Similarity of Drug Names and Confusion Errors:

Laboratory Experiments with Students and Pharmacists

Abstract

The purpose of this study was to test whether the probability of confusion errors is related to the similarity indices calculated by the formulas: “cos1”, “htco”, and “edit” (Tsuchiya et al., 2001). A pair of drugs was displayed right after a brief exposure of either one of the drugs and two numerals on the computer screen. The subject's task was first to answer the sum of the figures and then to identify which drug was displayed. Students and pharmacists participated in the experiments. The results showed that the rate of confusion errors increased as a function of the degree of similarity except for index “edit”. It was also demonstrated that the error rate was affected by the number of shared characters in the beginning and the ending of drug names. Further analysis of the results suggested the necessity of examining the effect of the coincidence of special features in a character and the effect of the similarity of the form of a character.