jrnls.org logo
Submit

BROWSE ALL JOURNALS

SEE ALL 19 JOURNALS
International Journal of Pharmaceutical InvestigationNOW IN
Review

Digital Surveillance and Artificial Intelligence in Detection of AEFI: A New Frontier in Vaccine Safety Monitoring

Soiba Siddiqui1, Madhan Ramesh1★, Prashanth Sathya Narayana2★ Corresponding author

  1. 1Department of Pharmacy Practice, JSS College of Pharmacy, JSS AHER, Mysuru, Karnataka,INDIA.
  2. 2Department of Paediatrics, JSS Medical College, JSS AHER, Mysuru, Karnataka, INDIA.

CORRESPONDENCE

Madhan Ramesh

Department of Pharmacy Practice, JSS College of Pharmacy, JSS AHER, Mysuru-570015, Karnataka, INDIA.

mramesh@jssuni.edu.in

Received: 26-05-2025; Revised: 09-07-2025; Accepted: 12-09-2025.

Volume 16, Issue 1 · pp. 51–57 · PUBLISHED Jan-Mar 2026 · DOI: 10.5530/ijpi.20260414

View on IJPI original site ↗

ABSTRACT

The integration of digital surveillance and Artificial Intelligence (AI) technology constitutes a significant shift in Adverse Event Following Immunization (AEFI) monitoring. This review explores the artificial intelligence based approaches such as machine learning, Natural Language Processing (NLP), and blockchain technology to improve vaccination safety surveillance. Artificial intelligence provides the immediate and accurate identification of safety signals by using a variety of data sources, such as electronic health records, social media platforms, and mobile health applications. Social media analytics have shown great promise for early detection of AEFIs, but blockchain protects the integrity and transparency of safety data. Despite these advances, difficulties remain, including underreporting, inequities in healthcare infrastructure, and ethical considerations including data privacy and biased algorithms. Low-resource environments, in specifically, experience challenges due to insufficient digital infrastructure and employee abilities. The assessment emphasizes the importance of combining old passive reporting systems with modern digital tools to develop a comprehensive and balanced AEFI monitoring framework. By addressing these barriers enabling global collaboration, digital technologies can increase the effectiveness of AEFI surveillance, increase confidence in vaccination programs, and improve emergency response times. This innovation in vaccine safety monitoring has the ability to address existing gaps and provide a strong framework for global immunization safety.

KEYWORDS

0% READ

FULL TEXT

INTRODUCTION

Maintaining public confidence in vaccination programs and ensuring the detection of unusual Adverse Event Following Immunization (AEFI) that may not be observed in clinical trials requires monitoring vaccine safety (Buttery and Clothier, 2022). Artificial Intelligence (AI) and digital surveillance are becoming more effective techniques for improving vaccine safety monitoring, especially when it comes to the identification of adverse events following vaccination. With the use of digital surveillance and artificial intelligence technologies, safety warnings might be identified more proactively and surveillance could occur almost instantly. Active surveillance systems that combine administrative, claims data and electronic health records are quickly emerging as a major global development trend in post-marketing vaccine safety monitoring (Sun et al., 2021).

Compared to more conventional passive monitoring techniques, this strategy enables the earlier identification of any safety concerns. Utilizing natural language processing techniques for social media monitoring has also demonstrated potential for early identification of mentions of Vaccine Adverse Events (VAEMs) (Khademi Habibabadi et al., 2022). AEFI surveillance includes mobile health technology as an additional frontier. Compared to typical passive reporting, the use of SMS prompts for consumer-based vaccination safety monitoring greatly increased AEFI detection rates (Gold et al., 2021).

It's significant to note that while digital approaches have benefits, conventional passive reporting systems continue to have issues. Only 56.2% of nations reported ≥10 AEFI per 100,000 surviving newborns in 2019, according to a worldwide review, pointing to limitations in regular AEFI reporting (Salman, 2021). Artificial intelligence-enabled natural language processing and picture analysis can quickly identify possible adverse events by extracting valuable insights from unstructured data, including social media posts and electronic medical records (Zeng et al., 2021). Digital surveillance and artificial intelligence will probably become more crucial as vaccination systems develop, especially in response to global health emergencies like COVID-19, in order to ensure vaccine safety and maintain confidence in immunization programs (Teo, 2021). By improving vaccine safety surveillance systems and enhancing their capacity to proactively identify and promptly resolve safety problems, the use of these innovative techniques may eventually improve public confidence in immunization programs (Ball, 2014).

