Computerized physician order entry impact on prescription medication errors in a pediatric unit

Background: Medication errors incidence is an important quality measure, and Hospital Pediatric departments are a particularly vulnerable area for medication errors. The adoption of Computerized Prescription Order Entry (CPOE) systems is generally associated with improvements in several medication-related quality measures. Objective: To evaluate the impact of a CPOE system on the medication ordering process’s frequency of errors two years after changing from a handwritten system. Setting: Data were collected both retrospectively and prospectively from 341 admissions in a pedriatic ward of a teaching hospital. Method: To perform this ambispective non-controlled interventional study pharmacists analyzed the medicine’s prescriptions over three 30-day different periods: the first one with handwritten prescriptions (historical, phase 1), a second just after launching a CPOE system (historical, phase 2), and a third two years after its implementation (prospective phase 3). Main outcome measure: The number of errors per 100 drug prescriptions was compared across the study’s phases. Results: Pharmacists identified 20 prescription errors among 403 drug orders (5 errors per 100 prescriptions) during phase 1. During the early phase of CPOE implementation, 37 errors were detected among 562 prescriptions (6.6 errors per 100 prescriptions), and during the late phase, 34 errors among 573 (5.9 errors per 100 prescriptions). Differences were not statistically significant. Dose-related errors (improper dose and wrong dose units) were significantly reduced from 2.7 per 100 prescriptions during phase 1 to 1.4 per 100 prescriptions during phase 3 (p <0.05). The use of CPOE leads to different types of errors being made by staff. Conclusion: The rate of prescription errors found in the present study for handwritten prescriptions was deemed lower than expected. In this context, the introduction of a CPOE system resulted in a slightly worse rate of prescription errors than the previous period using a handwritten system. However, the intervention reduced the incidence of dose-related errors