How should medication errors be defined? Development and test of a definition

Aims: Definitions of medication errors vary widely in the literature, and prevalence from 2-75% in part because of this lack of consensus. Thus, clarification of the concept is urgently needed. The objective was to develop a clear-cut definition of medication errors and specify relevant error types in the medication process. Methods: Based on existing taxonomy and through a modified Delphi-process consensus of definition and error types were reached among Danish experts appointed by 13 healthcare organisations and the project group. The experts prioritised five definitions of medication errors and score the relevance of 76 error types. Based on explicit criteria, the project group settled non-consensus cases. Results: The panel consisted of 12 physicians, seven pharmacists, and six nurses. Consensus was reached for the definition An error in the stages of the medication process - ordering, dispensing, administering and monitoring the effect - causing harm or implying a risk of harming the patient. Moreover, consensus for 60 of 76 error types was achieved. Applied to a historic dataset the definition reduced the number of medication errors from 34% to 7%. Conclusions: Experts deemed a definition using harm or risk of harm as cut-off point as the most appropriate in Danish hospital settings. In addition, they agreed on a list of 60 error types covering the medication process. Interestingly, a substantial lower occurrence of medication errors was found when applied to historic data. The definition is in accordance with international taxonomy, thus is assumed to be applicable to modern healthcare settings abroad.