Pharmacovigilance in the Age of Artificial Intelligence: A Theoretical Account of Signal Detection, Real-World Data and Drug Safety Surveillance in Under-Reporting Environments
DOI:
https://doi.org/10.68854/c282yc42Keywords:
pharmacovigilance, drug safety, adverse drug reactions, signal detection, artificial intelligence, naturalAbstract
Drug-safety surveillance depends on reports of adverse reactions, and most African health systems generate such reports rarely and late. A safety signal that would be detected within months in a high-income country may take years to emerge in these settings, or may never emerge at all. Artificial intelligence (AI) offers to detect signals from other sources, such as electronic health records, social media, pharmacy and insurance data and clinical text, and to triage reports faster. This theoretical paper develops an account of what AI can contribute to pharmacovigilance where the reporting base is thin. It draws on pharmacoepidemiology and regulatory science, on the literature on pharmacovigilance systems, and on recent research on machine learning and natural language processing for adverse-event detection and signal management. The analysis separates three functions that pharmacovigilance must perform: generating reports, where the data do not exist; detecting signals, where the data exist but the pattern is hidden; and acting on signals, where the pattern is found but the regulator does not respond. AI addresses the second function well, the first only where alternative data sources exist, and the third not at all. On this basis the paper advances a set of propositions concerning the data sources and methods suited to under-reporting environments, the risk of false signals from unrepresentative digital data, and the regulatory capacity needed to make AI-enabled surveillance consequential. Nigeria's pharmacovigilance system anchors the analysis, and substandard and falsified medicines are considered as a surveillance problem that AI may help address. The paper concludes that detection improved by artificial intelligence becomes consequential only where the regulatory capacity to investigate and act on signals already exists, so that investment in that capacity is a precondition for, rather than a complement to, investment in AI tools.
References
Chabalenge, B., Sahota, T., Ermolina, I., & Tanna, S. (2025). Substandard and falsified medicines in Africa: Healthcare systems challenges, supply chain issues, regulatory challenges and strategies to increase access to quality medicines. Frontiers in Pharmacology, 16, 1708784. https://doi.org/10.3389/fphar.2025.1708784 de Oliveira, P. M. N., Hartmann, K., Bhamare, C., & Lucchesi, M. B. B. (2025). Active vaccine safety surveillance in low- and middle-income countries: Challenges for vaccine manufacturers from emerging countries. Vaccine, 48, 126727. https://doi.org/10.1016/j.vaccine.2025.126727
Denck, J., Ozkirimli, E., & Wang, K. (2023). Machine-learning-based adverse drug event prediction from observational health data: A review. Drug Discovery Today, 28(9), 103715. https://doi.org/10.1016/j.drudis.2023.103715
Dut, G. M. (2026). Governing AI for pharmacovigilance in low-income countries: Systems perspective. Journal of Medical Internet Research, 28, e96374. https://doi.org/10.2196/96374
Golder, S., O'Connor, K., Wang, Y., Klein, A., & Gonzalez Hernandez, G. (2024). The value of social media analysis for adverse events detection and pharmacovigilance: Scoping review. JMIR Public Health and Surveillance, 10, e59167. https://doi.org/10.2196/59167
Golder, S., Xu, D., O'Connor, K., Wang, Y., Batra, M., & Gonzalez Hernandez, G. (2025). Leveraging natural language processing and machine learning methods for adverse drug event detection in electronic health/medical records: A scoping review. Drug Safety, 48(4), 321–337. https://doi.org/10.1007/s40264-024-01505-6
Hu, Q., Chen, Y., Zou, D., He, Z., & Xu, T. (2024). Predicting adverse drug event using machine learning based on electronic health records: A systematic review and meta-analysis. Frontiers in Pharmacology, 15, 1497397. https://doi.org/10.3389/fphar.2024.1497397
