The Blue Button Initiative has been touted as a powerful way for veterans to download a summary of their health records from their My HealtheVet account. Last August, President Obama announced this capability of the VA system, bringing national attention to it. The simple “Blue Button” feature allows a simple text download of information from the VA system’s Personal Health Record (PHR) – part of the HHS effort to “liberate data” from otherwise inaccessible sources. It was the result of a collaborative effort between the VA, CMS, the Department of Defense, along with the Markle Foundation’s Consumer Engagement Workgroup.
And now, the VA is announcing a Blue Button Prize Competition, hoping to encourage innovators to build PHRs that have Blue Button capability, and can install that PHR on 25,000 physician websites across the U.S.
Is there a problem with this?
One of the “fatal flaws” of the Blue Button data format is that it is a simple, untagged text file – the output of such a file can be seen here. That means that the data is not easily readable by computers – it is not a tagged XML file, and thus very hard for a machine to read. The VA has talked about tagging their data so that it can be parsed into more-generalizable XML files, but that has not yet occurred.
As some critics have stated, the Blue Button initiative is “give-me-my-damn-data compatible,” but not something that interoperates with other systems. As simple text, the output is read-only, and is not re-usable. That might be fine for a read-only patient-facing PHR, or a read-only document that can be imported into a physician-facing EHR. But it is not something that will yield data that a computer can then use, blend with other data in an EHR system, and synthesize a new summary based on VA+EHR data.
HHS Chief Technology Officer Todd Park spoke to the virtues of Blue Button, saying that it was a conscious decision by members of his team to create something quick-to-market, in order to “liberate the data,” and not wait for more-structured data standards for export/import to mature (that may take a while). There is always this balance – how to deliver something now, vs. how to deliver something fully functional.
There are other structured file formats that are required of Certified EHRs to be able to read and write. The Continuity of Care Record (CCR), Continuity of Care Document (CCD) and the Clinical Document Architecture (CDA) are formats in place now, and are used to create “standard” XML messages that could allow different EHRs to send data to each other, and import it in an automated way. The raw, unorganized text that is part of the Blue Button download adds another layer – it cannot be something that competes with the required formats for data exchange that are built into Certified EHR products currently.
So, how does the Blue Button Initiative fit into the larger EHR/PHR picture? This might be best answered by breaking it down into a few bullet points:
(1) Giving patients access to their VA data is an important step forward, and should be applauded. Patient access is a core value built into Meaningful Use, and promotes collaborative healthcare. The Blue Button represents a “quick and dirty” text-based approach to this, and is a step in a positive direction.
(2) The Blue Button data format is not amenable to re-use by physician EHRs. In its current format, it cannot be parsed into an XML format, and doesn’t fit with CCRs or CCDs. It would be a huge step forward (a next-step for the Blue Button development group within HHS) for an XML tagged Blue Button data file to be created. That way, developers and innovators can transform some future “Blue Button XML file” into other message formats (CCRs or CCDs) and we can be another step closer to true health data interoperability.
(3) Blue Button should not be seen as competitive to established data exchange formats. It has its role, but has room to grow.
For now, Blue Button functionality will render a read-only text file for a given patient, from their VA/DoD health record. This is very positive. Patients, in their PHRs, will be able to download and see their My HealtheVet data – and doctors can see this data (if their EHRs are equipped) as read-only documents similar to the current way of seeing external correspondence like PDF or text files. It will require human eyes to digest this information, which is fine for now. Actually having the ability of machine-use of the data, so that the imported data can be blended with other local data, and a synthesized record can be generated – that is still a ways off. But that is the direction we should be headed.
