Pharmacovigilance is the process of keeping an eye on the safety of all medications, including biological agents, vaccines, and herbal and complementary therapies 1. The research and practices around the identification, evaluation, comprehension, and avoidance of side effects or any other medication-related issue are known as pharmacovigilance. It is a crucial part of providing patients with care and using medications sensibly 2, 3. It is also known by a number of different names, including post-marketing surveillance, spontaneous reporting, adverse drug reaction monitoring, drug safety surveillance, and side effect monitoring 1.

Traditional approaches to adverse event reporting and identification are essential for keeping an eye on patient safety and pharmaceutical safety in healthcare settings 1, 7. The first approach, called voluntary reporting, is based on healthcare providers, patients, and caregivers informing regulatory bodies like the FDA or MHRA about suspected medication-related adverse events 7. This method offers real-world data on adverse events that occur during ordinary clinical practice and is simple and easy to understand 2, 7. However, voluntary reporting is vulnerable to bias and underreporting, whereby specific negative events may be overreported or underreported depending on personal experiences and perceptions, resulting in differences in the accuracy and comprehensiveness of the reported data. A methodical examination of patient records, including doctor notes and test findings, provides a structured method for discovering adverse occurrences. This process is known as a medical record or chart review 1, 2, 3. Even though this approach makes it possible to gather comprehensive clinical data, it can be labour- and time-intensive and may miss adverse events that are not sufficiently recorded in the records 2, 3. When providing clinical treatment, healthcare workers who engage in direct observation actively watch patients for indications of unfavourable events. With the help of this technique, bad events can be detected in real time, allowing for quick action to lessen patient harm 3. However, especially in busy hospital settings, it could require a lot of resources and miss unfavourable events that happen outside of the monitoring time 8. Patient and family reports offer important insights into patient perceptions and experiences with medication safety. Adverse responses can be reported by patients or caregivers to healthcare providers via satisfaction questionnaires or during clinic visits. Although these reports present distinctive viewpoints, they could be arbitrary and inconsistent with clinical evaluations, which could result in differences in the veracity of the reports. These conventional techniques, in spite of their shortcomings, support post-marketing surveillance initiatives by helping to detect and lessen adverse events related to drug usage 2, 3. To improve patient safety and streamline healthcare procedures, it is essential to comprehend the advantages and disadvantages of each approach when analysing and applying the data gathered 8.

Artificial intelligence (AI) is becoming more and more adept at performing tasks that humans do, but faster, cheaper, and more effectively. AI has enormous promise in the medical field. AI is progressively becoming a part of our healthcare ecosystem, just like it is in our daily lives 9. The combination of AI's many subdomains has sparked notable advancements and creative concepts in this dynamic context, which are transforming healthcare delivery systems worldwide. A key piece of technology in the healthcare industry, machine learning (ML) uses statistical models and algorithms to evaluate vast amounts of medical data and draw insightful conclusions 10. By predicting patient outcomes, spotting illness trends, and streamlining treatment plans using individualized patient data, machine learning (ML) algorithms enable healthcare systems to improve clinical decision-making. A powerful branch of machine learning called deep learning has shown impressive results in deciphering complicated medical imaging data, including CT, MRI, and X-ray images 9, 11. Radiologists and other healthcare professionals can better diagnose and plan treatments for diseases including cancer, cardiovascular disease, and neurological disorders by using deep neural networks, which provide accurate and efficient detection of abnormalities and diseases. The way medical professionals engage with enormous volumes of unstructured textual data, such as clinical notes, electronic health records (EHRs), and medical literature, has been completely transformed by natural language processing (NLP) technologies. In order to improve documentation accuracy, clinical coding, and information retrieval procedures, natural language processing (NLP) algorithms extract useful clinical information, identify trends in patient narratives, and aid in the semantic interpretation of medical papers. In the field of computer vision, artificial intelligence (AI)-powered systems improve the interpretation of diagnostic tests and medical images, providing highly accurate and efficient automated analysis and anomaly identification 12. Healthcare professionals can identify lesions, cancers, and other anomalies with the help of computer vision algorithms, which can result in early interventions and better patient outcomes. AI-enabled robotic technologies are revolutionizing medical interventions and surgery by facilitating less invasive procedures, remote consultations, and precision surgeries. Artificial intelligence (AI)-powered surgical robots improve patient safety and surgical results by increasing surgical accuracy, lowering operating risks, and speeding up recuperation periods. By utilizing domain-specific information and clinical guidelines, expert systems designed specifically for the healthcare industry give doctors clinical decision-making aids and decision support tools that maximize treatment plans and diagnostic workflows. These technologies increase patient safety in all healthcare settings, optimize clinical workflows, and improve care coordination. Large archives of patient data, clinical guidelines, and medical knowledge are organized using Knowledge Representation and Reasoning frameworks, which enable data-driven insights, predictive analytics, and evidence-based decision-making in healthcare delivery and research 8, 11. Reinforcement learning maximize patient outcomes and the provision of healthcare services by streamlining workflows, improving resource allocation, and boosting operational effectiveness. A new era of precision medicine, data-driven healthcare delivery, and personalized medicine is being heralded by the confluence of AI subfields within the healthcare domain. Healthcare companies may increase patient care quality, open up new avenues for innovation, and tackle challenging healthcare issues of the twenty-first century by utilizing AI technology 11

