Tests of New HIE Architecture SlatedONC Taking an Incremental Approach to Evaluating PCAST Plan
That's the word from Farzad Mostashari, who took over as head of the office on April 8. At the Health IT Policy Committee meeting Wednesday, Mostashari acknowledged that ONC plans to follow the committee's PCAST Report Workgroup's suggestion to test elements of the proposed architecture in phases, starting with providing patients access to their data via a portal. "It's a very creative, smart option for a way to move forward," he said.
"We have some breakthrough grants to our state health information exchange grantees in Indiana and Montana ... specifically looking at this issue," Mostashari told committee members Wednesday. "And there may be opportunities within the federal government to look for platforms for doing this sort of experimentation and learning."
The approach recommended by PCAST, calling for the eventual movement to one architecture for all health information exchange nationwide, represents a major departure from ONC's ongoing efforts to "let a thousand flowers bloom" and accommodate numerous HIE models, said Paul Egerman, co-chair of the PCAST Report Workgroup. But he stressed that the presidential council advocated an evolutionary approach to changing the HIE architecture. "This is not rip and replace," he said.
Universal Exchange LanguageThe PCAST report on information exchange - a complex document that's proven controversial - calls for developing an architecture based on a new universal exchange language and interlinked search capabilities, coupled with privacy and security safeguards (See: Embedding Privacy in EHRs). It calls for using metadata tags tied to individual data elements within electronic health records. The tags would contain descriptive information, such as patient consent to exchange the data.
In theory, the new architecture, for example, would help enable a clinician to make a query to track down information about a patient from multiple sources or a researcher to identify appropriate patients for a clinical study.
ONC, a unit of the Department of Health and Human Services, already has begun work on determining whether existing standards can be used to tag individual data elements within an EHR with information about patient identity, data sources or "provenance" and patient consent preferences, Mostashari said. Early this summer, ONC will issue a notice seeking "structured feedback on those particular metadata standards that we will have initially identified as being potentially suitable in the universal exchange language," he said.
The HIT Standards Committee ultimately will make recommendations on whether to include any metadata standards, such as for use in portals that patients use to access records, in the Stage 2 requirements for the HITECH Act electronic health record incentive program, Mostashari said. ONC is slated to draft a proposed rule for Stage 2 requirements by December.
PCAST CriticismEarlier, critics of the PCAST report portrayed as unrealistic its ambitious recommendation that the new architecture, including a universal exchange language, be required for future stages of the HITECH Act EHR incentive program (See: Universal Exchange Language Debated). And some said the council's report lacked enough detail to determine if its suggested approaches would adequately safeguard privacy.
Plus, some critics argued that instead of creating a new approach to HIE architecture from scratch, which could prove costly, federal authorities should focus on further development of existing standards, including Health Level Seven, and related implementation efforts. (See: EHR Interoperability: What Works?)
The PCAST report advocated an architecture that it contended could help pave the way toward the following:
- Every American would have electronic health records and would have the ability to exercise privacy preferences for how those records are accessed, consistent with law and policy.<.li>
- Subject to privacy and security rules, a clinician would be able to view all patient data that is available and necessary for treatment. The data would be available across organizational boundaries, using what amounts to "pull" technology.
- Authorized researchers and public health officials, subject to privacy and security rules, would be able to leverage patient data in order to perform multi-patient, multi-entity analyses. Researchers could pull necessary data for this type of secondary use.
Major Policy ConcernsOn Wednesday, the HIT Policy Committee accepted a detailed analysis of the presidential council's report from its PCAST Report Workgroup. Although the workgroup was not asked to judge the merits of the PCAST proposal, its letter, nevertheless, made it clear that the proposal raised policy and operational feasibility issues. As a result, it stated, "large operational tests are needed that resolve the policy and feasibility concerns."
The workgroup's letter stated "it is possible to implement the new exchange architecture in a series of incremental steps." It also noted several challenges involved in implementing the architecture concept. For example, pointing to PCAST's call for attaching privacy preferences to individual data elements within an EHR, the letter stated: "The extent of that granularity and the practicality of data element privacy choices are topics for review."
To read the workgroup's full report, click on PCAST Report WG Draft Letter in the April 13 HIT Policy Committee meeting materials.
Mostashari, acknowledging the substantial challenges involved in making the PCAST report's vision for health information exchange a practical reality, strongly endorsed the phased approach to testing the concept. "We're going to move forward on something that can really meet the vision but also be feasible to implement," he said.