Effect of a computerized prescriber-order-entry system on reported medication errors

The Institute of Medicine recently recommended that hospitals adopt computerized prescriber-order-entry (CPOE) systems with medication clinical decision support.1 In response to this recommendation, and with the intent of decreasing medication errors and improving patient safety, the University of North Carolina (UNC) Hospitals began implementing a CPOE system in October 2002. Some early attempts to implement CPOE at other institutions have failed, in large part because of physician dissatisfaction.2–,5 At one academic health care center, physicians registered their protest by writing entire orders in the “comments” section of computerized orders, thereby bypassing the potential of the system to check for the appropriate medication and dosage.2 Other systems, such as the one developed for Brigham and Women’s hospital in the 1990s, were incrementally implemented, providing an opportunity to demonstrate that CPOE was associated with an overall decrease in medication errors despite the introduction of some system-specific problems.3 CPOE has also been shown to be associated with shortened hospital stays and decreased costs.4,5