The impact of a computerized physician order entry system on medical errors with antineoplastic drugs 5 years after its implementation

Summary What is known and objective Computerized physician order entry (CPOE) systems reduce medical errors (MEs). Nevertheless, a CPOE system may also lead to new types of errors, especially when it is first implemented. The objectives of this study were to determine the impact of a CPOE on the number of MEs and to identify the types of MEs in prescriptions issued by the Haematology Department 5 years after the implementation of the CPOE system. Methods We conducted a prospective analytical study on the implementation of a CPOE system at the Pharmacy Department of the Hospital Ramon y Cajal (Madrid, Spain). The study comprised three phases: a pre‐implementation phase, an implementation phase conducted in the Haematology Department and a post‐implementation phase, which was conducted 5 years after the implementation of the CPOE system. One hundred and fifty prescriptions per pre‐ and post‐implementation phase were consecutively included in the study. A previously described classification scheme was used to detect and classify MEs. Results and discussion The implementation of a CPOE system was associated with a large reduction in MEs. One hundred and fourteen patients (pre‐implementation phase) were compared to 82 patients (post‐implementation phase). The total number of MEs per 100 patients decreased from 236·8 (95% CI: 212·1–261·3) to 10·9 (95% CI: 5·8–19·6), with an absolute risk reduction of 36·2 (95% CI: 32·6–39·9). The percentage of prescriptions with an ME decreased from 37·5% to 1·2% (P