Open source spurs e-record adoption

The move toward all-electronic health records (EHRs) is gathering momentum and support. Just over 38% of U.S. hospitals now use EHRs, according to a recent survey conducted by ASHP. That's up from 31% in 2007. In hospitals that use EHRs, 94% of pharmacies have access to patient records.

More information at the point of care is just the beginning. An EHR gives clinicians masses of population data that have not been readily available from paper records. A growing number of studies show that EMRs are a fast, efficient, and cost-effective means to improve care by identifying high-risk patients who would benefit from specific interventions. And, for the first time, EHRs are showing signs of affordability. Open-source software originally developed at Veterans Affairs is pulling EHR pricing down from stratospheric levels.

Bids from similar proprietary systems came in around $18 million, according to David Whiles, Midland information systems director. Midland is the first community hospital in the country to go live with the open-source EHR. "We are at the front edge of a move toward open-source products in healthcare," said Medsphere president/CEO Kenneth Kizer, M.D. "The biggest single advantage we have is cost. With OpenVista, you don't have the excessive fees associated with licensing proprietary software."

Kizer headed the VA health system from 1994 through 1999 and pushed development of what is currently the largest EHR system in the country. Because VistA was created with taxpayer dollars, the source code is in the public domain. An active developer and user community is constantly improving the open source system. Medsphere was founded in 2002 to rework VistA for the open-source Linux operating system.

At Midland Memorial, director of pharmacy services John Harrington is less concerned with cost than with functionality. Like most serious EHR systems, OpenVista combines patient charting, electronic order entry, pharmacy, labs, nursing, admitting, billing, and other hospital functions into a single interoperable system. Midland's old system, which was slated to lose vendor support in 2006, had standalone modules for pharmacy and other departments. Some department systems were interoperable but most were not. After 14 months of a planned 16-month rollout, Harrington said he is still looking for the downside to EHR. "The biggest advantage we now have is access to the entire patient medical record whenever there is a question," he said. "And we haven't even gotten to being able to do broad data mining and population studies."

Harrington could be even happier once he starts prospecting for high-risk patients. In Danville, Pa., Geisinger Health System tripled the rate of osteoporosis screening and treatment by mining EHR data for high-risk patients. And an earlier EHR study more than doubled the rate of influenza and pneumococcal vaccination among its high-risk patients.

Data mining is a simple concept, noted Geisinger researcher William Ayoub, M.D. Pick a condition with well-defined risk factors or treatment recommendations such as osteoporosis or influenza immunization, then screen the entire patient population electronically. High-risk patients are contacted by mail or phone and encouraged to come in for appropriate screening and/or treatment. The entire process can be automated, driving efficiency up and costs down.