The Evolution of AEFI Monitoring: From Paper to Pixels

Adverse Events Following Immunization monitoring has advanced significantly, switching from paper-based to computerized methods. Conventional AEFI monitoring was dependent on pre-existing reporting systems, which frequently resulted in delays between the reporting of an event and its occurrence (Khademi Habibabadi et al., 2022). Underreporting and a lack of timeliness were two issues with these passive monitoring techniques (Laryea et al., 2019). Modern methods, on the other hand, improve AEFI monitoring by utilizing digital technology (Psihogios et al., 2022). Using smartphone technology, for example, the SmartVax program actively requests AEFI reports from vaccination recipients through automated SMS questionnaires (Westphal et al., 2016). Even with older populations, this approach showed great acceptability and response rates, indicating that it is a potential tool for almost real-time vaccination safety monitoring. It's interesting to observe that social media has become a valuable additional resource for tracking side effects of vaccinations. High accuracy has been achieved in extracting mentions of vaccine adverse events from Twitter tweets using the VAEM-Mine approach, which blends topic modeling and classification techniques (Khademi Habibabadi et al., 2022).

The speed, completeness, and sensitivity of vaccination vigilance have all greatly increased with the switch from paper-based to digital AEFI monitoring systems. SafeVac and other mobile apps have demonstrated the viability of collecting AEFI data over extended periods (Nguyen et al., 2021). Tracking vaccination safety has shown to be successful using traditional AEFI monitoring methods, such as China's National Adverse Event Following Immunization Surveillance System (NAEFISS) (Hu et al., 2023). It is become more and more obvious that advances are moving into the digital realm. According to an evaluation of Ethiopia's preparedness for deploying active AEFI surveillance, technologies that support both paper-based and electronic-based recording are required (Zeleke et al., 2023). The utilization of international databases such as VigiBase for AEFI signal disproportionality analysis serves as another example of this transformation (Kim et al., 2024). The shift to digital has been accelerated by the COVID-19 pandemic. The vaccine safety service in Victoria, SAEFVIC, quickly modified its systems to handle higher report volumes by combining data from both active and spontaneous surveillance. They generated public-facing vaccination safety reports, created databases specifically for AESI for improved monitoring, and deployed technologies to automate, filter, and triage reports (Laemmle Ruff et al., 2022).

Harnessing Big Data: Social Media and AEFI Signal Detection

Social media platforms have become important sources of information for tracking vaccination safety signals and adverse occurrences after immunizations. Research has indicated the effectiveness of using social media data, specifically Twitter, in order to obtain a basic understanding of vaccination safety concerns (Habibabadi et al., 2023; Khademi Habibabadi et al., 2022). Using natural language processing techniques, the VAEM-Mine approach developed in one research, extracted mentions of vaccination adverse events from Twitter tweets with good accuracy (F1 score of 0.91) (Khademi Habibabadi et al., 2022). Additional studies showed that 3 million tweets on COVID-19 vaccinations were gathered and examined. Topic modeling and machine learning were used to detect individual vaccine reactions (Habibabadi et al., 2023). Through the provision of more timely data and the opportunity to capture instances that may be unreported through official channels, these social media-based initiatives can serve as a valuable supplement to standard AEFI reporting systems (Habibabadi et al., 2023; Khademi Habibabadi et al., 2022).

Additionally, social media and internet search engine results might be helpful tools for finding reports of AEFIs connected to cluster anxiety that might not be included in Conventional scientific research (Suragh et al., 2018). After developing which extracted Vaccine Adverse Event Mentions (VAEMs) from Twitter with high accuracy, 8,992 VAEMs were found among the more than 800,000 tweets pertaining to vaccines (Khademi Habibabadi et al., 2022). In the same way, different research found people who had directly experienced COVID-19 vaccination responses on Twitter and confirmed its findings using an established method for reporting vaccine reactions (Habibabadi et al., 2023). These methods get over the drawbacks of passive systems, such as their underreporting and lack of timeliness. Utilizing the potential of big data from social media and other digital sources to improve vaccination pharmacovigilance and AEFI signal detection is significant (Khademi Habibabadi et al., 2022).

