Bibliometric Analysis of Publications on Clinical Studies Leveraging Natural Language Processing During 2000 - 2023

The number of clinical studies using natural language processing is quite large. Therefore, it is important to examine in depth the development of clinical studies using Natural Language Processing over the years. However, there are a limited number of studies in the literature examining the research status of this field. The article presents a bibliometric analysis of studies on the keywords "clinical AND studies AND natural AND language AND processing" indexed in Scopus between 2000 and 2023. This study aims to evaluate academic outputs in the relevant field quantitatively, make sense of the data, reveal the state of scientific knowledge in the field, and give scientists a general perspective on the subject. Bibliometrix and Microsoft Excel programs were used for bibliometric analysis. Nineteen thousand two hundred seventy-three different authors identified a total of 4535 studies. 77.5% of these studies were research articles (3516), 14.8% were conference papers (669), 6.8% were reviews (307), and 0.9% were book chapters (43). Journal of Biomedical Informatics was the journal in which the most studies were published, with 226 articles. Only the United States (2637) contributed 58.1% to the studies. Liu, H. was the most prolific author, with 85 articles. Harvard Medical School was the most productive institution, with 304 studies. The most cited article was Discontinuation of Statins in Routine Care Settings, A cohort study.