We use cookies to make sure that our website works properly, as well as some ‘optional’ cookies to personalise content and advertising, provide social media features and analyse how people use our site. Further information can be found in our Cookies policy
Pharmacovigilance (PV) constitutes a fundamental pillar of global drug safety, encompassing the systematic detection, assessment, understanding, and prevention of adverse drug reactions (ADRs) and other medicine-related problems. Pharmacovigilance (PV) is an essential component of modern healthcare systems, designed to safeguard patients by continuously monitoring, detecting, and preventing adverse drug reactions (ADRs). Conventional PV practices, which rely heavily on spontaneous reporting systems, manual case assessments, and retrospective analysis, often encounter major shortcomings such as delayed identification of safety concerns, underreporting of cases, and an inability to efficiently analyze vast and heterogeneous data sources. With the rapid expansion of real-world evidence and digital health records, these limitations have created a pressing need for more advanced solutions. Artificial Intelligence (AI) has emerged as a transformative approach to address these challenges. By integrating machine learning (ML), natural language processing (NLP), and deep learning techniques, AI-driven systems can efficiently process large and complex datasets, including information from electronic health records, clinical trials, pharmacovigilance databases, and even patient-generated content from social media platforms. These technologies not only accelerate the detection of safety signals but also support the automation of individual case safety report (ICSR) management, improve cost-effectiveness, and facilitate active patient engagement through digital tools. Furthermore, AI offers opportunities for predictive pharmacovigilance, enabling early identification of at-risk populations and aligning drug safety monitoring with the principles of personalized medicine. Despite its promise, the widespread adoption of AI in PV is constrained by several obstacles. Issues such as inconsistent or biased datasets, lack of transparency in “black-box” algorithms, ethical concerns related to privacy and data protection, and uncertainties in regulatory acceptance remain critical challenges. In addition, the interoperability of AI systems with existing pharmacovigilance infrastructures and the risk of over-reliance on automation demand careful consideration. This paper provides a comprehensive analysis of the opportunities and challenges associated with the use of AI in pharmacovigilance. It also highlights current regulatory perspectives and underscores the importance of explainable AI models, harmonized international guidelines, and collaborative frameworks between industry, regulators, and healthcare providers. The integration of AI into pharmacovigilance represents not only a technological advancement but also a paradigm shift toward a more proactive, transparent, and patient-centered approach to drug safety.
Keywords
Pharmacovigilance (PV), Adverse Drug Reactions (ADRs), Artificial Intelligence (AI) Machine Learning (ML) ,Natural Language Processing (NLP), Drug Safety, Monitoring Signal Detection
Introduction
Pharmacovigilance (PV) is a vital discipline within pharmaceutical sciences and public health, dedicated to the continuous monitoring of drug safety across a product’s entire life cycle. Defined by the World Health Organization (WHO) as “the science and activities relating to the detection, assessment, understanding, and prevention of adverse effects or any other possible drug-related problems”, pharmacovigilance ensures that medicines remain not only effective but also safe for patients. It plays an indispensable role in strengthening trust in healthcare systems, regulatory authorities, and the pharmaceutical industry. The primary aim of PV is the early identification, assessment, and mitigation of adverse drug reactions (ADRs), which are a significant cause of morbidity, mortality, and healthcare expenditure worldwide. Indeed, ADRs account for 5–10% of hospital admissions and pose an increasing burden on public health systems, thereby making robust drug safety monitoring an international priority.
Traditional pharmacovigilance systems, which largely rely on spontaneous reporting mechanisms, retrospective analysis, and manual case review, have historically served as the backbone of drug safety. While they have successfully identified numerous drug-related safety concerns, these conventional approaches are increasingly limited in scope. Major shortcomings include underreporting of ADRs, delays in detecting potential safety signals, and the inability to efficiently analyze the massive, complex, and often unstructured healthcare data being generated in modern practice. With the rapid growth of global healthcare data arising from electronic health records (EHRs), patient registries, clinical trial databases, insurance claims, biomedical literature, and even patient discussions on social media traditional PV systems face an urgent need for technological reinforcement.
