The aim of this research was to investigate trends in the occurrence of medication errors; namely, preparation by a pharmacist of a medicine that does not match that described on the medical script. First of all, the medication error was evaluated in terms of two categories applied to the character of the name of the medicine : the physical character (such as string length, phonologic and number of matching character in drag name) and the meaning character (such a medical property). The differences between medicines mistakenly prescribed and correctly described on the medical script were assumed as distances, and these distances were calculated in each category. These distances supposed similarity between medicines mistakenly prescribed and those described on the medical receipt. In addition, the relations between calculated distances and factors related to pharmacists, such as years of experience, were investigated. Consequently, it was found that the occurrence of medication errors depends on the length of the string; moreover, it was suggested that the possibility of influence by similarities of character, the phoneme and the medicinal effect depended on the experience of pharmacy staff. This means that the occurrence of medication error does not depend on one specific factor, and that it probability increases when certain factors are combined. Additional, the strategy of "reading" by pharmacists differ from the general it, and the strategy of pharmacists depend on the cognitive factors.
抄録全体を表示