Machine Learning Algorithms in AEFI Pattern Recognition

In multiple aspects, AEFI detection and monitoring has been enhanced via machine learning and advanced data analysis techniques including (i) Near real-time surveillance: to detect any potential risks more efficiently, experts are investigating into techniques to gather data closer to real-time (Sun et al., 2021). (ii) Utilization of electronic healthcare data: there is an opportunity to improve AEFI signal monitoring with the increasing quantity of electronic healthcare data (Mesfin et al., 2020). (iii) Social media monitoring: Vaccine-related adverse events are gathered from social networking platforms such as Twitter using natural language processing techniques (Khademi Habibabadi et al., 2022). (iv) SMS-based surveillance: SMS-based surveillance significantly improved AEFI detection rates compared to conventional passive monitoring techniques (Gold et al., 2021).

Machine Learning (ML), which offers various advantages over conventional techniques, is becoming more widely acknowledged as a useful tool for tracking vaccination safety. Large amounts of data from a variety of sources, such as national claim databases and electronic health records, may be processed effectively by ML algorithms, which can then identify possible adverse effects following vaccination with high accuracy (Kim et al., 2021). For active surveillance systems, which proactively detect information on AEFI and promptly look into any safety warnings, this feature is extremely significant (Sun et al., 2020). Random forests, logistic regression, and support vector machines are a few supervised machine-learning techniques that have demonstrated potential for AEFI identification (Carrell et al., 2023). These techniques are capable of identifying possible safety signs by analyzing intricate patient data. While less common, artificial neural networks and naive Bayes classifiers have also been used (Koutanaei et al., 2015).

Improvements were observed in the prediction and detection of the severity of adverse effects following immunization using ML models. XG Boost models had been used in identifying and predicting mild to severe side effects (Levi et al., 2024). Clinical report AEs have been successfully identified and cataloged using Large Language Models (LLMs) (Li et al., 2024). Remarkably, other machine learning techniques have also been used in research related to vaccines, even though LLMs demonstrate promise in AE identification.

Electronic medical data have been analyzed using natural language processing algorithms in order to identify potential side effects of vaccinations. Using confidential coding, the MediClass system was able to identify both gastrointestinal-specific and generic AEFIs in clinical notes, surpassing conventional approaches that just use diagnostic codes (Hazlehurst et al., 2009).

Machine learning algorithms have shown significant results in a range of medical applications, such as the identification and categorization of unfavorable events that occur after vaccination schedules (Ghaffar Nia et al., 2023). Their ability to analyze large datasets, identify complex patterns, and make accurate predictions has made them invaluable tools in the field of public health surveillance (Zeng et al., 2021). Predictive analytics is one of the main uses of machine learning in adverse event after vaccination pattern detection (An et al., 2023). Healthcare practitioners may be able to detect high-risk people or populations for adverse events after vaccination by utilizing machine learning algorithms (Sidey-Gibbons et al., 2019).

Natural Language Processing: Decoding Patient-Reported AEFIs

An increasingly useful technique for examining and comprehending patient-reported adverse events after immunizations is natural language processing, or NLP. Natural language processing techniques might be utilized for gathering appropriate information from social media posts, patient records, and medical reports in order to identify relevant AEFI signs (Dong et al., 2024). These methods evaluate on a both syntactic and semantic level, addressing problems such as identified categorization, sentiment analysis, and determined classification (Pattanayak, 2023). Handling language appropriate to a certain area and individualized comprehension is one of the challenges in using natural language processing to the identification of AEFI signals (Gour, 2020).

Numerical language processing approaches have the potential to extract meaningful information from huge quantities of text data, such as reports on vaccination safety and patient suggestions (Dong et al., 2024). NLP-based methods have demonstrated potential in vaccination safety surveillance by recognizing and eliminating disproportionality signals associated with listed AEFIs, thereby enhancing signal management. 17% of disproportionality signals for COVID-19 vaccinations might be rejected using an NLP-based method, according to research that used the Vaccine Adverse Event Reporting System (VAERS). This might result in a better use of time and resources in signal management (Dong et al., 2024).

Researchers and medical practitioners may examine huge patient-reported adverse event file index databases more quickly and effectively by using natural language processing. This might result in better vaccination safety monitoring and more focused actions to allay public fears (Dong et al., 2024). The analysis of patient-reported data has become an essential component of modern healthcare due to the increased emphasis on patient-centered treatment and the increasing implementation of electronic health records (Li et al., 2022). Particularly, natural language processing methods have become an effective means of drawing conclusions from the unstructured textual data present in adverse event reports submitted by patients (Allabun and Soufiene, 2023). Although these details can be difficult to collect using typical structured data fields alone, natural language processing's ability to effectively extract and assess the occurrence and severity of adverse events is a considerable benefit (Young et al., 2019). In the context of patient safety monitoring and quality improvement activities, the utilization of computational approaches to analyze free-text notes may provide an enhanced understanding of the patient experience (Ozonoff et al., 2022).