This is where Artificial Intelligence (AI) offers transformative potential. Broadly defined, AI encompasses computational methods that simulate human intelligence to perform tasks such as reasoning, learning, problem-solving, and decision-making. In healthcare, AI has already demonstrated success in diagnostic imaging, predictive analytics, personalized treatment strategies, and drug discovery. In pharmacovigilance, AI technologies—particularly machine learning (ML), natural language processing (NLP), and deep learning—are emerging as powerful tools to enhance drug safety monitoring. Machine learning algorithms can be trained on large pharmacovigilance datasets to detect hidden or subtle patterns in adverse event reporting that may indicate emerging safety signals. For example, ML systems are capable of analyzing millions of patient records or spontaneous reports to identify early risk factors for ADRs that may otherwise go unnoticed through traditional surveillance. Natural language processing extends this capability by enabling the automated analysis of unstructured safety data, such as clinical case narratives, electronic health record notes, and even patient experiences expressed on social media platforms. These tools unlock vast amounts of previously underutilized information, expanding the evidence base for signal detection. Deep learning models, inspired by the structure of the human brain, add yet another layer of sophistication by processing high-dimensional, complex datasets—such as genomic profiles or imaging data—to uncover rare but serious ADRs that are difficult to detect with conventional methods.
The integration of AI into pharmacovigilance represents not merely a technical enhancement but a paradigm shifts in drug safety surveillance. Traditionally, PV has been reactive—responding to reported adverse events after they occur. AI, however, offers the potential to transition towards a proactive and predictive system, capable of anticipating risks before they materialize. This aligns with the emerging principles of precision medicine, where drug safety and efficacy can be personalized to the genetic, demographic, and lifestyle profiles of individual patients. In this way, AI promises to make pharmacovigilance not only faster and more efficient but also more patient-centered. Nevertheless, despite its transformative potential, the adoption of AI in pharmacovigilance presents significant challenges. Issues of data quality, interoperability across healthcare systems, and the “black-box” nature of advanced AI algorithms pose serious barriers to trust and implementation. Ethical considerations—including patient privacy, informed consent, and algorithmic bias—must also be carefully addressed to ensure the responsible use of AI. Furthermore, regulatory frameworks are still evolving. Agencies such as the U.S. Food and Drug Administration (FDA), the European Medicines Agency (EMA), and the WHO are actively exploring standards to validate, monitor, and govern AI-driven pharmacovigilance systems, emphasizing the need for transparency, accountability, and human oversight. Given this context, the present paper seeks to explore the dual dimensions of AI in pharmacovigilance: its vast opportunities to revolutionize drug safety monitoring and the challenges that must be overcome to ensure its ethical, effective, and sustainable adoption. By examining current applications, real-world case studies, regulatory perspectives, and future directions, this work aims to provide a comprehensive understanding of how AI can strengthen pharmacovigilance systems and ultimately enhance patient safety and public health outcomes worldwide.
Role of Artificial Intelligence in Pharmacovigilance :
Artificial Intelligence (AI) is transforming pharmacovigilance (PV) by enhancing the ability to detect, assess, and prevent adverse drug reactions (ADRs) more efficiently than traditional methods. Its role extends across multiple stages of drug safety monitoring, from data collection to signal detection and regulatory compliance.
Improved Drug Safety : AI enhances overall drug safety by enabling early detection of adverse drug reactions (ADRs) from vast, diverse data sources such as electronic health records (EHRs), social media, and spontaneous reporting systems. By analyzing real-world data in real time, AI helps reduce risks, supports proactive safety interventions, and ultimately safeguards public health.
