In Joseph Conn's "Defending CPOE," which appeared in the March 14, 2005 issue of Modern Healthcare, Conn quotes Suzanne Delbanco, chief executive officer of the Leapfrog Group, regarding a recent study reported in the Journal of the American Medical Association showing computerized physician order entry induced medication errors.
Delbanco comments that the studied CPOE system was "quite an old system" and that "no system is any better than how it is designed and implemented." It is this quality of analysis and perspective that is so unsettling to some in the medical community.
The first problem is that CPOE advocates report studies that compare a focused application of a current and optimized CPOE version with a non-CPOE system that is weak in design. Why should demonstrably flawed albeit suboptimal CPOE systems be immune from critical analysis using a defense that argues "it's not the system, it's the application" when the non-CPOE systems to which they are compared are not?
A second problem is that CPOE advocates apparently dismiss the paint-on-the-hands test for CPOE. Comparing a recent generation CPOE system in an idealized application environment and then extrapolating that data to the real world of imperfectly applied, incompatible, previous-generation CPOE applications strikes some experienced people as inherently flawed science if not disingenuous.
The Food and Drug Administration has approved many medications because they were shown to save lives. Some of these have later been taken off the market because they also killed many people. Let's be a little more discerning regarding the reported benefits of CPOE.
Coke R. Smith, M.D., medical director, Sunnyside Community Hospital