Review and Analysis of Massively Registered Clinical Trials of COVID-19 using the Text Mining Approach

Objective: Immediately after the outbreak of nCoV, many clinical trials are registered for COVID-19. The numbers of registrations are now raising inordinately. It is challenging to understand which research areas are explored in this massive pool of clinical studies. If such information can be compiled, then it is easy to explore new research studies for possible contributions in COVID-19 research. Methods: In the present work, a text-mining technique of artificial intelligence is utilized to map the research domains explored through the clinical trials of COVID-19. With the help of the open-- source and graphical user interface-based tool, 3007 clinical trials are analyzed here. The dataset is acquired from the international clinical trial registry platform of WHO. With the help of hierarchical cluster analysis, the clinical trials were grouped according to their common research studies. These clusters are analyzed manually using their word clouds for understanding the scientific area of a particular cluster. The scientific fields of clinical studies are comprehensively reviewed and discussed based on this analysis. Results: More than three-thousand clinical trials are grouped in 212 clusters by hierarchical cluster analysis. Manual intervention of these clusters using their individual word-cloud helped to identify various scientific areas which are explored in COVID19 related clinical studies. Conclusion: The text-mining is an easy and fastest way to explore many registered clinical trials. In our study, thirteen major clusters or research areas were identified in which the majority of clinical trials were registered. Many other uncategorized clinical studies were also identified as “miscellaneous studies”. The clinical trials within the individual cluster were studied, and their research purposes are compiled comprehensively in the present work.