Kiguba, R., Zakumumpa, H., Ndagije, H. B., Mwebaza, N., Ssenyonga, R., Tregunno, P., Harrison, K., & Pirmohamed, M. (2023). Facilitators and barriers to uptake of the Med Safety mobile app for adverse drug reaction reporting by health workers in Uganda: A qualitative study. Drug Safety, 46(6), 565–574. https://doi.org/10.1007/s40264-023-01303-6
Menang, O., Kuemmerle, A., Maigetter, K., & Burri, C. (2023). Strategies and interventions to strengthen pharmacovigilance systems in low-income and middle-income countries: A scoping review. BMJ Open, 13(9), e071079. https://doi.org/10.1136/bmjopen-2022-071079
Menang, O., van Eeuwijk, P., Maigetter, K., de Soyres-Kuemmerle, A., Agbenu, E., & Burri, C. (2025). Building functional and sustainable pharmacovigilance systems: An analysis of pharmacovigilance development across high-, middle- and low-income countries. Therapeutic Advances in Drug Safety, 16, 20420986251342941. https://doi.org/10.1177/20420986251342941
Moyo, E., Moyo, P., Mangoya, D., Imran, M., & Dzinamarira, T. (2023). Adverse drug reaction reporting by healthcare providers in sub-Saharan Africa: A scoping review of the challenges faced and the strategies to address the challenges. International Journal of Africa Nursing Sciences, 19, 100639. https://doi.org/10.1016/j.ijans.2023.100639
Ndagije, H. B., Walusimbi, D., Atuhaire, J., & Ampaire, S. (2023). Drug safety in Africa: A review of systems and resources for pharmacovigilance. Expert Opinion on Drug Safety, 22(10), 891–895. https://doi.org/10.1080/14740338.2023.2251375
Noguchi, Y., Tachi, T., & Yoshimura, T. (2025). Is it appropriate to conduct a disproportionality analysis using a spontaneous reporting database to investigate whether drug-related adverse events are dose-dependent? Frontiers in Pharmacology, 16, 1563524. https://doi.org/10.3389/fphar.2025.1563524
Ogar, C. K., Sisay, M. M., Roque-Pereira, L., Leopold, C., Souverein, P. C., Mantel-Teeuwisse, A. K., & De Bruin, M. L. (2026). Evaluating the (comparative) safety profile of the novel oral polio vaccine type 2 using individual case safety reports in VigiBase. British Journal of Clinical Pharmacology. Advance online publication. https://doi.org/10.1002/bcp.70580
Oyebode, O., & Orji, R. (2023). Identifying adverse drug reactions from patient reviews on social media using natural language processing. Health Informatics Journal, 29(1), 14604582221136712. https://doi.org/10.1177/14604582221136712
Painter, J. L., Chalamalasetti, V. R., Kassekert, R., & Bate, A. (2025). Automating pharmacovigilance evidence generation: Using large language models to produce context-aware structured query language. JAMIA Open, 8(1), ooaf003. https://doi.org/10.1093/jamiaopen/ooaf003
Sankar, C., Meyer, J. C., Schönfeldt, M., Gunter, H., Dawood, H., Sekiti, V., Pickard, N., Mubaiwa, L., Mawela, D., Dlamini, S., Peter, J., Spencer, D., Gray, C., Patel, V., Bamford, L., Sehloho, T., & McCarthy, K. (2025). Vaccine safety surveillance in South Africa through COVID-19: A journey to systems strengthening. Vaccine, 46, 126535. https://doi.org/10.1016/j.vaccine.2024.126535
Uesawa, Y. (2026). Interpreting positive and inverse disproportionality signals in spontaneous reporting systems: A claim-strength framework with direction-aware evaluation. Drug Safety. Advance online publication. https://doi.org/10.1007/s40264-026-01726-x von Kreijfelt, G., Zwaveling, J., van Lint, J. A., Visser, L. E., & Kant, A. C. (2026). Strengthening signal detection by proactive adverse event case identification in electronic health records using a text-mining tool. Drug Safety. Advance online publication. https://doi.org/10.1007/s40264-026-01739-6
Winkler Martinez, E., Tfayli, A., Chaminade, P., & Tfaili, S. (2026). Vibrational spectroscopies for the identification of substandard and falsified drugs: Handheld and portable devices, multispectral and hyperspectral imagers. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 357, 127836. https://doi.org/10.1016/j.saa.2026.127836
Downloads
Published
Issue
Section
License
Copyright (c) 2026 IIAR Journal of Health, Medicine, and Biomedical Research
This work is licensed under a Creative Commons Attribution 4.0 International License.