Blockchain Technology: Ensuring Data Integrity in AEFI Reporting

Adverse Events Following Immunization reporting can greatly benefit from the Distributed Ledger Technology (DLT) of blockchain (Xu et al., 2022). AEFI data integrity and traceability may be guaranteed by blockchain's distributed and immutable nature (Liu et al, 2020; Roberts et al., 2023). The distributed ledger technology's Data Block Matrix (DBM) version permits controlled data revision or deletion while maintaining integrity (Roberts et al., 2023). DLT on blockchain improves data security, privacy, and integrity for healthcare applications. It enables the secure transmission and administration of confidential medical information, such as AEFI reports and patient records (Liu et al, 2020; Thakur, 2022). Enhancing confidence and transparency in the healthcare field blockchain's decentralized structure ensures that data is immutable and resistant to modification (Alkan, 2021; McBee and Wilcox, 2020). Blockchain can offer a simple to use, transparent, and safe way to record and track adverse events in the background of AEFI reporting. Further, Data Block Matrix (DBM), a DLT variation that permits controlled data alteration or deletion while ensuring integrity assurance. DBM is particularly useful for applications that must adhere by privacy requirements that require the deletion of user data upon request due to its unique features. The integrity of the data is preserved since the system can keep an immutable ledger of each transaction, guaranteeing that once AEFI data is stored, it cannot be modified or withdrawn (Dhanalakshmi and Babu, 2019). It is important, while that the immutability of blockchain may provide difficulties in maintaining to privacy laws including General Data Protection Regulation (GDPR), which mandate the capability to eliminate personal information upon request (Suripeddi and Purandare, 2021). However, to address this, developments like the Data Block Matrix have been developed, which permits regulated data modification or deletion while maintaining the blockchain's integrity-preserving aspects (Roberts et al., 2023).

Ethical Considerations in AI-Driven AEFI Surveillance

AI-driven AEFI surveillance systems raise important ethical considerations that need to be carefully addressed. Privacy and data protection are key concerns when implementing AI-based surveillance of AEFI. These systems often rely on collecting and analyzing large amounts of personal health data, which must be handled securely and in compliance with regulations (Zeleke et al., 2023). Ensuring public confidence requires transparency, responsibility, accountability and comprehensibility in the AI systems used for AEFI identification and analysis (Cheong, 2024; Habli et al., 2020). The ethical implications of AI applications in healthcare must be carefully assessed since they have the potential to greatly affect patient outcomes, privacy, and overall quality of treatment (Rogers et al., 2021; Solomonides et al., 2022). In order to address concerns like beneficence, nonmaleficence, autonomy, and justice—basic concepts in medical practice—ethical standards for AI in healthcare are required (Solomonides et al., 2022). These suggestions might help in ensuring that sure AI systems are developed and utilized in an approach that prioritizes patient safety first, maintains individual rights, and encourages impartial healthcare findings (Tahri et al., 2023).

Global Implications:Bridging AEFI Reporting Gaps in Low-Resource Settings

Public trust and safety in vaccination programs are the primary concerns, but in low-resource areas with potentially inadequate healthcare infrastructure and diagnostic abilities, it might be difficult to conduct thorough monitoring and reporting of adverse events that occur after immunization (Constantine et al., 2018; Lamichhane and Neupane, 2022). Over the past ten years, aims to improve vaccination safety monitoring systems across the world had made progress, but there are still large gaps, especially in low-resource settings (Salman, 2021). To improve AEFI surveillance, some countries have used digital methods. Australia has established programs such as Vaxtracker and SmartVax, as well as Canada utilizes the Canadian National Vaccine Safety (CANVAS) Network (Psihogios et al., 2022). A major challenge is Healthcare Workers' (HCWs') lack of training and understanding of reportable AEFIs, reporting systems, and processes (Abdu et al., 2022; Aborigo et al., 2022). Research showed that healthcare workers frequently are hesitant to disclose AEFIs as they concern about the consequences for themselves and the public's opinion (Lv et al., 2022; Thomas et al., 2021). The fear of raising unnecessary public alarm about vaccines is a prominent barrier, with 61.76% of healthcare workers citing this as a reason for not reporting AEFIs (Lv et al., 2022). Furthermore, 15.7% of respondents reported that the fear of facing personal consequences discourages them from reporting (Thomas et al., 2021). Reporting is also discouraged by fear of consequences on an individual's part or of getting blamed by superiors (Aborigo et al., 2022; Gidudu et al., 2020). Additionally, AEFI surveillance participation among HCW is decreased by excessive workloads, a lack of enthusiasm, and insufficient feedback on reports that are turned in (Aborigo et al., 2022). Further, AEFI discussion during supervisory visits was substantially linked to an increase in HCW reporting (Gidudu et al., 2020). This emphasizes how crucial it is for leaders to continue to communicate and offer assistance. Structural challenges might be addressed by making reporting procedures simpler, providing rewards, and including AEFI reporting into regular monitoring (Aborigo et al., 2022). Multiple strategies are required to close these gaps. More timely AEFI identification may also be supported by utilizing digital advances and social media monitoring (Gold et al., 2021; Nyambayo et al., 2022). Strong pharmacovigilance programs are ultimately necessary to sustain public confidence in vaccinations and overcome resistance (Nyambayo et al., 2022; Yamoah et al., 2023).