Reliable Monitoring : Traditional PV systems often face issues like underreporting and delayed analysis. AI tools ensure continuous and reliable monitoring of drug safety signals by using machine learning algorithms to scan multiple databases simultaneously. This leads to faster identification of emerging safety concerns and strengthens post-marketing surveillance.
Automated Adverse Event Detection : AI, particularly through Natural Language Processing (NLP), can automatically extract and interpret potential ADR information from clinical notes, patient reports, literature, and even social media discussions. This automation reduces manual workload, increases efficiency, and minimizes human errors in adverse event reporting.
Signal Detection : AI-based models are highly effective in identifying hidden patterns, rare events, and complex associations between drugs and adverse events that may go unnoticed with conventional statistical methods. Machine learning and deep learning approaches improve the accuracy of signal detection, allowing for quicker and more reliable safety signal generation.
Data Triage : Pharmacovigilance databases contain enormous volumes of safety reports, many of which may be duplicates or of low clinical value. AI supports intelligent data triage by prioritizing high-risk cases, filtering duplicates, and classifying reports based on severity and clinical relevance. This enables PV professionals to focus resources on the most critical safety issues.
Data Integration : Drug safety data is fragmented across clinical trials, EHRs, mobile health apps, regulatory reports, and literature. AI facilitates seamless integration of multi-source, multi-format data into a unified platform for comprehensive analysis. Such integration supports holistic decision-making and enhances the benefit- risk assessment process.
Opportunities of Artificial Intelligence in Pharmacovigilance
Artificial Intelligence (AI) offers transformative opportunities in pharmacovigilance, addressing the limitations of traditional methods and enabling a more predictive, efficient, and patient-centric approach to drug safety monitoring. The key opportunities include:
Enhanced Efficiency and Automation : AI can automate repetitive PV tasks such as case processing, duplicate detection, literature screening, and adverse event (AE) coding. This reduces manual workload and operational costs, allowing pharmacovigilance professionals to focus on complex decision-making.
Real-Time Signal Detection : Traditional PV relies on retrospective analyses, often leading to delayed recognition of drug safety issues. AI enables real-time safety signal detection by continuously scanning electronic health records (EHRs), medical claims, and patient-generated data, ensuring faster intervention.
Utilization of Big Data ; Modern healthcare generates massive amounts of structured and unstructured data from EHRs, clinical trials, registries, wearables, and social media. AI can handle and analyze this “big data,” extracting meaningful insights that traditional statistical methods cannot manage effectively.
Improved Accuracy in Adverse Event Identification : Natural Language Processing (NLP) allows AI to analyze unstructured texts such as clinical notes, patient narratives, and online discussions. This improves the detection of underreported or rare adverse events, which may otherwise go unnoticed.
Personalized Medicine and Predictive Safety : AI supports the shift towards personalized pharmacovigilance by predicting ADR risks in specific patient subgroups based on genetic, demographic, and lifestyle data. This predictive capability enables proactive risk mitigation and safer prescribing practices.
Global Data Integration and Harmonization ; AI facilitates integration of safety data from multiple sources and geographies, helping regulatory authorities and pharmaceutical companies maintain globally harmonized pharmacovigilance systems. This strengthens international collaboration in drug safety monitoring.
Regulatory Support and Compliance : Regulatory agencies like the FDA, EMA, and WHO are exploring AI-based tools to enhance safety monitoring and regulatory decision-making. AI can improve compliance by ensuring timely detection, reporting, and evaluation of ADRs.
Cost-Effective Pharmacovigilance : By reducing manual reporting burdens, automating workflows, and improving efficiency, AI offers pharmaceutical companies a more cost-effective model for long-term PV operations while maintaining high safety standards.
Patient Engagement and Social Media Monitoring : AI can monitor social media and patient forums to capture real-world evidence about drug experiences. This provides new opportunities for early detection of safety concerns directly from patients, improving patient-centric pharmacovigilance.