Artificial intelligence (AI) has swiftly transformed various industries, including pharmacovigilance, enabling more accurate and efficient monitoring and evaluation of adverse drug reactions. Predictive modelling exemplifies AI's proactive approach to identifying potential safety concerns related to pharmaceuticals, enhancing patient safety by detecting adverse events early 13. AI facilitates the analysis of large datasets from diverse sources like social media and EHRs, fully identifying potential signals of adverse drug reactions and enabling prompt responses when necessary 12, 14. Despite its revolutionary potential, AI integration in pharmacovigilance faces obstacles such as data privacy issues and regulatory framework adaptation. However, the potential benefits of improving patient outcomes and medication safety underscore the importance of continued research and development 15. The convergence of AI with pharmacovigilance signifies a transformative moment in drug safety surveillance, providing early detection of safety concerns and expediting investigation of adverse drug reactions. Methods for incorporating AI into pharmacovigilance include NLP for extracting insights from unstructured data sources and predictive analytics for projecting adverse occurrences based on data trends 14, 16. Real-world uses of AI in pharmacovigilance automate the identification and categorization of adverse drug responses from diverse data sources and scan social media for possible safety alerts, minimizing time and resources needed while improving accuracy and consistency. AI-enhanced pharmacovigilance presents both potential and challenges, such as resolving data privacy issues and modifying regulatory frameworks. Future developments in AI-powered medication safety monitoring systems promise proactive and patient-centred methods through ongoing innovation and improvement 17. Case studies demonstrate the effectiveness of AI algorithms in identifying early safety signals and enhancing overall patient safety outcomes 13. AI integration into pharmacovigilance signals a new phase of proactive, data-driven methods for tracking medication safety, with enormous potential to enhance patient outcomes and healthcare administration 15.

Automation powered by artificial intelligence (AI) has become a ground-breaking advancement in pharmacovigilance, radically altering the ways in which adverse events and safety signals are recognized, assessed, and handled in the pharmaceutical and healthcare industries 15. Clinical knowledge, manual review, and retrospective analysis of data from individual case reports, epidemiological research, and clinical trials were major components of traditional approaches 14. But these methods' scalability, efficiency, and vulnerability to biases and human error were severely limited. A paradigm shift in pharmacovigilance is anticipated with the introduction of AI-driven automation, which uses advanced algorithms, machine learning models, and natural language processing (NLP) approaches to quickly and efficiently assess large amounts of real-world data sources 12, 15, 16. Artificial intelligence (AI) algorithms are able to identify trends, correlations, and anomalies that may indicate bad responses or new safety issues by examining electronic health records, adverse event reports, medical literature, and social media posts 14. Even though AI-driven automation has the potential to be revolutionary, there are a number of obstacles and restrictions. Adopting AI technologies necessitates large infrastructure, computational resources, and regulatory compliance investments 17. Moreover, ongoing algorithmic validation, monitoring, and enhancement activities are required to guarantee the accuracy, dependability, and generalizability of AI-driven systems 14, 18. The usefulness and efficiency of AI in augmenting medication safety monitoring and regulatory decision-making procedures are demonstrated by real-world instances 11.

Examples of how AI technology might transform pharmacovigilance procedures include the FDA's Sentinel Initiative, Advera Health Analytics' SignalMine, Oracle Health Sciences' Argus Safety, AstraZeneca's AI-driven pharmacovigilance system, and IBM Watson for Drug Safety.

The use of artificial intelligence (AI) in pharmacovigilance presents a viable route to better patient outcomes and medication safety monitoring 24. Issues with data availability and quality are the most significant since they directly affect the accuracy and dependability of AI-powered forecasts and analyses 25. It is challenging to perform the in-depth analysis required for effective pharmacovigilance because of the incompleteness of data reporting, inaccuracies, biases, and data silos. Problems with data quality and availability can delay regulatory action and harm patient outcomes 25, 26. Examples of real-world experiences include the Essure and Vioxx (rofecoxib) incidents. Moreover, interpretability and transparency pose significant challenges; complex AI models sometimes act as "black boxes," making it challenging for stakeholders to understand the choices that are made 24. Algorithmic bias, a lack of standardized assessment tools, and regulatory compliance all of which necessitate careful consideration and methodical approaches make these problems more challenging. To fully leverage artificial intelligence's potential in pharmacovigilance, problems with data fragmentation, standardization, and technological compatibility must be resolved 24, 26. Facilitating collaboration among regulators, technology developers, and stakeholders is imperative in order to promote transparency, enhance data quality, and adhere to legal obligations. By getting past these challenges and limitations, AI-driven pharmacovigilance can enhance patient safety and healthcare outcomes internationally 24.

