Hacking the flu emergency at CrisisCamp Boston


Last month at the Center for Civic Media we held CrisisCamp Boston – an event that is part of the global Crisis Commons organization and sprung out of the Hurricane Hackers group that began life in the Center for Civic Media. There were three motivations for organizing the event: to build on the success of the Sandy group and move forward with those projects, to tackle an immediate and local issue (Boston’s flu emergency) and to experiment with a new hackathon / workshop format.

Building on learnings expressed by Denise and Pablo (among others), Pablo, Lyre and I wanted to find a format that would bring a more diverse range of participants into the process – particularly policymakers and subject experts with no software development experience – and that would draw on specific use cases. In other words, in a rapid R&D format, to spend much more time on the R than is typical at hackathons. Over the weekend the research time was very productive, covering some of the on-the-ground issues faced by a city responding to the flu (or comparable medical emergency) and brought in experts such as Aaron Kite-Powell, an epidemiologist at Lincoln Labs (and former health department officer in Florida and Maryland), Seán Fay of the Healthcare Emergency Medicine program at BU, Justin Kates from the Nashua Office of Emergency Management and Mike Evans, a civic technologist at the New Urban Mechanics, who recently launched the Flu Shot webapp.

Screen Shot 2013-02-05 at 14.14.13

Mike kicked off the event by showcasing Flu Shot and explaining how the team had adapted code from Chicago’s Department of Public Health to deploy the tool in under an hour. The group then began rapid brainstorming around the flu emergency and formulating suggestions of possible projects. Among the many suggestions, the two projects that teams decided to work on were:

1. A mobile app for first responders in an emergency situation to quickly see an aggregated, geo-located collection of all the relevant public data available – e.g. the location of shelters, incident reports and contact information.
2. Flu Reporter, a standardised mechanism for doctors to quickly report flu outbreaks to their county or state health department. Not all states require doctors to report flu cases, and the process tends to be ad-hoc and done entirely by phone calls, so often these reports aren’t made very quickly, meaning that a potential early warning signal is missed.

One of the challenges of a hackathon or workshop event that brings together policymakers and software developers is how to keep both types of participant engaged at all times – even when it’s not “their turn” to take the lead. That’s especially difficult when the coding begins. Policy people may say – “I don’t know how to code, so I’ll leave you software people to it.” But of course those subject experts are essential at this point. You don’t just hand coders an IA in an event of this kind and expect a solution. Hacking together solutions means iterating, and that requires strong sector knowledge and a constant persistent focus on the use case. A great tool with no use case is not a useful outcome when you’re seeking to respond to a real (future) crisis.

Both groups set down the early shoots of projects – in both cases, mobile apps aimed at a very specific user group. You can see some of that preliminary work on github and the research documents, but there’s a long way to go. As with all one-off events, the challenge is to keep the flow of enthusiasm going. That’s what we’re hoping to do with future CrisisCamp events. And it can work: at this CrisisCamp some members of the group returned to work on InnCrisis, a fundraising / donation platform that links donors directly to the recipients of the aid they provide. That project continues to be developed.

The Flu Reporter group also found that though the app concept had a clear use case, it’s only a part of the solution. In fact a more immediate issue to tackle might be to standardize the format in which doctors report disease cases – a national set of open reporting standards, from which an API could be created. It wasn’t an issue that the group tackled directly during the event, but it’s an interesting potential project for the future.

Check out the raw documentation from the event to dig deeper into the process. If you’d like to join the CrisisCamp Boston group and get notified about future events, sign up here. Huge thanks to everyone who participated (in no particular order): Anne Donohue, Michael Evans, Jon Lin, Ryan Kahn, Masha Josefson, Ryan Burns, Aaron Kite-Powell, Joe King, Luisa Beck, Al Willis, Diane Williams, Kate Balug, Seán Fay and Justin Kates.

Finally, students in the Cambridge area who are interested in exploring better ways to run hackathons and design workshops might be interested in the Civic Media Collaborative Design Studio course at MIT this semester, taught by Federico Casalegno and Becky Hurwitz and involving several community partners.

Cross-posted at rodrigodavies.com.