National and Regional Health Information Networks – History Lesson

All the buzz in the U.S. right now is related to the strong push towards national and regional health information exchanges. In April 2004, the President announced the intention to provide within 10 years an electronic health record to every U.S. resident for the purposes of improving medical care and reduce cost. As a result, the Office for the National Coordinator for Healthcare IT (ONCHIT) was created.

The first job for ONCHIT was to outline its goals in the Strategic Framework for Action (SF) published in July of 2004. In November 2004 ONCHIT published a Request For Information (RFI) asking for input on how a National Health Information Network (NHIN) might look. ONCHIT asked 24 questions covering everything from the architecture, the governance and the system of incentives to make the whole thing work. We at Microsoft have been very involved in this process and as one of the major software vendors in the U.S. and worldwide, we were very pleased with the whole effort. Although technology and software will not be the only components of the solution, they will certainly play an important role.

In order to respond to the ONCHIT RFI, we did the responsible thing and instead of individually addressing the response we partnered with other major IT vendors to form the Interoperability Consortium (IC). The other companies that participated in the IC were: Accenture, Cisco, Computer Science Corporation (CSC), Hewlett Packard (HP), IBM, Intel and Oracle.

The IC’s response to the HHS RFI was a whopping 140 pages. I'll spare you the details, however, broadly we proposed that the NHIN architecture should be based on a service oriented, and be a distributed and federated model using Web Services to interconnect all the different components and services.

One thing was evident from the beginning given the privacy and security concerns here in the U.S.  A centralized architecture based on the model of the UK NHS was not going to work. What was going to work was instead an architecture based on a simple, distributed model that would scale to a national level and allow for multiple value-added services to be gradually deployed over time.

Another key aspect we considered was the local nature of healthcare delivery, where most of the interesting stuff happens in a well defined geographic location. The Regional Healthcare Information Organization (RHIO) is the aggregation of several healthcare providers and healthcare organizations that would share a common backbone for clinical data exchange. The term "regional" is somewhat misleading. In fact, a RHIO could be an entire Integrated Delivery Network (IDN) such as Kaiser or Mayo Clinic and not be tied to a specific geography.

We suggested that the NHIN would be formed by the aggregation of RHIOs across the country, all talking to each other using interoperable communication mechanisms such as Web Services.

After receiving over 400 responses, ONCHIT released a document outlining the major findings and preparing the ground for a Request For Proposal (RFP) issued in June of 2005.

In the meantime we have been involved with RHIO projects like the one between Massachusetts, Indiana and California, led by MA-SHARE and Connecting for Health. The RHIO prototype is now operational and the three states are able to communicate and exchange clinical data using a variety of HL7 v2 (using the HL7 2.xml standard) and HL7 v3 over Web Services. The three different RHIOs run a variety of commercial (Microsoft in MA and Sun/Java in IN) and open source (in CA) Web Services stacks and everything works seamlessly across the different platforms.

This same group, led by CSC, was awarded one of the four contacts for NHIN prototypes part of the ONCHIT RFP. We hope to take this prototype to the next level and come up with a set of recommendations for architecture and communication protocols for ONCHIT. I'll be working with this team, providing architectural advice and expertise on Web Services and how they relate to HL7 v3.

After the all contracts have been awarded, the situation looks something like this:

  • ONCHIT is proposing the use cases where we need to focus in this first phase of the development of the NHIN
  • The Healthcare IT Standards Panel (HITSP), part of ANSI, will issue recommendations on what standards will need to be employed. Microsoft is member of HITSP and will be actively participating in the process.
  • The Health Information Security and Privacy Collaboration (HISPC), overseen by RTI International, is developing security and privacy guidelines and solutions.
  • The Certification Commission for Health Information Technology (CCHIT), part of HIMSS, is figuring out how to certify products and solutions to comply with the recommendations of HITSP and HISPC.
  • The 4 prototypes, one of which we are part, will be initially feeding the process and ultimately will represent some kind of reference implementation and will be required to ultimately all talk to each other.

Sounds simple, right? Well, unfortunately it's not 🙂

This is the end of our history lesson. The rest is still being written. Let me know if you want me to go into more detail of the way we think about RHIOs and the NHIN at Microsoft.

Until then, ciao!

Comments (5)

  1. Clipperz says:

    Personal health records are definitely the more common type of sensitive data we produce and own. Nonetheless we are not really in control of those data. They are spread in folders, binders and drawers at hospitals, homes and physician offices. Most pati

  2. Jianhua says:

    What is the consideration to have four prototypes? Is it because ONCHIT would like to compare these four and select a best one, or it does not want to see only one prototype dominating the market (like Mircosoft dominating PC operating systems)? What will happen next? Will ONCHIT keep all of the four and requires them to talk to each other, or combines the best features from these four and creates a new one that has all of these good features? If possible, please share your thought on these questions. Thank you.


  3. There is a good article that oulines ONCHIT’s strategy when it comes to the different prototypes at

    This should answer your questions.

  4. If you pay any attention to what’s happening in the U.S. around the National Health Information Network

Skip to main content