Issue link: https://www.e-digitaleditions.com/i/1494960
12 Pharmaceutical Technology ® Quality and Regulatory Sourcebook March eBook 2023 PharmTech.com IPOPBA - STOCK.ADOBE.COM A s the use of machine learning (ML) in pharmacovigilance grows, so does the need for large and comprehensive data sets to ensure reliable and accurate mod- els. However, obtaining and maintaining such data can be a challenge, as data gaps can arise for a variety of reasons. These gaps can have significant impli- cations for the accuracy and effectiveness of ML models, and it is therefore important to explore new strategies for closing these gaps. Data gaps can pose challenges to the development of reliable ML models in pharmacovigilance. There are various strategies that organizations can use to overcome these challenges, including active surveil- lance, using diverse and representative data sources, employing data augmentation techniques, and im- plementing data governance and management prac- tices. By leveraging these strategies, organizations can effectively close data gaps and achieve more accu- rate and effective ML models in pharmacovigilance. Choosing the right technology partner is important when implementing data-driven technologies in pharmacovigilance and on the need for continuous monitoring, evaluation, and improvement of the mod- els. Overcoming data gaps and achieving reliable ML models in pharmacovigilance can be done through a combination of data management, governance prac- tices, and effective technology partnerships. Strategies for closing data gaps The following are some strategies for closing data gaps in ML for pharmacovigilance. Automated data extraction with active surveillance. Automated data extraction with living review, more popularly known as active surveillance, is an import- ant tool. The living review approach is a strategy that Overcoming Data Gaps in Pharmacovigilance Pankaj Bhardwaj, serves as a senior product manager, and Ryanka Chauhan serves as a product manager; both at Datafoundry for DF mSignal AI. By leveraging certain strategies, organizations can effectively close data gaps and achieve more accurate and effective machine learning models in pharmacovigilance.