Challenges of AI in Pharmacovigilance :
Artificial Intelligence (AI) is increasingly being adopted to enhance drug safety surveillance; however, its implementation in pharmacovigilance is not without significant challenges. These limitations need to be addressed before AI can be fully integrated into global drug safety systems.
Data Quality and Availability
Problem: AI systems rely heavily on the quality and completeness of data. Adverse Drug Reaction (ADR) reports are often incomplete, inconsistent, or contain non-standard terminologies. Social media and patient forums add further complexity due to unstructured, noisy, and subjective data.
Impact: Poor-quality data can result in inaccurate signal detection, false positives, or failure to identify real safety concerns.
Example: If ADR reports lack dosage or patient demographics, AI models cannot generate precise safety predictions.
Algorithmic Transparency and the “Black-Box” Problem
Problem: Many AI algorithms, particularly deep learning, operate as “black boxes” with limited interpretability. They provide predictions without explaining the reasoning process.
Impact: Lack of transparency hinders trust among regulators, clinicians, and pharmaceutical companies.
Example: An AI model may flag a drug as risky but fail to explain whether it was due to dosage, comorbidity, or patient age.
Regulatory and Legal Challenges
Problem: Regulatory agencies like the FDA, EMA, and WHO have not yet established uniform global standards for AI-driven pharmacovigilance tools. Questions about validation, accountability, and liability remain unresolved.
Impact: Delays in approval and limited adoption of AI solutions in drug safety monitoring.
Example: If an AI system fails to detect a life-threatening ADR, it is unclear whether responsibility lies with the software developer, regulator, or pharmaceutical company.
Ethical and Privacy Concerns
Problem: AI in PV often requires access to sensitive patient data (electronic health records, clinical trials, insurance claims). This raises concerns regarding privacy, data ownership, and informed consent.
Impact: Breach of confidentiality or misuse of health data could lead to ethical violations and legal penalties.
Example: Use of patient social media data without consent to track ADRs can violate privacy laws like GDPR (Europe) or HIPAA (USA).
Integration with Existing Pharmacovigilance Systems
Problem: Current PV systems rely on structured databases such as EudraVigilance (Europe), FAERS (U.S.), and VigiBase (WHO). Integrating AI tools with these legacy systems is technically complex due to interoperability issues.
Impact: Delayed implementation and duplication of work in signal management.
Example: An AI tool might not properly synchronize with FAERS data formats, leading to mismatched ADR records.
Resource, Cost, and Skill Gaps
Problem: Implementing AI requires high-performance computing, skilled professionals (data scientists, AI experts, PV specialists), and continuous system maintenance.
Impact: Many low- and middle-income countries (LMICs) may struggle with adoption due to financial and human resource limitations.
Example: A small pharmacovigilance unit in a developing country may not afford advanced AI platforms or expert training.
Bias and Generalizability Issues
Problem: AI models are only as good as the data they are trained on. If datasets are biased toward certain populations, drugs, or regions, the results may not generalize to global populations.
Impact: Potential health inequities and under-detection of ADRs in underrepresented groups.
Example: An AI trained mostly on Western patient data may fail to capture drug safety issues prevalent in Asian or African populations.
Overreliance and Human Oversight
Problem: Excessive reliance on AI might reduce human vigilance and critical thinking in ADR evaluation.
Impact: Important clinical context and expert judgment may be overlooked.
Example: AI may misclassify a serious ADR as “non-serious” without physician
Regulatory Perspectives on AI in Pharmacovigilance :
The integration of Artificial Intelligence (AI) into pharmacovigilance has attracted significant attention from global regulatory authorities, as it has the potential to transform drug safety surveillance while raising questions of compliance, reliability, and accountability. Since pharmacovigilance is a highly regulated domain that directly impacts patient safety, regulatory bodies such as the World Health Organization (WHO), the U.S. Food and Drug Administration (FDA), and the European Medicines Agency (EMA) are closely evaluating how AI technologies can be adopted without compromising ethical and legal standards.
