Identification of Adverse Drug Event–Related Japanese Articles: Natural Language Processing Analysis

Abstract

Background Medical articles covering adverse drug events (ADEs) are systematically reported by pharmaceutical companies for drug safety information purposes. Although policies governing reporting to regulatory bodies vary among countries and regions, all medical article reporting may be categorized as precision or recall based. Recall-based reporting, which is implemented in Japan, requires the reporting of any possible ADE. Therefore, recall-based reporting can introduce numerous false negatives or substantial amounts of noise, a problem that is difficult to address using limited manual labor. Objective Our aim was to develop an automated system that could identify ADE-related medical articles, support recall-based reporting, and alleviate manual labor in Japanese pharmaceutical companies. Methods Using medical articles as input, our system based on natural language processing applies document-level classification to extract articles containing ADEs (replacing manual labor in the first screening) and sentence-level classification to extract sentences within those articles that imply ADEs (thus supporting experts in the second screening). We used 509 Japanese medical articles annotated by a medical engineer to evaluate the performance of the proposed system. Results Document-level classification yielded an F1 of 0.903. Sentence-level classification yielded an F1 of 0.413. These were averages of fivefold cross-validations. Conclusions A simple automated system may alleviate the manual labor involved in screening drug safety–related medical articles in pharmaceutical companies. After improving the accuracy of the sentence-level classification by considering a wider context, we intend to apply this system toward real-world postmarketing surveillance.

How can natural language processing help model informed drug development?: a review

“…NER identifies medications and their attributes (dosage, route, duration, and frequency), indications, ADEs, and severity. 129 , 130 Word Sense Disambiguation is used to further filter the identified entities and confirm their contextual sense. The relation extraction task identifies relations between the named entities: medication-indication and medication-ADE.…”

“…The relation extraction task identifies relations between the named entities: medication-indication and medication-ADE. 129 Word embeddings are utilized to vectorize the input for training an ML 130 or DL 131 model to identify and classify ADEs. Numerous publicly available NLP systems have been extended to perform ADE detection tasks, including MedLEE, 132 MetaMap, 25 cTAKES, 27 MedEx, 126 and GATE.…”

Natural Language Processing: from Bedside to Everywhere

“…Another simple approach to ADE detection is classification-based IE, which detects ADE information mentioned in a document by sentence- [110,111] or document-level [111] classification. For instance, Ujiie et al [111] proposed a machine learning-based method to first classify each sentence of case reports into ADE-suggesting or not, and then to identify the documents that report any ADEs based on the sentence-level classification results.…”

Year 2020 (with COVID): Observation of Scientific Literature on Clinical Natural Language Processing

“…Indeed, there is a significant number of existing corpora, datasets and resources available in English. Yet, we observe an increasing number of publications dedicated to other languages and a greater variety of languages: Arabic [ 20 ], Chinese [ 21 22 23 24 25 26 ], Croatian [ 27 ], Finnish [ 28 , 29 ], French [ 30 , 31 ], German [ 32 33 34 ], Hebrew [ 35 ], Italian [ 36 37 38 ], Japanese [ 39 , 40 ], Korean [ 41 , 42 ], Norwegian [ 43 ], Portuguese [ 44 ], Spanish [ 45 46 47 48 ], Swedish [ 49 ], and Turkish [ 28 ]. Overall, we believe that the trend observed in previous years is continuing.…”