The Future of AEFI Monitoring:Predictive Models and Real-Time Analysis

Technological developments in artificial intelligence and machine learning changed the health care sector while unlocking up possibilities for innovative techniques of post-licensure vaccination safety surveillance (de hond et al., 2022). Increasingly, adverse reactions following vaccination signals are being detected in near real-time by using routinely obtained electronic healthcare data, such as immunization and medical records (Mesfin et al., 2019). In AEFI surveillance, data comparability and standardization are essential. To compare data from surveillance systems and clinical trials, common case definitions and protocols for case classification, recording, and data presentation must be developed (Bonhoeffer et al., 2002).

There is a trend towards implementing active surveillance and other methods into vaccination safety programs, as passive reporting systems have certain limits. These include keeping a look toward new data sources, actively screening for certain illnesses, and utilizing real-time techniques to identify changes in vaccination safety data (Crawford et al., 2014). The utilization of predictive analytics in monitoring unfavorable occurrences after vaccination has increased in significance as healthcare organizations work to improve patient safety and outcomes (Buttery and Clothier, 2022). Predicting and analyzing AEFI data is gradually utilizing machine learning and deep learning algorithms (Sghir et al., 2023). Large numbers of various data forms, including unstructured and structured information from several sources, may be processed according to these strategies. Platforms for real-time streaming, such as Apache Flink, can collect, transfer, and preprocess data streams in order to use deep learning models for predictive analytics (Dinakar and Vagdevi, 2023).

Challenges and Limitations of Digital AEFI Surveillance System

In digital AEFI reporting, completeness and data quality are key considerations. Two major issues that need to be addressed are incomplete AEFI case investigations and delays in prompt AEFI identification (Mesfin et al., 2020). Limitations in some nations are also mentioned, such as the lack of data provided through passive surveillance systems and the delays in investigating into serious adverse events. In particular circumstances, there are problems with regulatory policies, guidance documents, and reporting requirements. For instance, the pharmacovigilance system in Pakistan was lacking enough regulatory policies or reporting requirements at initially (Hagos et al., 2024). The absence of rules, regulations and proclamations pertaining to AEFI surveillance was also identified as a gap in Ethiopia. It could be challenging to accommodate recording methods that are both paper-based and electronic-based. It was determined that in order to successfully execute active AEFI surveillance, a system that can handle both forms was required (Zeleke et al., 2023). Compared with standard reporting systems, social media monitoring has potential as an additional data source, but there are still difficulties in authenticating and analyzing the data (Habibabadi et al., 2023). Underreporting remains a major problem in the field of digital AEFI monitoring systems. For instance, between 1991 and 2001, just 2.7 AEFI incidents per 100,000 vaccination doses were documented in Switzerland, indicating a significant underreporting rate (Schumacher et al., 2010). Similarly, with recent modifications to the surveillance system, the number of major AEFIs reported in India remains significantly lower than predicted (Joshi et al., 2018). According to Zimbabwe's AEFI surveillance system review, incomplete case investigations and delays in AEFI detection are ongoing challenges (Nyambayo et al., 2022). International data comparison is limited by the lack of consistency in AEFI definitions and reporting throughout monitoring systems. According to a comprehensive evaluation of 74 research on maternal immunization, there was an abundance of variation in the definition, estimation, and reporting of AEFIs (Fulton et al., 2015).

