About two weeks ago, I got a generous invite from Stefan Weitz over at Bing to be the product guy for their first ever TEDActive Hackathon. The format was up in the air for awhile, and eventually settled on the format I prefer (which could only loosely be called a Hackathon): a single team (me, James Cox from Smokeclouds, Drew Volpe from Locately, Alan Cannistraro from Apple, student Rafael Cosman, and Zach Szukala from Seso) solving the singular problem of helping kids with asthma, as proposed by TED Fellow Sandeep Sood from Monsoon Company.
So, kids with asthma. Huge problem (25% of all ER visits, tons of deaths, and sucky quality of life for lots of people). The team hit a little hiccup to begin with as people all wanted to collaborate on the ideation phase. Which is great, as long as we don’t hold that all ideas are created equal: the whole reason Product with a capital P exists is because some people have spent a long time figuring out how to do this right. That said, it was a volunteer hackathon, so making everyone feel included was as much a part of the goal as saving kids with asthma.
The largest effect of that was the core app had to be built around SMS. Which is tough: if the goal is to get kids to use something, you want something with the strongest possible promoting pressure, which generally means access to as much rich content as you can get. That said, SMS means that it is cheap to deploy around the world and widens the audience, so even if it lowers engagement within a particular population, hopefully the increase in population overall makes up for it.
Which lead me to defining the problem. The first place people went was about preventing asthma attacks by avoiding triggers, which I think is a logical first step. Problem is, there are already apps that do journaling and after consulting with some epidemiologists (special thanks to Kate Weinberger from Columbia), it turns out that the causes of asthma are so manifold that trying to determine the exact cause of different kids asthma was difficult. And moreover, even if we knew the triggers, we had no way of using big data to warn them about them: pollen could be literally one tree, five city blocks away, that only stirred things up when the wind was blowing the right direction.
So I pushed for a product that ignored the actual triggering sources and aimed to do three things for all kids:
a) prevent attacks (which when you are targeting kids in general means breathing and body exercises that increase lung capacity and core muscle groups)
b) reduce severity of attacks before they happen (with similar exercises as preventing attacks, but focusing more on the forcing of air in and out of the lungs when constricted)
c) reduce severity of attacks while they are happening (which means staying calm, using the inhaler and breathing together, and knowing when to call for help)
Trouble is, apparently pulmonologists and scientists like to squabble about exactly how you do those three things. Given the severity of the problem and the number of people studying the problem, I was a little surprised that so little had been done in this area. They couldn’t even tell me if it was better to rest or exercise after you’ve had an attack and recovered!
Like every problem, however, there is some opportunity in there. Assuming we could get widespread usage of our apps, we’d actually generate the data that would answer those questions, which makes for an unofficial d) prevent attacks through research.
So having figured out a long list of stretches and breathing exercises we needed to get people to do, I now had to figure out a way to get people to do them via text. Enter the tamagotchi. We ended up giving every kid a “village” and then creating tasks that helped them grow it – essentially, a caretaker game, just like the tamagotchi. The tasks, however, sneakily get people to do the exercises I wanted.
Here’s an example. Your village is on fire…blow it out. So do it, right now: blow out a fire. 99% of people purse there lips and blow, which is actually a form of resistive breathing that, low and behold, helps strengthen core lung function. And a simple command gets most people to do it, without going into long details about the right motion.
Turns out, we could figure out a whole pattern of these that resulted in different stretches and exercises (my favorite, suggested by Sandeep, was “hug a really fat panda”). And there is a side benefit to the village tasks: they get kids familiar with taking instructions via text, which comes in handy for “attack” part of the app.
Having figured out preventive exercises, we needed to create something for when kids were actually having an attack, which led me to create the largest flowchart of my life. Basically, when a kid has an attack, they text in to a specific number and presto, we start asking them simple questions to help triage them. They find an adult, use the inhaler and breathe appropriately, and if they’re enrolled in the larger game, get some cheering on from their villagers.
And yes, it is still really just a bunch of text and we don’t know yet if that’s enough to get kids motivated. But I think even if the village metaphor doesn’t catch on, even just a free program that helps deal with an actual attack is valuable. And because we lucked out and had an Apple developer in Alan, we also go to produce a simple iPhone app based on the same principles: blow into your mic in a particular pattern in order to play the game and build your resistive breathing.
The results are out at breatheapps.org and Sandeep’s company has promised to continue the work. For my part, I’ll be staying involved, particularly in order to keep up with the data itself. Can a simple text intervention get kids to do this? Will it prevent attacks? Will fewer kids go to the hospital? We won’t know for awhile (assuming this catches on at all) but even if we get a few kids to breathe better, it was certainly worth a few sleeplessness nights at TEDActive.
Also published on Medium.