Electronic Health Records
Written by Peter A. Buxbaum
The Department of Defense’s Military Health System has declared its commitment to computerize health and medical records for some time. Over the last few months alone, however, those declarations have been matched with action.
The issue of the adoption of electronic health records within the military operates on many levels, as is reflected in recent developments. The DoD issued a recent policy mandating the collection of battlefield medical data on a single platform and has launched a prototype personal health record system to help MHS beneficiaries manage their own health information.
The technology issue that looms largest over MHS remains interoperability with the Department of Veterans Affairs medical record, a step which has been mandated by Congress to ensure that military medical information flows seamlessly to the VA system as returning warriors make the transition. Up until recently, that effort has been characterized by interim and incremental measures. But late in 2008, MHS and the VA announced the adoption of a new road map, which they hope will promote a more comprehensive level of interoperability.
In November 2008, Assistant Secretary of Defense for Health Affairs S. Ward Casscells released a policy requiring all services to use the Theater Medical Information Program—Joint (TMIP-J) for the collection and storage of all theater health care data on servicemembers.
TMIP-J is a suite of software tools fielded by the Army’s Medical Communications for Combat Casualty Care (MC4) organization and supports clinical care documentation and health surveillance and provides tools to track patients, supplies and equipment. DoD’s new policy assures that all of these activities are performed on the same platform, allowing them to be included in the central Clinical Data Repository, MHS’s data warehouse.
“Effective immediately, only the TMIP-J suite applications are authorized for use to collect and store theater health care-related data on servicemembers,” said Casscells’ memorandum. “Services must complete transition to the exclusive use of TMIP-J applications and ensure the use of TMDS across the continuum of care within 90 days of the date of this memorandum.”
The policy came to ensure that the DoD’s clinical data repository contains all health care data collected on deployed servicemembers and to ensure that medical information is accessible to the VA, the policy memo also noted.
Even prior to the establishment of the policy, “the services had already been ramping up their use of TMIP-J software,” said MHS chief information officer Charles Campbell.
According to Campbell, the Army has 100 divisional units fielded; the Marine Corps has 62 forward resuscitative sites operational in Iraq; the Air Force has two theater hospitals and over 10 forward resuscitative sites in Iraq, Afghanistan and Qatar; and the Navy is using the TMIP-J software on both the USS Ronald Reagan and the USS John C. Stennis.
Medical clinicians have documented 2.3 million outpatient encounters since January 2005 and over 5.6 million inpatient encounters since June 2007 using TMIP-J, Campbell said.
“That information is being transferred to the Theater Medical Data Store and Joint Medical Workstation for viewing by combatant commands and/or the VA for follow-on care,” he added. The TMIP-J policy represents an advance in the collection of health information by the military. But the ultimate goal of the digitization of medical information is not merely to collect data for MHS and VA but to make that information more available to the patients themselves through personal health records, or PHRs. The theory behind PHRs is that better informed patients will take more responsibility for managing their own health, leading to better health outcomes, as well as lower costs.
To that end the Military Health System launched a prototype personal health record in December 2008 designed to help members of the military and their families manage their health information. Dubbed MiCare, the PHR was introduced at Madigan Army Medical Center in Tacoma, Wash.
The PHR is not a singular concept, but includes a number of different capabilities, all or some of which may be included in any given system. PHRs can provide a variety of functions such as electronic repositories, where patients can securely and privately store health-related data and share that data with one or more health care providers at the patient’s discretion. Still other PHR capabilities allow patients to refill prescriptions, set appointments and view medical records.
MiCare began as a pilot project with Microsoft Corp. and Google in March 2008. Its purpose is to replace MHS’s existing Tricare Online and to incorporate features offered by commercial PHR vendors. Tricare Online, which began as an information portal in December 2007, has rolled out only a limited number of features since then.