CONCLUSION

The integration of AI and digital surveillance in AEFI monitoring has completely transformed vaccine safety surveillance. While traditional systems faced challenges like underreporting, delays, and inconsistent data quality, advances in AI-driven tools, such as machine learning algorithms and natural language processing, have enabled real-time detection and analysis of adverse events. The usage of blockchain protects data security, while social media and mobile health apps have increased the rate and depth of safety signal detection, particularly during health emergencies. Despite these advances, issues like as data privacy, ethical concerns, and poor acceptance in low-resource areas remain.

REFERENCES

As published

Showing references and in-text citations exactly as published.

  1. 1.Abdu, N.; Mosazghi, A.; Yehdego, T.; Tesfamariam, E. H.; Russom, M.. Knowledge and perceptions of nurse practitioners on adverse events following immunization and barriers to reporting in the central region, Eritrea: A cross-sectional study. Drug, Healthcare and Patient Safety. 2022;Volume (14):125–134. https://doi.org/10.2147/DHPS.S363925DOIGOOGLE SCHOLAR
  2. 2.Aborigo, R. A.; Welaga, P.; Oduro, A.; Shaum, A.; Opare, J.; Dodoo, A. et al. Optimising reporting of adverse events following immunisation by healthcare workers in Ghana: A qualitative study in four regions. PLOS One. 2022;17(12):Article e0277197. https://doi.org/10.1371/journal.pone.0277197DOIGOOGLE SCHOLAR
  3. 3.Alkan, B. Ş.. Real-time Blockchain accounting system as a new paradigm. Muhasebe ve Finansman Dergisi. 2021;41–58. https://doi.org/10.25095/mufad.950162DOIGOOGLE SCHOLAR
  4. 4.Allabun, S.; Soufiene, B. O.. Study of the drug-related adverse events with the help of electronic health records and natural language processing. International Journal of Advanced Computer Science and Applications. 2023;14(6). https://doi.org/10.14569/IJACSA.2023.01406148DOIGOOGLE SCHOLAR
  5. 5.An, Q.; Rahman, S.; Zhou, J.; Kang, J. J.. A comprehensive review on machine learning in healthcare industry: Classification, restrictions, opportunities and challenges. Sensors. 2023;23(9):4178. https://doi.org/10.3390/s23094178DOIGOOGLE SCHOLAR
  6. 6.Ball, R.. Perspectives on the future of postmarket vaccine safety surveillance and evaluation. Expert Review of Vaccines. 2014;13(4):455–462. https://doi.org/10.1586/14760584.2014.891941DOIGOOGLE SCHOLAR
  7. 7.Bonhoeffer, J.; Kohl, K.; Chen, R.; Duclos, P.; Heijbel, H.; Heininger, U. et al. The Brighton Collaboration: Addressing the need for standardized case definitions of adverse events following immunization (AEFI). Vaccine. 2002;21(3-4):298–302. https://doi.org/10.1016/S0264-410X(02DOIGOOGLE SCHOLAR
  8. 8.Buttery, J. P.; Clothier, H.. Information systems for vaccine safety surveillance. Human Vaccines and Immunotherapeutics. 2022;18(6):Article 2100173. https://doi.org/10.1080/21645515.2022.2100173DOIGOOGLE SCHOLAR
  9. 9.Carrell, D. S.; Gruber, S.; Floyd, J. S.; Bann, M. A.; Cushing-Haugen, K. L.; Johnson, R. L. et al. Improving methods of identifying anaphylaxis for medical product safety surveillance using natural language processing and machine learning. American Journal of Epidemiology. 2023;192(2):283–295. https://doi.org/10.1093/aje/kwac182DOIGOOGLE SCHOLAR
  10. 10.Cheong, B. C.. Transparency and accountability in AI systems: Safeguarding wellbeing in the age of algorithmic decision-making. Frontiers in Human Dynamics. 2024;6:Article 1421273. https://doi.org/10.3389/fhumd.2024.1421273DOIGOOGLE SCHOLAR
