Decision support could help doctors catch "zebras"

Internist Michael Barr, vice president for practice advocacy and improvement at the American College of Physicians, says there's a place for decision support software in EHRs. Although he doesn't endorse any particular program, he believes there's value in the kind of software that can help jog physicians' memories about unusual diagnoses at the point of care. Among the programs designed to do this are Isabel, Iliad, D-Explain, and QMR (InfoTech Bulletin, June 8, 2007).

"A lot of medicine is pattern recognition, pulling together pieces of a puzzle and saying, 'aha, this is the differential diagnosis that comes up,'" notes Barr. "Then add a physician's judgment, and you get the filter that only comes with experience and training. A certain kind of clinical decision support can help identify unusual conditions-sometimes referred to as 'zebras.' These utilities are available now on handheld computers, and we hope that eventually, they'll be integrated in EHRs so they become part of the clinician workflow."

Barr believes physicians could also benefit if they could access clinically relevant information and guidance at the point of care, such as the tools in the ACP's Physician Information and Education Resource (PIER). And he'd like to see EHR makers include other forms of clinical decision support, as well. EHRs could generate "recommendations based not just on clinical indicators and age and sex but also on patient preferences," he says. For example, a clinical decision support tool could advise a physician to prescribe an alternative medication regimen if a patient couldn't follow the most common recommendation because of dementia or other patient-specific factors.

The problem with any point-of-care decision support tool, however, is that it tends to slow physicians down. "In a fee-for-service environment, you're paid on how many patients you see, not necessarily on how well you do with respect to quality," notes Barr. "Hopefully, as we move toward the recognition of what it takes to produce high-quality outcomes, it'll make sense for physicians to use the kinds of tools that are out there and those that have yet to be developed."