Esther Dyson is an angel investor focusing largely on health issues, chairman of EDventure Holdings and founder of the Health Initiative Coordinating Council (HICCup). Photo credit: Flickr user Bush 41 Library.
First, let me do what you should always do when looking at data or content anywhere: Ask where it came from and why you are seeing it. Who is showing it to you, and what do they want from you? In this case, my role is a trifecta: I will be one of the advisers who review the semifinalists for Knight News Challenge: Health; my new nonprofit startup, HICCup, plans to submit an entry; and I am writing this post as a presumed knowledgeable, objective outsider. The truth of that is for you to judge now that you know the facts, but yes, I will recuse myself from judging HICCup's entry or any other submission that competes with it in more than a general way.Related Links
"What's next in Knight News Challenge: Health" by Chris Sopher
"Bring your best ideas; deadline nears for News Challenge: Health" by Chris Sopher
"Knight News Challenge: Health opens with inspiration phase, additional prizes from collaborators" by Raina Kumra and John Bracken
"Announcing key collaborators and details of Knight News Challenge: Health" by John Bracken and Chris Barr
"Data essential to promoting healthy habits" by Nirav R. Shah
"Data provides a focus for community action" by Bryan Sivak
"News Challenge: Make APIs not apps, health CEO says" by Lucky Gunasekara
"How data-driven solutions can transform health" by Lexie Komisar
"California HealthCare Foundation: The data stops here" and "It takes a community to humanize health data" by Andy Krackov
"Pizza tracker versus patient tracker" by M. Bridget Duffy
It will be up to the other advisers to assess whether they think I'm biased. In fact, I may well be overly generous to the other applicants in compensation. That's one problem with data. A finger on the scales doesn't always press where you might expect.
But let's get to health data in particular. It's known for being really squishy: Other than life or death, and sometimes even then, there's usually room for argument: How bad is the condition? Are the cells precancerous? Even if the data are precise—215 mg/dl—you can argue what they mean for a particular individual: Is this above their norm? What should be their norm, given their particular combination of weight, diet, height, genetic makeup and other specifics, including the timing of the blood draw? There's complexity within a single individual, and also complexity within populations.
In the end, the purpose with community data, as with self-tracking or medical data, should be more than just collection and analysis. The data should be used to monitor and change things for the better, whether for an individual losing weight, for a hospital improving safety and patient outcomes (and reducing costs), or for a community making it easier and safer for its members to exercise.
But there are lots of problems with health data. Many populations are ignored; others are studied with an edge. One thing I noticed as a cosmonaut in training is that most people with “good”—i.e. extensive—health records are sick people. We simply don't examine most healthy people other than astronauts that closely. When it was discovered that I have Barrett syndrome—an unusual condition that can turn into esophageal cancer—it wasn't good news, but we don't really know how bad the news is either. It’s discovered in most people when there’s a reason for putting a tube down their throat, but there may be many others, symptomless like me, for whom it never becomes a problem. We just don't know. We know mostly about the progression for people who do have symptoms.
But data is really important for making medical decisions and also for making population health decisions. Although things are improving with the Accountable Care Act and different attitudes towards measurement, we as a country have spent billions of dollars on health care programs without much insight into their impact. Sometimes we didn’t bother to measure; other times, the impact was lost within a welter of other factors. (And sometimes, perhaps, the data didn’t prove the point that someone was trying to make so they were ignored.)
Compared to what we know about actual medical practices, data about health and everyday behavior is even scarcer. What do normal people—people who don't monitor their food intake with MyFitnessPal—eat? How much do the non-trackers—no Fitbits or FuelBands or Shines—move?
Looking for data to guide the activities of HICCup, I have found many health interventions, but few serious studies of their impact. Most of it is focused on inputs: What did the program do? How many people were involved? Health data on long-term impacts is scarce.
But I believe that's about to change, in part because of programs like News Challenge: Health. As an adviser, I hope to see lots of creative projects that will engage ordinary people in collecting and analyzing health-related data—not just medical data, but observations about the real word they live in. How do people's eating habits reflect the restaurants they live near? How does the weather affect their sleeping patterns? Could we encourage first-graders to watch the impact of diet and exercise on pet mice as an extracurricular exercise?
As for HICCup, our goal is to help communities foster radical improvements in their members’ health through a broad set of interacting health initiatives, and to keep meticulous records as they do so. Understanding what you're doing is the best way to make it possible for others to follow your example.
Knight News Challenge: Health will offer a share of more than $2 million for innovative ideas that harness information and data for the health of communities. To join the conversation, visit newschallenge.org, or participate in an upcoming discussion at our events around the country. Submissions for the challenge will be accepted from Sept. 3 through Sept. 17.