  11. 11.Constantine, M.; Cremance, T.; Juru, T. P.; Gerald, S.; Notion, G. T.; Peter, N. et al. Evaluation of the adverse events following immunization surveillance system in Guruve district, Mashonaland Central 2017. Pan African Medical Journal. 2018;31(1). https://doi.org/10.11604/pamj.2018.31.202.16573DOIGOOGLE SCHOLAR
  12. 12.Crawford, N. W.; Clothier, H.; Hodgson, K.; Selvaraj, G.; Easton, M. L.; Buttery, J. P.. Active surveillance for adverse events following immunization. Expert Review of Vaccines. 2014;13(2):265–276. https://doi.org/10.1586/14760584.2014.866895DOIGOOGLE SCHOLAR
  13. 13.de Hond, A. A. H.; Leeuwenberg, A. M.; Hooft, L.; Kant, I. M. J.; Nijman, S. W. J.; van Os, H. J. A. et al. Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: A scoping review. npj Digital Medicine. 2022;5(1):2. https://doi.org/10.1038/s41746-021-00549-7DOIGOOGLE SCHOLAR
  14. 14.Dhanalakshmi, S.; Babu, G. C.. An examination of big data and blockchain technology. International Journal of Innovative Technology and Exploring Engineering. 2019;8(11):3118–3122. https://doi.org/10.35940/ijitee.K2497.0981119DOIGOOGLE SCHOLAR
  15. 15.Dinakar, J. R.; Vagdevi, S.. Real-time streaming analytics using big data paradigm and predictive modelling based on deep learning. International Journal on Recent and Innovation Trends in Computing and Communication. 2023;11(4s):161–165. https://doi.org/10.17762/ijritcc.v11i4s.6323DOIGOOGLE SCHOLAR
  16. 16.Dong, G.; Bate, A.; Haguinet, F.; Westman, G.; Dürlich, L.; Hviid, A. et al. Optimizing signal management in a vaccine adverse event reporting system: A proof-of-concept with COVID-19 vaccines using signs, symptoms, and natural language processing. Drug Safety. 2024;47(2):173–182. https://doi.org/10.1007/s40264-023-01381-6DOIGOOGLE SCHOLAR
  17. 17.Fulton, T. R.; Narayanan, D.; Bonhoeffer, J.; Ortiz, J. R.; Lambach, P.; Omer, S. B.. A systematic review of adverse events following immunization during pregnancy and the newborn period. Vaccine. 2015;33(47):6453–6465. https://doi.org/10.1016/j.vaccine.2015.08.043DOIGOOGLE SCHOLAR
  18. 18.Ghaffar Nia, N.; Kaplanoglu, E.; Nasab, A.. Evaluation of artificial intelligence techniques in disease diagnosis and prediction. Discover Artificial Intelligence. 2023;3(1):5. https://doi.org/10.1007/s44163-023-00049-5DOIGOOGLE SCHOLAR
  19. 19.Gidudu, J. F.; Shaum, A.; Dodoo, A.; Bosomprah, S.; Bonsu, G.; Amponsa-Achiano, K. et al. Barriers to healthcare workers reporting adverse events following immunization in four regions of Ghana. Vaccine. 2020;38(5):1009–1014. https://doi.org/10.1016/j.vaccine.2019.11.050DOIGOOGLE SCHOLAR
  20. 20.Gold, M. S.; Lincoln, G.; Bednarz, J.; Braunack-Mayer, A.; Stocks, N.. Consumer acceptability and validity of m-health for the detection of adverse events following immunization–the Stimulated Telephone Assisted Rapid Safety Surveillance (STARSS) randomised control trial. Vaccine. 2021;39(2):237–246. https://doi.org/10.1016/j.vaccine.2020.12.011DOIGOOGLE SCHOLAR
  21. 21.Gour, A.. AI-based natural language processing (NLP) systems. Journal of Algebraic Statistics. 2020;11(1):48–58.GOOGLE SCHOLAR
  22. 22.Habli, I.; Lawton, T.; Porter, Z.. Artificial intelligence in health care: Accountability and safety. Bulletin of the World Health Organization. 2020;98(4):251–256. https://doi.org/10.2471/BLT.19.237487DOIGOOGLE SCHOLAR
  23. 23.Hagos, A. A.; Sahile, Z.; Ahmed, W.; Phanouvong, S.. Leveraging COVID-19 vaccine safety monitoring in Ethiopia and Pakistan to enhance system-wide safety surveillance. Global Health: Science and Practice. 2024;12(Suppl. 1). https://doi.org/10.9745/GHSP-D-23-00161DOIGOOGLE SCHOLAR
  24. 24.Hazlehurst, B.; Naleway, A.; Mullooly, J.. Detecting possible vaccine adverse events in clinical notes of the electronic medical record. Vaccine. 2009;27(14):2077–2083. https://doi.org/10.1016/j.vaccine.2009.01.105DOIGOOGLE SCHOLAR
