Friday I finished another grueling 27-hour coding marathon at Cajun Code Fest in Lafayette, LA. Last year my team won the grand prize of $25,000 (along with bragging rights, mine and Betty's picture in the paper, and a "Mom's Favorite" award for a year), but this year we weren't so lucky.
You can't win all the time. Or maybe we could have won, had I only sabotaged the other teams' computers. Damn my sportsmanship!
This year's theme was "Own Your Own Health." That's a pretty general theme, and my team - consisting of me, my coworker John Harvey, Chris Burriss (my teammate from last year) and Chris's fiance Tara - couldn't decide what we wanted to do. We literally discussed this topic for three weeks straight.
During this process I found all the best food and drinks in the CC's and Starbuck's surrounding my house. Did you know that they have elephant ears at the Starbuck's in the Barnes and Nobles at Perkins Rowe?! They also have pizza pretzels. I think I single-handedly supported the Starbucks baristas for three weeks.
We had a thousand ideas, some good, some bad, and a few that we thought were great. We sketched out information exchanges, patient portals, phone apps, fitness and diet wizards, artificial intelligence for health apps, and the APIs to connect everything in the world... you name it, we thought of it and fought about it. Probably more than once.
To keep things light, I kept throwing out the idea of a phone app that would find the two laziest people in a certain vicinity, alert TMZ, send out notifications on Facebook and Twitter to all their friends, and then watch the two unlucky contestants fight for the right to not be Supreme Unhealthiest Person Alive. It would be like an instant reality show. I suggested it so often that now I'm thinking of really making that app.
But my personal two favorite pitches were as follows:
1. I wanted to make an abstracted health score based on health data. People could upload their health data into Microsoft HealthVault, which is a platform to store health data, which over 200 devices can write to. Then we'd create scores based on facets of your health - like say, Strength, Dexterity, Stamina, etc. You might recognize those as Dungeons and Dragon-type scores, and you're right - that's what I was going for. We'd create the algorithm to devise the health score, then we'd create an API to use it in a game, and then we'd try to have some Microsoft XBox Kinect game developers incorporate it into their titles. Why not just have Kinect see your movements? Why not have it also let you - all of you, even your health - be a part of the game? Maybe you can't storm Normandy as fast in Call of Duty if you don't have good health, or maybe you unlock secret items if you hit certain health milestones. I pitched this idea to Sean Nolan, the chief architect of Microsoft HealthVault, and he thought it was an interesting idea, although he could have been just humoring me; and
2. I wanted to make a Disney PhotoPass of Health. You get a receipt from a doctor, it has a code, you enter that code into an account online, and BOOM you see your data. Link your health insurance ID to your new HealthPass account (as I call it) and then you automagically get anything that goes through your insurance. Want a copy of your health records? We'd have them. The best part is that you wouldn't have to already be a part of the system to end up getting the data - just enter the code that you got from the doctor at whatever date and time is convenient for you, and then you'd have your info.
We ended up not doing either. Instead, we came up with the idea of putting the patient's health data under source control. For you non-software people out there, source control is what software people use to keep track of who checked in software files that make up an application. For example, I'm currently working on a project at work that involves 30 people in-house and 60 people offshore. That's 90 people working on the same set of files. How do we keep from waiting on other people to finish with a file before we check in, or make sure that we're not overwriting other people's files, or stop ourselves from killing each other? Source control.
We reasoned that if 90 people working on the same project from all over the world can manage to work with the same set of data at the same time with very few issues, then surely my primary care physician, ENT, allergist and radiologist can all work with the same set of data with no issues. Whenever I'd go to the doctor, I wouldn't have to tell them what medications I'm taking or what drugs I'm allergic to - that information would already be accessible to them. They'd be able to maintain their own version of my patient chart, but could sync it up with my patient chart under source control. We'd write APIs for other software developers to include in their apps, so that they could easily (and cheaply) interact with this unified patient record.
I write billing software for a healthcare company, and so I know that your doctor gets paid for every visit you make to him - but not for long. We're moving to an outcomes-based payment system where doctors will get paid more for treating you better. Why keep paying a doctor who makes you continuously go to him to get treated? Pay doctors more to treat you better. Doctors get paid per visit, and that's why you get double-booked and have to wait in the waiting area for an hour. Having access to a unified patient record would ultimately allow that doctor to provide better care for you because he or she would have a better picture of your health, which would then allow that doctor to make more money in the outcomes-based reimbursement system.
It's all about the Benjamins, baby.
It would also help with insurance costs. Insurance would ultimately pay less because patients would be visiting their doctors less frequently, because they'd be healthier. I talked with a health insurance broker at the conference, and she liked our idea so much that she gave me her number and wants my team to help her figure out if we can actually integrate our idea with some of her clients.
A lot of other teams at Cajun Code Fest worked the analytic edge, saying that they could provide information for scientists, governments or patients to show how their health fit in with the grand scheme of other patients, using Big Data analytic approaches. We included that in our app as well. But we thought that giving patients advice on how to improve their health is risky if you don't have the full picture of the patient's health record, and so that's the issue that we were trying to solve.
We liked our idea, and I'm very proud of my team. We didn't win, but we made it into the top 6 out of 17 teams. I think we got 5th or 4th. I also got free crawfish, which was the whole reason for signing up in the first place, so I'm happy.
The first place winners created an application to track a patient's medicine. I really liked this app. You could put your phone up to a bottle of medicine, and it would mark that you had taken that med. You could also print out a bar code or something similar to put on your weekly medicine pill box, in case you take a lot of pills and get them organized for an entire week. Last, if you didn't take your medicine, it could alert someone so that they could check on you. I have some friends who take a lot of medicine for various reasons, and when I told them about that app, their eyes lit up.
I also liked that application from the approach of a provider. My grandparents are in retirement homes, so they are given their medication by a nurse. How cool would it be for a home health nurse or nurse in a retirement home to scan in when a patient took a med, and then that person's family could look at the health record online.
There was a student group who created an app to track exercises - not just by miles run or how many times you exercised, but by intensity. I thought it was a great looking application, and was really impressed by the amount of information that the app provided.
Thanks to the great people at Cajun Code Fest for putting on another wonderful event. I'm always amazed at how many incredibly talented people we have in this area, and I always feel re-energized after just being around them.
We'll see you next year.
Thunderbolting Your Video Card
4 weeks ago