The VA also has a PHR called MyHealtheVet, an effort initiated in 2003 and one of the oldest of personal health records systems. MyHealtheVet, which also started out as an information portal, has succeeded in incorporating more electronic features than MHS’s TRICARE Online before it was replaced by MiCare. “One of the goals we have in MHS is to make sure that our patients are all well-informed consumers who take active participation in their health care,” said Campbell. “We believe that MiCare facilitates that particular goal. The more folks own their information, the more informed they will be and they will be better able to manage their own health care.”
MiCare will allow beneficiaries to choose between the Microsoft HealthVault and Google Health systems. Both will serve as patient-controlled repositories for health information and will provide access to demographic information, medication lists, allergy data, lab and radiology results, upcoming appointments, and documentation from the military electronic health record. Both systems will also provide the ability to store health records obtained from civilian providers, plans and pharmacies. The beneficiary will also be able to choose to share this with health care providers.
“When we started working on our PHR, we were looking to see if there was something out there commercially that we could use so that we didn’t have to build and manage it ourselves,” said Campbell. “Then Microsoft announced HealthVault.”
Using a commercial system like HealthVault means that Microsoft, which offers the service free of charge to users, has the burden of data storage, operating a help desk and providing technical support, Campbell noted. When Google announced its PHR offering soon after, MHS saw the opportunity to offer its participants a choice.
“One thing we strive for is patient choice,” said Campbell. “These are two well-known, large companies and we ended up working with both.” The PHRs of MiCare subscribers will receive feeds from AHLTA, MHS’s electronic record system, as well as entries from non-MHS sources on prescriptions, allergies, lab and radiology results, and clinical notes.
“We have an initial subscription for 250 users,” said Colonel Keith Salzman, M.D., chief of informatics at the Western Regional Medical Command and at Madigan. “We want to see how the workflow proceeds and how the information is displayed. From there we will want to refine the system to make sure users receive useful, clean information and work out any bugs and problems.”
Users are required first to sign up with Microsoft and/or Google to set up their PHRs. Active duty personnel may then authorize MHS to transfer data to the PHR online using their common access card (CAC). Other MHS beneficiaries must sign an authorization document at an MHS facility.
“We are just now beginning to evaluate the system and to solicit feedback,” said Rick Barnhill, a Madigan program manager. “So far we have found that people interested in managing their personal and family care are signing and wanting to participate.”
For Campbell, promoting the use of PHRs involves more than cost considerations. “If a servicemember is on leave and driving across country and gets injured, he would have access to his health care information and could give access to that information to whichever health care provider he happens to see,” he explained. “We think that is going to be very helpful.” MHS plans on introducing PHR connectivity for providers who are outside of the military system. “Sixty-five percent of the care we provide is outside our direct care military system and is not captured in our electronic health record,” Campbell noted. “Our PHR will be based on the Department of Health and Human Services standards for sharing health information. As the nation moves toward that standard, we will be able to more easily share information with civilian providers.”
MHS also plans to make MiCare interoperable with MyHealtheVet PHR, according to Campbell, but has not started on that process yet. “We consider this to be a very important piece of the information sharing between the two organizations,” Campbell said. “A lot of our patients move over to the VA so we want” the two PHRs to interoperate.
Interoperability between the DoD and VA PHRs is an issue that will be tackled down the road, Campbell suggested. Of more immediate concern to both departments is fulfilling the Congressionally mandated goal of interoperability between MHS’s electronic health record with the system used to support the treatment of veterans.
To date, the departments have undertaken this project by implementing incremental information sharing capabilities. They have implemented the Bidirectional Health Information Exchange (BHIE), which provides a real-time interface between AHLTA and the VA’s electronic health record system, known as VistA and the Clinical Data Repository/Health Data Repository (CHDR), which combines both systems’ data repositories to enable the exchange of drug interaction and allergy information for shared patients.
A VA clinician viewing patient information in VistA, for example, can click on a BHIE icon to see of if there is any updated information from AHLTA, the military health record, on the issue the clinician is investigating for that patient. But in November 2008, DoD and VA announced they will be seeking to enhance the interoperability of their electronic patient clinical data by migrating their respective electronic health records systems to a service-oriented architecture. A service-oriented architecture refers to the development of software capabilities through the integration of loosely coupled, reusable components, as opposed to the point-to-point integration between standalone systems.
