EXPLORING THE POSSIBILITIES OF AUTOMATED MONITORING OF MEDICAL AND PHARMACEUTICAL SCIENTIFIC LITERATURE FOR PHARMACOVIGILANCE PURPOSES

Abstract

The aim of the work . To assess the potential and effectiveness of automated monitoring of medical and pharmaceutical literature for pharmacovigilance purposes. The research evaluates the impact of automation on analysis duration, accuracy of identifying relevant information, and reduction of workload for specialists, based on data collected from Ukrainian periodicals published in 2024.

Materials and Methods. The analysis included 113 Ukrainian medical and pharmaceutical open-access periodicals actively published during 2024. Optical Character Recognition (OCR) and text analysis methods were applied to automatically detect mentions of medicinal products. Time expenditure for data processing was compared between manual and automated approaches.

Results and Discussion. A total of 482 journal issues were analyzed. The automated approach reduced the average time required to process a single article from 11 to 1.5 minutes, resulting in a sevenfold decrease in time expenditure. The study also examined the frequency of journal issue publication and compared periodicals with the highest and lowest publication rates in 2024. At the same time, the importance of retaining the role of a specialist for result validation and decision-making was emphasized. Key advantages of automation were identified, including improved responsiveness, reduced risk of human error, and continuous monitoring capabilities.

Conclusions. The study demonstrated that automation of medical and pharmaceutical literature monitoring can significantly reduce time costs (by a factor of seven) and lower the risk of human error. Future research may focus on optimizing detection algorithms, improving the handling of diverse data formats, and integrating automated systems into the broader information infrastructure of pharmaceutical companies.