1) Regulatory Concerns :
Data Quality and Standardization: Regulators stress that AI systems must rely on accurate, complete, and standardized data. Poor data quality or lack of harmonization between databases can lead to misleading safety signals.
Transparency and Explainability: AI models, especially deep learning algorithms, often function as “black boxes.” Regulators emphasize the need for explainable AI so that safety decisions can be justified and audited.
Bias and Fairness: Regulatory agencies are cautious about AI systems unintentionally introducing bias (e.g., underrepresenting minority groups in safety data), which could compromise patient safety.
Validation and Reliability: AI tools must undergo rigorous validation to demonstrate accuracy, reproducibility, and reliability before they can be used in regulatory submissions.
2) Key Guidelines and Frameworks
FDA (United States): The FDA has issued guidance on the use of AI/ML in medical devices and is exploring frameworks for AI in drug safety surveillance. The FDA’s Sentinel Initiative uses advanced analytics, including AI, to monitor real-world data for drug safety signals.
EMA (Europe):The EMA has acknowledged AI’s potential in pharmacovigilance and issued a draft reflection paper (2021) on AI/ML in the lifecycle of medicines, including post-marketing safety.Emphasis is placed on human oversight, risk management, and ensuring that AI complements rather than replaces human expertise.
WHO (Global):WHO encourages the integration of digital health technologies into pharmacovigilance but highlights the importance of global harmonization, especially for low- and middle-income countries. WHO’s Uppsala Monitoring Centre (UMC) is already experimenting with machine learning to enhance signal detection in the global ADR database (VigiBase).
3) Ethical and Legal Considerations
Patient Privacy and Data Protection: Regulatory frameworks such as GDPR (General Data Protection Regulation, EU) and HIPAA (Health Insurance Portability and Accountability Act, USA) mandate strict safeguards for patient-identifiable data used in AI-driven PV systems.
Accountability: Clear responsibility must be defined in cases where AI-driven systems fail to detect or incorrectly report adverse drug reactions. Regulators emphasize that final accountability lies with human experts, not AI tools.
4) Future Directions
Development of global regulatory harmonization to ensure consistency in AI adoption across countries.
Promotion of explainable AI (XAI) frameworks to enhance trust between regulators, industry, and the public.
Use of regulatory sandboxes (controlled experimental settings) to test AI systems in pharmacovigilance before large-scale deployment.
Collaboration between pharmaceutical companies, technology developers, and regulators to create standardized validation protocols for AI applications in PV.
CONCLUSION :
Artificial Intelligence (AI) is emerging as a transformative force in pharmacovigilance (PV), offering solutions to challenges such as underreporting, delayed signal detection, and the management of large, complex datasets. By applying technologies like machine learning, natural language processing, and deep learning, AI enables faster detection of adverse drug reactions (ADRs), real-time monitoring, and integration of diverse data sources, ultimately improving drug safety and patient outcomes. These opportunities make PV systems more proactive, efficient, and patient-centered. However, AI adoption in PV also faces barriers, including data quality issues, algorithmic transparency, ethical concerns, and regulatory uncertainties. While AI can significantly enhance drug safety, it cannot fully replace human expertise; rather, it should complement professionals in decision-making. Regulatory authorities such as the FDA, EMA, and WHO emphasize the need for transparency, accountability, and harmonized frameworks to guide safe implementation. In summary, AI has the potential to revolutionize pharmacovigilance by shifting it from reactive to predictive drug safety monitoring. With responsible use, clear regulations, and ethical safeguards, AI can strengthen global PV systems, improve trust in healthcare, and ultimately protect public health.
REFERENCES
Desai, M. K. (2024). Artificial intelligence in pharmacovigilance: Opportunities and challenges. Perspectives in Clinical Research, 15(3), 116–121.