That decision represents the acceptance of recommendations made by government consultancy Booz Allen Hamilton following a study that was completed in September. “The Booz Allen study looked at three different options,” said Stephen Jones, principal deputy assistant secretary of defense for health affairs at a Pentagon press conference, “building on the current approach, replacing AHLTA with VistA, and replacing both with another commercial off-the-shelf solution.”
The favored option will allow DoD and VA to “move forward with information sharing in a compatible way.” The Booz Allen analysis demonstrated that there was a 97-percent overlap of requirements among the two departments with respect to inpatient care electronic health records, said Robin Portman, who led the Booz Allen team that undertook the study. Once it was determined that it was possible to merge the requirements of the two departmental systems, Booz Allen set about to analyze how to go about it and developed a series of options. “The status quo is the path they are on today,” said Portman, “with both organizations trying to update their current systems and doing point-to-point integrations based on their priorities for the need for data sharing.”
Booz Allen also considered whether VA could use AHLTA or DoD could use VistA, whether they should invest in a fully integrated, commercial off-the-shelf (COTS) product or take a best-of-breed approach, in which different high-performing functional components are integrated together.
The panel settled on the SOA approach after considering the risks and costs of each option, according to Portman. “On the clinical side the recommendation is that DoD and VA establish an interagency service-oriented architecture and implement services to support provider access to all the information they require,” she said. “The departments will need to work together to identify what the common services will be and how to sequence them.”
The common service specifications must be coupled with common data standard and ontologies once the architecture is established, Portman noted. “When getting off the ground, they need to identify common services both organizations deem to be high priority,” she added. “They will then use a joint governance structure to test the process underneath the services to be sure both organizations will be able to work together to implement this kind of solution.”
Portman believes that one benefit to the implementation of a service-oriented architecture will be to facilitate the exchange of information with providers outside of the DoD and VA systems. “We didn’t want to come up with a solution that precluded sharing information with outside providers,” she said.
The Booz Allen recommendation leans heavily on the development of a joint governance structure between DoD and VA. “It is going to require a strong community of interest backed by the leadership in both departments,” said Portman.
Governance, rather than technical issues, may indeed be the sticking point when it comes to the implementation of a joint service-oriented architecture. Congress has already mandated the establishment of a joint DoD/VA interagency program office to facilitate health information exchange, yet a report from the Governmental Accountability Office released in July 2008 characterized progress in that area as sluggish.
“The departments’ effort to set up the program office is still in its early stages,” the report stated. “Leadership positions in the office are not yet permanently filled, staffing is not complete, and facilities to house the office have not been designated.”
Congress directed the departments to have the office fully operational by September 30, 2009, but the GAO report cast doubt on whether that mandate can be fulfilled. “Without a fully established program office and a finalized implementation plan with set milestones, the departments may be challenged in meeting the required date for achieving interoperable electronic health records and capabilities,” the report concluded.
Portman agreed that governance issues present a major challenge. “We do highlight the need for everyone not to underestimate the challenges of a joint effort like this one,” she said. “Establishing a true interagency program office can be challenging.” Portman pointed with approval to joint DoD/VA efforts to date, including the establishment of the BHIE and CHDR. She also noted that Booz Allen is recommending that the SOA be implemented through a series of pilot programs and that the departments establish priorities and sequencing for the rollout of services. But Arthur Wu, staff director at the House Veterans Affairs subcommittee on oversight of investigations, reacting to the announcement of the joint DoD/VA SOA, was skeptical about the ability of DoD and VA to work together. “We’ve been after them for years,” he said. “It’s a question of pushing two cultures together.”
For her part, Portman would rather view the silver lining than the cloud. “This will allow both organizations to think futuristically about their requirements and drive industry common services specifications for COTS products,” she said.
In other words, if DoD and VA can get their acts together around a joint SOA, they will not only promote interoperability between their two systems but will likely influence the adoption of technology standards for the health care industry as a whole. ♦