No abstract

Systems reviewed in this manuscript have strengths such as clinical detail, a large enough sample size to capture rare adverse events, and/or a patient denominator internal to the database. Few systems include all of these attributes. Pediatric drug safety would be better informed by utilizing multiple systems to take advantage of their individual characteristics.

Introduction The US FDA is interested in a tool that would enable pharmacovigilance safety evaluators to automate the identification of adverse drug events (ADEs) mentioned in FDA prescribing information. The MITRE Corporation (MITRE) and the FDA organized a shared task-Adverse Drug Event Evaluation (ADE Eval)-to determine whether the performance of algorithms currently used for natural language processing (NLP) might be good enough for real-world use. Objective ADE Eval was conducted to evaluate a range of NLP techniques for identifying ADEs mentioned in publicly available FDA-approved drug labels (package inserts). It was designed specifically to reflect pharmacovigilance practices within the FDA and model possible pharmacovigilance use cases. Methods Pharmacovigilance-specific annotation guidelines and annotated corpora were created. Two metrics modeled the experiences of FDA safety evaluators: one measured the ability of an algorithm to identify correct Medical Dictionary for Regulatory Activities (MedDRA ®) terms for the text from the annotated corpora, and the other assessed the quality of evidence extracted from the corpora to support the selected MedDRA ® term by measuring the portion of annotated text an algorithm correctly identified. A third metric assessed the cost of correcting system output for subsequent training (averaged, weighted F1-measure for mention finding). Results In total, 13 teams submitted 23 runs: the top MedDRA ® coding F1-measure was 0.79, the top quality score was 0.96, and the top mention-finding F1-measure was 0.89. Conclusion While NLP techniques do not perform at levels that would allow them to be used without intervention, it is now worthwhile exploring making NLP outputs available in human pharmacovigilance workflows.

Introduction On 4 February, 2020, the Secretary of the Department of Health and Human Services declared a public health emergency related to coronavirus disease 2019 (COVID-19), and on 27 March, 2020 declared circumstances existed to justify the authorization of the emergency use of drug and biological products (hereafter, “drugs”) for COVID-19. At the outset of the pandemic with uncertainty relating to the virus, many drugs were being used to treat or prevent COVID-19, resulting in the US Food and Drug Administration’s (FDA’s) need to initiate heightened surveillance across these drugs. Objective We aimed to describe the FDA’s approach to monitoring the safety of drugs to treat or prevent COVID-19 across multiple data sources and the subsequent actions taken by the FDA to protect public health. Methods The FDA conducted surveillance of adverse event and medication error data using the FDA Adverse Event Reporting System, biomedical literature, FDA-American College of Medical Toxicology COVID-19 Toxicology Investigators Consortium Pharmacovigilance Project Sub-registry, and the American Association of Poison Control Centers National Poison Data System. Results From 4 February, 2020, through 31 January, 2022, we identified 22,944 unique adverse event cases worldwide and 1052 unique medication error cases domestically with drugs to treat or prevent COVID-19. These were from the FDA Adverse Event Reporting System (22,219), biomedical literature (1107), FDA-American College of Medical Toxicology COVID-19 Toxicology Investigator’s Consortium Sub-registry (638), and the National Poison Data System (32), resulting in the detection of several important safety issues. Conclusions Safety surveillance using near real-time data was critical during the COVID-19 pandemic because the FDA monitored an unprecedented number of drugs to treat or prevent COVID-19. Additionally, the pandemic prompted the FDA to accelerate innovation, forging new collaborations and leveraging data sources to conduct safety surveillance to respond to the pandemic.