Schmider, J., Kumar, K., Nair, A., Goldberg, S., & Mayo, F. (2018). Innovation in pharmacovigilance: Use of artificial intelligence in adverse event case processing. Clinical Pharmacology & Therapeutics, 104(1), 26–29.
Saravanan, P., & Iyer, G. (2022). Industry perspective on artificial intelligence/machine learning in pharmacovigilance. Perspectives in Clinical Research, 13(3), 178–183.
Duke, J., Banerjee, A., & Reddy, S. (2021). Artificial intelligence for pharmacovigilance signal management. Drug Safety, 44(9), 943–952.
Ohana, B., Sullivan, J., & Baker, N. (2021). Validation and transparency in AI systems for pharmacovigilance: A case study on literature monitoring. Drug Safety, 44(12), 1301–1312.
Ward, I. R., Raza, S., & Smith, T. (2021). Explainable artificial intelligence for pharmacovigilance: What features are important when predicting adverse outcomes? Journal of Biomedical Informatics, 122, 103896.
Biseda, B., & Mo, K. (2020). Enhancing pharmacovigilance with drug reviews and social media using natural language processing. Frontiers in Pharmacology, 11, 870.
Hakim, J. B., Wang, Z., & Atreja, A. (2024). Guardrails for large language models in pharmacovigilance: Ensuring safety in AI applications. NPJ Digital Medicine, 7(1), 145.
Baselie, Y., Joshi, V., & Krishnan, S. (2019). Artificial intelligence within pharmacovigilance: Cognitive services framework. Indian Journal of Pharmacology, 51(6), 365–370.
Chino, F., Desai, R., & Subramanian, S. (2024). Artificial intelligence in pharmacovigilance: Expanding to materiovigilance and virtual trials. Perspectives in Clinical Research, 15(4), 182–189.
Bolislis, W. R., Fay, M., & Kühler, T. C. (2020). Regulatory perspective on artificial intelligence in drug development and pharmacovigilance. Clinical Pharmacology & Therapeutics, 107(4), 819–821.
Harpaz, R., Callahan, A., Tamang, S., Low, Y., Odgers, D., Finlayson, S., & Shah, N. H. (2017). Text mining for adverse drug events: The promise, challenges, and state of the art. Drug Safety, 40(12), 1171–1197.
Van Stekelenborg, J., et al. (2019). Signal detection methodologies to support pharmacovigilance: Artificial intelligence and data mining. Pharmacoepidemiology and Drug Safety, 28(9), 1235–1242.
Singh, R., & Lal, R. (2022). Application of machine learning and artificial intelligence in pharmacovigilance: An Indian perspective. Journal of Pharmacovigilance & Drug Safety, 19(2), 75–82.
Trifirò, G., et al. (2019). Big data and pharmacovigilance: Data, algorithms, and regulatory needs. Drug Safety, 42(12), 1201–1212.
Banerjee, A. K., et al. (2019). Artificial intelligence in pharmacovigilance: Scoping review and regulatory considerations. Drug Safety, 42(5), 569–580.
Liu, X., & Chen, H. (2021). Deep learning for adverse drug reaction detection from electronic health records. Journal of the American Medical Informatics Association, 28(1), 15–23.
Wisniewski, A. F., et al. (2016). Application of data mining in pharmacovigilance. Fundamental & Clinical Pharmacology, 30(5), 620–630.
Smith, A. R., & Roberts, J. (2020). Natural language processing in pharmacovigilance: Opportunities for signal detection. Therapeutic Advances in Drug Safety, 11, 2042098620947815.
Ibrahim, S., & Babar, Z. U. D. (2023). Artificial intelligence and pharmacovigilance in low- and middle-income countries: Opportunities and challenges. Journal of Pharmaceutical Policy and Practice, 16(1), 112.
Reference
Desai, M. K. (2024). Artificial intelligence in pharmacovigilance: Opportunities and challenges. Perspectives in Clinical Research, 15(3), 116–121.
