apple announced it would be trying to do just that with its new Healthkit app (“Could it change healthcare?”). They’ve partnered with Epic Systems, the Mayo Clinic, and other health systems as part of its rollout. The app will come standard on the iPhone. Dan Diamond wrote in Forbes that this “could be transformative for mobile health strategy.” Let’s hope so. Otherwise, no amount of Big Data will save healthcare from resembling the tangled, time-wasting, gas-burning, soul-destroying traffic jam that it does today.