关键词:  , MedDRA；术语；翻译错误；语义冲突；关联分析

Abstract:

Objective： To detect the semantic conflicts between the Chinese and English versions of the terms with the same code in the English-Chinese translation of the Medical Dictionary for Regulatory Activities （MedDRA） based on the association analysis algorithm， and correct translation errors.Methods： Taking a total of 78 283 Lowest Level Terms （LLT） in MedDRA 26.1 Chinese version as the research object， with MedDRA 26.1 English version as the correct reference， the research process includes 4 steps， which are determining and classifying Chinese terms for verification， analyzing the association between Chinese and English texts， customizing filtering rules， and manually verifying the results. The criterion for suspected errors in English translation to Chinese is when terms with the same code exhibit conflicting semantics in Chinese compared to English， or when Chinese terms fail to reflect strong associative semantics present in English.The algorithm of association strength adopts the fourfold table method and the information component that are commonly used in pharmacovigilance signal detection.Results： Through association analysis， rule filtering and manual verification， a total of 11 mistranslated terms with semantic conflicts and 8 terms with poor translation accuracy in the Chinese version of MedDRA 26.1 were found in this study. The mistranslated and poorly translated terms in English-Chinese translation involve the English roots "hyper-" and "hypo-" as well as the English words "increased" and "decreased".Conclusion： In this study， the mistranslation problems were discovered using the semantic conflicts between the Chinese translations and the original English texts of the MedDRA terms with the same code through the association analysis algorithm. The approach is simple and effective， and can provide feasible technical support for the maintenance of the MedDRA term set and improve the semantic consistency among the multilingual versions of MedDRA.

Key words: MedDRA； , Terminology； , Mistranslation； , Semantic conflicts； , Association analysis