Every year around the time of the health and government data extravaganza in Washington, Health DataPalooza, it’s reason to do an accounting of how far we’ve come in terms of accessing health data and using it as a foundation for value-based medicine. NPR says we have reached our “Awkward adolescence” (echoing Susannah Fox) with health data—lots of amazing things happening, but not a lot of impact.
Of course there’s plenty more work to do to make health data more accessible, more liquid and more private, but the progress since Health DataPalooza started less than 5 years ago is amazing, and we should take note, then come back to the paths forward.
This year, the big news was the FDA announced “OpenFDA”, available via an open API, with information on adverse drug events. Time will tell what the release will mean in terms of delivering health, but with over a thousand datasets now release by HHS alone, we are seeing a wave of new capability, even if data stories take time to tell.
Meanwhile, 80% of healthcare is now digitized, more than doubling from just a few years ago.
Samsung and Apple see the potential for accessing and harvesting health data and are moving into the fray to create personal health tracking hubs. There are many more examples showing that health data has, indeed, reached the limelight. Big health data, is now often called “the new oil”, and it’s already serving as a key resource in driving economics and powering countless new companies.
But that’s not what this post is about.
All this data is great, but it doesn’t take big data or rocket science to figure out what’s killing us, and how it might be prevented.
If you’d like to see what is killing us, check out this (small) data tool: http://vizhub.healthdata.org/gbd-cause-patterns/
Stop the presses: It’s us.
You take smoking, diabetes, obesity, cardiovascular disease and alcohol out of the mix, and the vast majority of those in the developed world would live to 90+. For data to really have an impact on health, it’ll have to have an impact on us. Many of these disease are diseases of behavior. We can debate how difficult it is to change behavior, or what biochemistry, genetics or other factors drive behavior, but most of our health problems could be prevented by making different choices. Consumers are going to need to care about it and use it.
There are bright spots that this is possible. Engagement rates reach 70% among institutions who do it well, but it takes leadership.
The reorganization of the ONC without a consumer office doesn’t show a lot of confidence that they are going to lead the way.
How do we fix this?
We’re nearing the point where we’ll be able to capture someone’s vital signs every minute of every day via Samsung, Apple, and many others. Will all this measurement save us from ourselves? Can we truly get prevention, or do will we just get better at heading off problems at the last minute? While preventing heart attacks is great, as a new iWatch is rumored to do, it would be even better if we could fix the unhealthy state that makes them possible before a last-minute intervention is necessary.
With that in mind, here are my wishes and a few predictions for the next phase of health care and health technology (now forever linked) and the road to solving health care with health data:
1. We need to create tools that can actually measure and impact behavior on what goes into us, not just stats on where we are and how we’re moving.
At the end of the day, we’re going to need to measure and provide feedback on input on intake as much as output. We’ll need to not only sense motion and vital signs but also what we’re putting into our bodies in terms of food, drink and chemicals, and start to change it. There has been work on tooth sensors to measure intake and Apple and others appear to be working on hydration sensors. It’ll be exciting to see developments in these areas in the coming years.
2. We need to better understand what drives metabolic disease. Metabolism-related killers are becoming our primary killers, but many normal weight people, in addition to obese people, die of metabolic disease. There’s still a lot we don’t understand about prevention and the disease. Yet metabolic disorders such as diabetes are taking an ever-greater toll and half the country will be at risk for diabetes by 2020. That’s a lot of suffering, a lot of death, and an enormous cost.
3. We need to prepare for the fight of a generation. Metabolic diseases are killing us in ever-larger numbers. The more we measure what’s driving costs, as we collect more and more Health Data, we’re going to run straight into a very big wall of conclusion: sugar is killing us.
With the release of FedUp, the idea of sugar as a culprit for our health care woes is starting to hit the mainstream. If the fight against control of tobacco was tough (and by no means won), the fight against sugar will be 10x harder.
4. We need to correlate outcomes and environment. That means we need to understand the network behavior of the health care system. We’ll learn a lot from the 125,000 people who die per year from not adhering to their medications. Why aren’t they taking them on time? What’s preventing people from treating themselves?
For that we need to understand things at a systems level and better correlate with the social determinants of health. As Atul Gawande pointed out, yet again, at health DataPalooza, the overall vulnerability of a population is what’s drives our biggest health costs. The intersection of socio-economic/social determinants and network behavior will help us solve major hotspots, major sources of cost and suffering.
5. This one might be obvious, but we need to be better at predicting with data. EHRs like their name implies, are records, focused on the past. We need electronic health systems that are predictive. Apple and Samsung or others will do it, and they appear to be correctly focused on a new kind of technology for the new business model of health care, focused on risk spread among all players (and value place on prediction).
Dave Chase, CEO of Avado, now part of WebMD, issued a stern warning to healthcare providers and their approach to healthIT on Susannah Fox’s blog:
“Just as it was easy to dismiss Google, craiglist, ebay, groupon, foursquare, facebook, etc. so too are the Iora Healths, Caremores, HealthCare Partners, Edison Health, One Medical, Surgery Center OK, Paladina Health, etc. ,but their value proposition is compelling. All of those players are deploying health IT in a radically different way than incumbents. Those orgs and their supporting technology take it for granted that patients are a core member of the care team, have access to their data and generally are using IT for competitive advantage.”
6. We need better rules on ownership and rights around health data, we could start with a Health Data Bill of Rights. In the consumer space, the rise of Snapchat and Whatsapp are indicative of a rise in the awareness and need for privacy. In health care, it will take time, but as health data gets “consumerized” with Apple and Samsung entering the fray, I predict the needs will become more and more apparent.
We need to work on rules and awareness to make health data more private and at the same time more easily exchanged. I don’t know exactly what that will look like but I, like many others, get the sense the answer may come through the blockchain. Fred Wilson at Union Square Ventures sees it as driving the next big investment cycle, after social and now mobile. He says, “our 2014 fund will be built during the blockchain cycle”. More on that in an upcoming post.
What do you think? What do we need to do to solve health care with health data?