Abstract
Japanese nurses are now being required to use research in clinical practice. As a result, they increasingly need to read nursing research literature written in English. Online machine translation is a convenient tool that may address any existing language barrier. The quality of machine translation has been evaluated using various methods; however, its reliability for nursing literature is unknown. In this study, existing methods for evaluating online machine translation quality were examined for structural accuracy and intelligibility of translations of nursing literature, and the usability of machine translations is discussed. In total, 297 English sentences from nursing study abstracts were evaluated for accuracy and intelligibility; 75 word items were identified as specific terms. Interrater reliability of the evaluation method was measured using two raters. The results showed the strong reliability of the existing evaluation method. Also, a possibly usable online machine translation system being offered in Japan was identified. Specific terms seemed to influence the extent of the intelligibility of the translations. Future studies need to include more samples of the nursing population to examine general perceptions of the usability of online machine translation systems for reading English nursing literature among the Japanese nursing population.