Artificial intelligence (AI) has the potential to have a significant influence on drug safety surveillance by changing the way adverse events are identified, examined, and handled across the course of pharmaceutical product development 27. Advancements in recent times highlight the various uses of AI technology to improve the efficacy, precision, and clarity of pharmacovigilance procedures 24. Massive volumes of structured and unstructured data from a variety of sources, including wearable technology, social media, medical literature, and electronic health records (EHRs), may be analysed by AI algorithms 23. For example, IBM Watson for Drug Safety uses artificial intelligence (AI) to evaluate real-world data and identify possible indicators of adverse drug reactions (ADRs) more effectively than using conventional technique 12. Trust among stakeholders is increased with explainable AI frameworks, such as the Interpretable Deep Learning for Drug-Induced Liver Injury (IDILI), which offer transparent insights into AI-generated predictions. The IDILI framework was created by Stanford University researchers to forecast drug-induced liver damage events. The model's predictions are explained by the researchers using observable properties that are derived from biological pathways and molecular structures 28. The FDA's Sentinel Initiative serves as an example of how real-world data (RWD) integration complements traditional clinical trial data by providing insights into patient outcomes and medication safety profiles in real-world situations. The FDA's Sentinel Initiative analyses RWD from electronic healthcare databases using AI and machine learning algorithms to track the safety of pharmaceuticals and spot possible side effects instantly. When used on adverse event reports from databases such as the FDA Adverse Event Reporting System (FAERS), temporal pattern mining techniques allow the discovery of new safety issues as well as long-term patterns. Using temporal pattern mining tools, research published in Pharmacological Safety examined adverse event data from FAERS to find temporal clusters of adverse events linked to certain medicines or drug classes 12, 28, 29. Distributed AI frameworks allow for safe cooperation and knowledge exchange while maintaining data privacy. Examples of these frameworks are federated learning platforms like the PharmAI consortium. The PharmAI collaboration created a federated learning platform that enables regulatory bodies, pharmaceutical firms, and healthcare providers to work together to train AI models on dispersed datasets while maintaining the privacy of sensitive patient data. Personalized risk assessment and treatment techniques are made possible by machine learning algorithms that identify predictive biomarkers for medication safety 12, 17, 19. In order to uncover genetic variations linked to drug-induced liver damage (DILI) and create predictive biomarker panels for determining the risk of DILI in patients receiving particular pharmacological treatments, a research study employed machine learning algorithms. Automated case processing and signal identification are made easier by natural language understanding (NLU) approaches, which are best shown by NLP systems for adverse event report analysis. In order to expedite the discovery of possible safety signals, a research team created a natural language processing (NLP) system that can analyse and summarize adverse event reports sent to regulatory bodies 16, 23. The assessment of the causative linkages between medications and adverse events is improved by the use of causal inference models, such as the framework created by Harvard Medical School researchers. More accurate evaluations of drug safety profiles are made possible by the framework, which estimates the causal impact of medications on particular adverse events using observational data and causal graph models 29. The construction of AI models for adverse event detection is accelerated, and data labelling is optimized through the use of semi-supervised and active learning approaches. AI-based signal detection systems perform better when active learning techniques prioritize the classification of adverse event reports based on their potential relevance and informativeness 14, 29.

The use of AI in pharmacovigilance presents exciting opportunities for the future that will transform the management and monitoring of drug safety 30. AI technology can improve adverse event detection's effectiveness and precision, enabling real-time monitoring and analysis of large datasets from various sources. Proactive actions to reduce harm are made possible by these systems' faster and more accurate ability to recognize patterns, trends, and possible threats than by more conventional techniques 9, 15, 21. Furthermore, regulatory agencies and medical experts may take prompt and decisive action in response to growing safety issues by using AI-driven predictive analytics 14. AI integration also has the potential to enhance pharmacovigilance reporting requirements and expedite regulatory compliance procedures 24. All things considered, the use of AI in pharmacovigilance presents a revolutionary way to improve drug safety monitoring and protect public health. The following suggestions for further study and advancement have been made in order to effectively use AI in pharmacovigilance. Pharmacovigilance stakeholders may use the revolutionary potential of artificial intelligence (AI) technology to improve patient care, regulatory decision-making, and drug safety monitoring by adopting these guidelines. Global public health will benefit from these initiatives as safer and more effective pharmaceuticals become a reality.