Analysis of human errors caused by radiological technologists during radiotherapy with using of the questionnaire and text mining

Abstract

Modern radiotherapy has been improved and diversified, however, various human errors occur during treatment. Relationship between errors and workload was examined using text mining. A questionnaire survey was conducted executed with radiological technologists (n=665) that were working in Gunma Prefecture (n=553) and at hospitals affiliated with a university in Tokyo (n=112), in order to investigate errors made by the technologists. The results were analyzed using morphological analysis software "Win Cha 2000" and morphemes were extracted. Cluster analysis was also conducted. The results indicated that the errors of over, or under irradiation caused by input mistakes were the most frequent type of error. Morphemes of error conditions were "patient," "irradiation," and "therapy" according to the order of the frequency of these errors, and morphemes of the cause of error were "patient," "confirmation," and "irradiation." The method of analysis used in the present study made it possible to analyze the causes and conditions of errors both quantitatively and qualitatively.