The Direct Project Goes Live

For those not familiar with the Direct Project, it started out about a year ago as NHIIN Direct and then was renamed several months ago. The goal was to collaboratively develop a specification to enable point-to-point health information exchange for providers, patients, and others. The timetable for this project seemed awfully ambitious to me, but I was wrong. There were several main drivers: the meaningful use stage 1 requirements for information exchange capabilities, the desire to use existing technologies to reach the goal, and a philosophy to make sure to incorporate the needs of "the little guy." Now, it is just a year later and the first pilot implementations of the Direct Project are being rolled out. By any measure, this is an impressive accomplishment.



Early on, the project work group went through a lively discussion concerning adoption of the transport standard to be used. SOAP, REST, and SMTP were among the protocols considered. Finally a consensus was reached to adopt SMTP with s/mime for security. Once SMTP was selected, the project moved forward more rapidly and with less controversy. The specification enables point-to-point exchange of a variety of health information in a fashion similar to email.


The pilots that have been announced to date seem designed to fulfill relatively limited use cases. Uses of Direct will expand rapidly as the safety and efficacy of the project are validated by experience gained in the production environment. Vendors will need to implement Direct software specification in their systems and an infrastructure of HISPs (Heath Internet Service Providers) will need to be established to serve all areas of the country.


In his Halamka blog Dr. John Halamka explained how the Direct specification was used to send data from his CCD/CCR at BIDMC to his PHR in Microsoft Health Vault. It took engineers at both ends about a day to get all the configurations set up. Certificate management had to be accomplished. When they were done, he was then able to extract "atomically" section data from the summary documents and organize it in a manner that gave him great flexibility in managing his health information in the PHR. The workflow to update the PHR sounds manual now but it could be automated for future uses. This potentially removes one of the barriers for more widespread adoption of PHRs by the public. Until now, it has been very difficult to import and organize personal health data.


The Direct Project removes an important barrier to more widespread information exchange. I am enthusiastic about its potential. It uses open source software and appears to be relatively easy to implement because existing technology is utilized. I am sure that many clinicians, especially those in small organizations and practices, will use Direct for their information exchange needs in the near future. But there are some caveats. EHR vendors will need to incorporate the software into their systems. (Epic and Allscripts, for example, demonstrated impressively the use of Direct at HIMSS11 Interoperability Showcase last week.) Will this trigger a need to update certification of the EHR software? As more sites use the specification another problem can be anticipated. One needs to know addressing information for the point-to-point exchange. Address management was not an issue addressed by the Direct Project. In other words, provider directories were not included in the specification. So users will need to develop a means to store frequently used addresses (a contact list like that included in many email programs) and an easy way to discover unknown addresses. Finally, Direct utilizes a "push" mechanism. Queries for data about a patient are not possible. Direct may not work well for after hours information exchange needs or for use in emergency situations.


My prediction is that the Direct Project will be utilized by many organizations intending to jump start meaningful health information exchange. The business case for its use is less complicated than for many more capable models for health information exchange. The Direct Project has proved to be the paradigm case for the successful, rapid, development of health information technology projects through a public-private collaborative. It should serve as a model for future health IT projects of similar scope. See the ONC video announcing early success with the project Direct Project Video.