The issues surrounding computerized physician order entry, or CPOE, are so complex that even what counts as a full-fledged system or implementation is the topic of much debate. The Leapfrog Group, for instance, reports that just over two percent of American hospitals can boast of a complete CPOE system, but critics say that the group’s benchmarks are too strict. First Consulting Group puts the number a bit higher, at five percent. A more recent analysis by A.O. Jech says that while 1 in 3 hospitals have implemented at least partial systems, less than 10 percent of the orders are being entered on those systems and are used by less than 10 percent of physicians.

Now a study due to be published in the Marsh/April issue of the Journal of the American Medical Informatics Association says that almost one in ten hospitals have implemented CPOE. “I think safety and overall efficiency are the major drivers (for adoption,” says one of the new study’s authors, Joan S. Ash, an associate professor at the Oregon Health and Science University. “It’s more efficient for the hospital even if it’s not more efficent for the individual physician. The big impediment is that physicians simply don’t have time for it and the benefits to them don’t make up for the time it takes.”

Ash’s point is echoed throughout academia and the healthcare IT industry. Although CPOE’s benefits are widely known – one system at Brigham and Women’s Hospital reduced error rates by 55 percent and serious medication errors by 88 percent; another study (at Wishard Memorial Hospital in Indianapolis) showed that CPOE could reduce patient stays by nearly a full day and reduce charges by 13 percent — there is resistance from physicians because of entry times and poor work-flow models.

Last year’s physician revolt at Cedars-Sinai is a perfect example. Former chief of staff Stephen Uman told the American Medical News that hundreds of physicians complained that the CPOE system administrators chose “was poorly designed, slowed patient care and was dangerous because orders weren’t being transmitted or were getting lost.” Ultimately, doctors forced the system off-line.

Uman said they won’t miss it. “For somebody ... who has 15 to 20 patients in the hospital, when you add three or four minutes per order or five to 10 minutes per patient, that comes up to two or three hours of extra time a day,” he said.

In the Winter 2004 issue of the Journal of Healthcare Information Management, Patricia P. Sengstack and Brian Gugerty observe that because “ it is a given that work processes will change, a plan to analyze and work through these changes is necessary. Response times, ordering times and communication time all need to be reviewed in terms of what effect CPOE will bring.”

But not everyone thinks that a CPOE system will wreak havoc with existing workflows. Marie DiFran-cesco and Terri Andrews, writing in the same issue of JHIM, describe the CPOE implementation at Alamance Regional Medical Center in far more glowing terms. They say “virtually all of ARMC’s 240 affiliated physicians get test results online, with more than 80 percent of the physicians entering orders for medications, lab and radiology tests through the CPOE system... Doctors are increasingly relying on CPOE as part of a ‘best practices’ model in medicine.”

As just two measures of the system’s success, ARMC has seen a 58 percent drop in laboratory duplicate testing, saving the hospital $110,000 every year. Total order time shave been reduced from 86 minutes to just three minutes “due to the instantaneous transmission of the order to the pharmacy.”

Ash points to two success stories: El Camino Hopsital in Mountain View, Calif., and the Kaiser Permante Northwest Epic system. She says that the differentiator in these systems isn’t the elegance of the CPOE interface as it is the system’s fucntionality. “We’ve observed physicians happily using old ugly interfaces,” Ash says. What counts is “the overall ease of use of the system, as well as the speed.”