Joint Hearing of the Committee on Homeland Security and Governmental Affairs of the U.S. Senate and the Committee on Veterans’ Affairs of the U.S. House of Representatives at 1:00 p.m. CDT.
Witness Testimony of Colonel Keith Salzman, M.D., MPH, FAAFP, FACHE, Western Region Medical Command and Madigan Army Medical Center, Chief of Informatics, Tacoma, WA, Department of the Army, U.S. Department of Defense
Chairman Mitchell, Congresswoman Brown-Waite and distinguished Members of the Subcommittee, thank you for inviting me to discuss the information technology sharing project between Madigan Army Medical Center and VA Puget Sound. I am Colonel Keith Salzman, a physician and a DoD/AMEDD leader in the newly emerging discipline of Informatics and it is my privilege to serve as the Chief of Informatics at Western Regional Medical Command/Madigan Army Medical Center where we enjoy a long history of command support for our work in Informatics.
I arrived at Madigan as the announcement was made that Madigan and VA Puget Sound would be working together to share electronic clinical information. I joined the team as a steering committee member. While the submission for the information sharing project occurred prior to my arrival, I have been on the project since its inception and continue to the present. We have completed all of the business plan objectives in 3 years of a 4 year project and are using the remaining funds to provide additional requests for document exchange that support Poly-Trauma information needs as well as other key documents and data types that contribute to extending interoperability, on the approval of the DoD-VA oversight committee members.
The Madigan-VA Puget Sound project arose in response to congressional requirements for the DoD and VA to each contribute set aside funding for 4 years to collaborate on sharing clinical information and care to improve healthcare services to shared patients. At the outset of this testimony I would underscore our assessment that; the choice on many levels between ‘either’, ‘or’ is more appropriately answered as ‘both’. I will explain as I review this project.
This particular demonstration was undertaken in response to section 722 of the FY 2003 National Defense Authorization Act which required no less than three demonstration projects of DoD/VA coordinated systems involving budget/financial management; staffing/assignment; and Information Management/Information Technology (IM/IT). Madigan and VA Puget Sound were selected for this project based on the established clinical sharing that was in place and the need to improve the exchange of clinical information to provide care for the mutual patients cared for at Madigan Army Medical Center, and the American Lake and Seattle VA centers that make up the VA Puget Sound Healthcare System.
The initial challenges surrounded the learning required to overcome the first ‘either-or’ proposition of who drove the project: enterprise or the local site. A critical first lesson learned was-‘both’. The local site had access to the clinical end user community and the requirements necessary to improve the flow of information while the enterprise had ownership of the architecture and systems in which requirements would be built and deployed. At the outset it is important to state that while this project is a demonstration project, all of the deliverables are being used by the enterprise systems of both the DoD and VA in production, in near real time (meaning seconds to minutes as a rule, not instantaneous or days to weeks).
After the initial assembly of local and enterprise teams and review and approval of a detailed business plan, the teams moved forward with iterative delivery of tangible products implemented and delivered for use in enterprise systems (SHARE for the DoD view) and Remote Data View in the Computerized Patient Record System (CPRS the VA view) of the Bi-Directional Health Information Exchange (BHIE) validated dual beneficiary patients. The work cycles for this project were generally 6-9 months in duration.
A second lesson learned was that while each system had its own test patients, shared test patients served the same purpose for interoperability (that purpose being validating information compilation and flow within the shared framework). While not as profound, the benefit of ‘either-or’ answered in ‘both’ facilitated testing, training and expansion of functionality.
The critical dialogue between clinical end user and the development team at the local level, combined with an active dialogue between local and enterprise team members, ensured that a principle of software development (namely to correct functional problems as they are identified in the design phase) proceeds iteratively and cost effectively. The savings can be significant over allowing major design problems to persist into production. This exemplifies another ‘both’ solution to an ‘either-or’ proposition.
Regarding requirements specifications, we observed that keeping the user requirements in sight while drafting the statement of work and contracting progress will save re-doing a product after-the-fact. A case in point is work on delivery of specified note types. The initial requirement was for 7 note types. Through a disconnected process of contracting, the requirement was interpreted as all notes, creating an information retrieval and storage problem, unintended consequences of assumptions made by contractors making assumptions about what the end users really needed. The experience was used later in our development of requirements by keeping an open dialogue between the end users and the enterprise-another ‘both’ solution.
With regard to the elephant in the room—establishing either AHLTA or VistA across both Departments-we observed the following:
There are strengths and weaknesses in both systems that complement each other. AHLTA is integrated world-wide and available 24/7. There are functionality problems that are being worked to improve use at the clinical and business level. VistA shows the benefits of local design in its adoption by end-users who are more inclined to buy into a product they created. The downside is the historic lack of configuration management. I use management intentionally as against configuration control. The VA faces big challenges in reorganization and must be careful not to destroy the strategy that delivered its success while addressing its Achilles heel of decentralized, unmanaged growth. The cost of imposing one system on both organizations now would be prohibitive. Establishing interoperability and designing a strategy of convergence over the next 10-20 years will allow a ‘both’ solution that capitalizes on best practices and less disruptive changes to either system.
By using an interoperable approach, the DoD and VA, who own about 50% of the penetration of the Electronic Medical Record (EMR) on the national level, can pave the way for interoperability as use of the EMR extends from large organizations to the small provider groups and individual patients who constitute the majority of the nation and who are not benefiting from an EMR. We are using the strategy of interoperability to extend to our indirect care providers in TRICARE and CHAMP-VA to capture the documentation that occurs outside of our EMRs. The extension makes a natural bridge to Regional Health Organizations.
A key to success in our strategy was to use messaging standards (HL7 (Health Level 7), Clinical Document Architecture (CDA), Release 1 and 2), which conform to security documentation requirements and integrate with the enterprise constraints from the local level. As stated at the beginning, this partnership between a local development cell immersed in the end user environment and the enterprise for configuration management is a critical model/partnership to succeed in developing software and hardware solutions for clinical-business processes that support healthcare delivery for our beneficiaries.
An observation regarding (commercial off the shelf) COTS solutions for federal agencies is that common products such as identity management and Single Sign On/Context management solutions can be purchased in bulk with significant efficiencies for the government.
In the end, we found that crossing new frontiers in collaborative work between federal agencies and local/enterprise ends of those agencies underscored our finding that ‘both’ solutions work better than ‘either-or’ solutions.
These comments summarize what I would offer as a steering committee member engaged in this project from the start. Subject to your questions I would like to thank the Subcommittee again for allowing us to share our insights on this critical work that is progressing successfully. I would also encourage Congress to continue its support of this program and each of the agencies involved. I look forward to your questions.
Appendix A: Data Currently Being Shared
- Outpatient medications
- Lab – Chemistry, Hematology, Micro, Path, etc
- Radiology Text Reports
- Pre and Post Deployment Assessments
- Post Deployment Health Re-Assessment
- Discharge Summaries (DoD Essentris Sites and VA)
- MAMC legacy outpatient notes to VA
- Theater Clinical Data
- Op Reports, Surgical Reports, History & Physical, Consult Results and Progress Notes (Fall 2007)