Overview of hData

Project hData is an open source, open technology initiative of The MITRE Corporation. It was started in the context of experience gained from creating Project Laika, an open source validator for the HITSP/C32 profile of the HL7 Continuity of Care Document (CCD). During this project we documented the problems of creating, exchanging, and validating CCDs and started to think about remedies for these issues. The outcome is hData, a set of specifications focusing on simplicity and scalability in health data representation and exchange. By working with our federal sponsors, we were able to present hData to a number of groups with excellent feedback.
Project hData was first presented to a wider audience at the 2009 Balisage conference, which included a white paper. Since then, additional development work has gone into the project and two specifications (Record Format and Transport & Packaging) and one security requirements document have been published.

Mapping hData Features to NHIN Direct Design Principles and Rules


hData Feature
Benefit
NHIN Direct Design Principle
(this mapping is not complete)
NHIN Direct Design Rule
1.
Small data objects
  • Significantly simpler schemas and parsers
  • Extensibility model
  • Dynamic updates
  • Low barrier to entry
1, 2, 10
1, 2, 4
2.
Hierarchically ordered linked data
  • Simpler parsers
  • Dynamic updates
  • Rich content context through linking
2
2, 4, 5
3.
Rich Meta-data model for content
  • Allows to sign and/or encrypt data
  • Any valid Internet Media type including DICOM or JPEG images, S/MIME content, etc. can be transported
  • Source information
1
1
4.
Strict separation of technical framework and clinical content
  • Modularity and extensibility
  • Different communities of interest (COI) focus on different specialities
  • Support of multiple data representation through single interface
1, 2, 5, 10
6
5.
RESTful Transport
  • Dynamic updates and subscription
  • Highly scalable
  • Low barrier to entry
5, 8, 9
3, 5, 6, 8
6.
Distributed data storage
  • Keep data local to data originator for timely updates, but allow remote caching
  • Eliminate single point of attack
5, 8
1, 8
7.
Patient-centric security model
  • Increase acceptance by patients for EHR systems by allowing them to control data access (within limits of law)
  • Enable secure dynamic discovery of distributed patient records
  • Prepare for fully mobile health care through cryptographic protection
4, 8
5, 6, 8
8.
Open Source, open technology, open IPR
  • Free to adopt, implement, extend
  • No royalties, since MITRE is a non-profit organization chartered to work in the public interest
10
3, 4


Mapping to NHIN Direct User Stories


The following table maps the user stories to hData functionality.
User Story
hData Features
Additional Information
1, 2, 3
1, 4, 5, 7
At last year's NIST 5th Annual Security Automation Conference we presented a scenarios that covers most of these three user stories. Please check this presentation for more information.
4
1, 4, 5, 7
In our model, this is very similar in implementation to stories 1, 2, and 3
5
1, 2, 5, 6
Labs (and similar entities) will not necessarily need to run an hData host and client, but pull and push data through the basic RESTful interface.
6
1, 5
Similar to user story 5., where only a small set of information (and a much smaller set of health data schema needs to be implemented)