Schmider, J., Kumar, K., Nair, A., Goldberg, S., & Mayo, F. (2018). Innovation in pharmacovigilance: Use of artificial intelligence in adverse event case processing. Clinical Pharmacology & Therapeutics, 104(1), 26–29.
Saravanan, P., & Iyer, G. (2022). Industry perspective on artificial intelligence/machine learning in pharmacovigilance. Perspectives in Clinical Research, 13(3), 178–183.
Duke, J., Banerjee, A., & Reddy, S. (2021). Artificial intelligence for pharmacovigilance signal management. Drug Safety, 44(9), 943–952.
Ohana, B., Sullivan, J., & Baker, N. (2021). Validation and transparency in AI systems for pharmacovigilance: A case study on literature monitoring. Drug Safety, 44(12), 1301–1312.
Ward, I. R., Raza, S., & Smith, T. (2021). Explainable artificial intelligence for pharmacovigilance: What features are important when predicting adverse outcomes? Journal of Biomedical Informatics, 122, 103896.
Biseda, B., & Mo, K. (2020). Enhancing pharmacovigilance with drug reviews and social media using natural language processing. Frontiers in Pharmacology, 11, 870.
Hakim, J. B., Wang, Z., & Atreja, A. (2024). Guardrails for large language models in pharmacovigilance: Ensuring safety in AI applications. NPJ Digital Medicine, 7(1), 145.
Baselie, Y., Joshi, V., & Krishnan, S. (2019). Artificial intelligence within pharmacovigilance: Cognitive services framework. Indian Journal of Pharmacology, 51(6), 365–370.
Chino, F., Desai, R., & Subramanian, S. (2024). Artificial intelligence in pharmacovigilance: Expanding to materiovigilance and virtual trials. Perspectives in Clinical Research, 15(4), 182–189.
Bolislis, W. R., Fay, M., & Kühler, T. C. (2020). Regulatory perspective on artificial intelligence in drug development and pharmacovigilance. Clinical Pharmacology & Therapeutics, 107(4), 819–821.
Harpaz, R., Callahan, A., Tamang, S., Low, Y., Odgers, D., Finlayson, S., & Shah, N. H. (2017). Text mining for adverse drug events: The promise, challenges, and state of the art. Drug Safety, 40(12), 1171–1197.
Van Stekelenborg, J., et al. (2019). Signal detection methodologies to support pharmacovigilance: Artificial intelligence and data mining. Pharmacoepidemiology and Drug Safety, 28(9), 1235–1242.
Singh, R., & Lal, R. (2022). Application of machine learning and artificial intelligence in pharmacovigilance: An Indian perspective. Journal of Pharmacovigilance & Drug Safety, 19(2), 75–82.
Trifirò, G., et al. (2019). Big data and pharmacovigilance: Data, algorithms, and regulatory needs. Drug Safety, 42(12), 1201–1212.
Banerjee, A. K., et al. (2019). Artificial intelligence in pharmacovigilance: Scoping review and regulatory considerations. Drug Safety, 42(5), 569–580.
Liu, X., & Chen, H. (2021). Deep learning for adverse drug reaction detection from electronic health records. Journal of the American Medical Informatics Association, 28(1), 15–23.
Wisniewski, A. F., et al. (2016). Application of data mining in pharmacovigilance. Fundamental & Clinical Pharmacology, 30(5), 620–630.
Smith, A. R., & Roberts, J. (2020). Natural language processing in pharmacovigilance: Opportunities for signal detection. Therapeutic Advances in Drug Safety, 11, 2042098620947815.
Ibrahim, S., & Babar, Z. U. D. (2023). Artificial intelligence and pharmacovigilance in low- and middle-income countries: Opportunities and challenges. Journal of Pharmaceutical Policy and Practice, 16(1), 112.
10.5281/zenodo.16904100
