When the Direct Project (a component of the Nationwide Health
Information Network) announced its first
hackathon yesterday I felt personally gratified as well as excited
to see the achievement of this milestone. The hackathon will take
place on October 27 and 28 and can benefit from the participation of
any programmers using Java and C#.
The Nationwide Health Information Network is the U.S. government’s
major open source initiative in health care. You could argue that
VistA, from the U.S. Department of Veterans Affairs, is more important
(it certainly is a vastly bigger code base), but VistA so far is
sparsely adopted outside the government, whereas the Nationwide Health
Information Network is positioned to become the platform for all
hospitals, clinics, doctors’ offices, and other health care
institutions to exchange information throughout the country.
The basic goal of this network is to allow health care providers to
exchange data on patients who move from one institution to another.
This could be an everyday occurrence such as a referral, or an
emergency situation such as a patient’s ER visit during travel. The
network will also facilitate the collection of data to government
agencies that do important work in health care statistics and
evidence-based medicine. What makes this all hard is the strict
privacy requirements that call for careful authentication and secure
data transfer; that’s why special code is necessary. Intermediaries
are also required to help health care providers authenticate each
other, and help providers that don’t have special security-enhanced
software to encrypt their data exchanges.
The original network rested a complex SOAP implementation that had
only scattered implementations and required most institutions to hire
consultants. The Direct Project will reimplement the security and
authentication through simpler protocols, starting with garden-variety
Doctors span a wide range of technical capabilities. Some are barely
storefront operations who consider themselves lucky to have PCs with
consumer-grade email clients. Others can afford special software that
supports S/MIME for encryption. The Direct Project has to encompass
all these participants. So the interface presented to health care
providers will be as simple as possible, but the implementations have
to be sophisticated and flexible.
This project has been conducted with the highest degree of openness
from the start. Anyone who’s interested can join a working group (I
dropped in on the Documentation and Testing group to review documents)
and a wide range of volunteers from major health care providers and
EHR vendors have been collaborating. From the conference calls and
email I’ve been on, things look very collegial and orderly. The
upcoming hackathon is the natural next stage in this open process.
The Nationwide Health Information Network has held hackathons before,
but this one is the first for the Direct subproject and shows that
it’s reaching a viable stage. A reference implementation for the
platform is nearly ready, but that’s only one node in a fairly
complicated architecture. For doctors to connect to the network,
client software and other mediators are needed.
So if you’re a programmer with an interest in health care, check out
the hackathon. It’s a chance to see where health care is going in the
United States, and help make it happen.