Military pushes for patient-centric structured-data EMR

In an update to its Composite HealthCare System patient record, the Department of Defense Military Health System has moved away from an institution-centered electronic medical record and toward a more centralized model that focuses on the patient.

"We have a unique mission in the military," Captain Robert Wah, director of information management in the Military Health System, said during a Tuesday morning education session at the HIMSS meeting.

A more patient-focused model would allow physicians to follow patients from the battlefield to Bethesda, he said.

CHCS II, which was deployed worldwide in January 2004, represents the military's next-generation EMR. It includes structured data collection, a global database, and clinical functionality.

To date, 39 of 139 military treatment facilities have gone online with the system. More than 14,000 encounters with the CHCS II occur every day, and it currently receives historical data from 44 of 102 host sites.

The military turned to commercial off-the-shelf software to create its computerized patient record.

"We're not in the business of writing software," Wah said.

An important capability of the new system is the addition of a structured data input capacity. The CHCS II does not simply convert a paper chart into electronic images, it actually creates an electronic, structured database that can be used for datamining purposes, he said.

Colonel Bart Harman, deputy director and chief medical information officer, summarized the potential applications for the system, which include improving the health of the general population, facilitating both medical and bioterrorism surveillance, and aiding disease management.

Organizations that do not have an easily mineable database might use proxy data to track the possible outbreak of bioterrorism, Harman said. In a population unfamiliar with smallpox, for instance, physicians would not be looking for that disease and might initially make a diagnosis of chicken pox. In the absence of a mineable database, organizations tracking disease outbreaks would seek spikes in chicken pox in older patients as a proxy for a smallpox outbreak.

With a more structured database such as that used in the CHCS II, investigators could search for specific groupings of symptoms, such as blurry vision, dysphagia, paralysis in both legs, and respiratory failure. By looking only for patients presenting with all four of these symptoms, a structured database could be mined for specific instances of botulinum toxin, for which these four symptoms are very specific.

This powerful built-in detection mechanism is one of many reasons to push for a strong medical record foundation built around a structured database format, Harman said.