  25. 25.Hu, R.; Liu, Y.; Zhang, L.; Kang, G.; Xu, B.; Li, M. et al. Post-marketing safety surveillance for both CRM197 and TT carrier proteins PCV13 in Jiangsu, China. Frontiers in Public Health. 2023;11:Article 1272562. https://doi.org/10.3389/fpubh.2023.1272562DOIGOOGLE SCHOLAR
  26. 26.Joshi, J.; Das, M. K.; Polpakara, D.; Aneja, S.; Agarwal, M.; Arora, N. K.. Vaccine safety and surveillance for adverse events following immunization (AEFI) in India. The Indian Journal of Pediatrics. 2018;85(2):139–148. https://doi.org/10.1007/s12098-017-2532-9DOIGOOGLE SCHOLAR
  27. 27.Khademi Habibabadi, S.; Delir Haghighi, P.; Burstein, F.; Buttery, J.. Vaccine adverse event mining of Twitter conversations: 2-phase classification study. JMIR Medical Informatics. 2022;10(6):Article e34305. https://doi.org/10.2196/34305DOIGOOGLE SCHOLAR
  28. 28.Khademi Habibabadi, S. K.; Palmer, C.; Dimaguila, G. L.; Javed, M.; Clothier, H. J.; Buttery, J.. Australasian institute of digital health summit 2022-automated social media surveillance for detection of vaccine safety signals: A validation study. Applied Clinical Informatics. Australasian Institute of Digital Health Summit. 2023;14(1):1–10. https://doi.org/10.1055/a-1975-4061DOIGOOGLE SCHOLAR
  29. 29.Kim, S.; Bea, S.; Choe, S.-A.; Choi, N.-K.; Shin, J.-Y.. Autoimmune disorders reported following COVID-19 vaccination: A disproportionality analysis using the WHO database. European Journal of Clinical Pharmacology. 2024;80(3):445–453. https://doi.org/10.1007/s00228-023-03618-wDOIGOOGLE SCHOLAR
  30. 30.Kim, Y.; Jang, J.-H.; Park, N.; Jeong, N.-Y.; Lim, E.; Kim, S. et al. Machine learning approach for active vaccine safety monitoring. Journal of Korean Medical Science. 2021;36(31). https://doi.org/10.3346/jkms.2021.36.e198DOIGOOGLE SCHOLAR
  31. 31.Koutanaei, F. N.; Sajedi, H.; Khanbabaei, M.. A hybrid data mining model of feature selection algorithms and ensemble learning classifiers for credit scoring. Journal of Retailing and Consumer Services. 2015;27:11–23. https://doi.org/10.1016/j.jretconser.2015.07.003DOIGOOGLE SCHOLAR
  32. 32.Laemmle-Ruff, I.; Lewis, G.; Clothier, H. J.; Dimaguila, G. L.; Wolthuizen, M.; Buttery, J. et al. Vaccine safety in Australia during the COVID-19 pandemic: Lessons learned on the frontline. Frontiers in Public Health. 2022;10:Article 1053637. https://doi.org/10.3389/fpubh.2022.1053637DOIGOOGLE SCHOLAR
  33. 33.Lamichhane, B.; Neupane, N.. Improved healthcare access in low-resource regions: A review of technological solutions. 2022.GOOGLE SCHOLAR
  34. 34.Laryea, E. B.; Frimpong, J. A.; Noora, C. L.; Tengey, J.; Bandoh, D.; Sabblah, G. et al. Evaluation of the adverse events following immunization surveillance system, Ghana, 2019. PLOS One. 2022;17(3):Article e0264697. https://doi.org/10.1371/journal.pone.0264697DOIGOOGLE SCHOLAR
  35. 35.Levi, Y.; Brandeau, M. L.; Shmueli, E.; Yamin, D.. Prediction and detection of side effects severity following COVID-19 and influenza vaccinations: Utilizing smartwatches and smartphones. Scientific Reports. 2024;14(1):6012. https://doi.org/10.1038/s41598-024-56561-wDOIGOOGLE SCHOLAR
  36. 36.Li, I.; Pan, J.; Goldwasser, J.; Verma, N.; Wong, W. P.; Nuzumlalı, M. Y. et al. Neural natural language processing for unstructured data in electronic health records: A review. Computer Science Review. 2022;46:Article 100511. https://doi.org/10.1016/j.cosrev.2022.100511DOIGOOGLE SCHOLAR
  37. 37.Li, Y.; Li, J.; He, J.; Tao, C.. AE-GPT: Using large language models to extract adverse events from surveillance reports-a use case with influenza vaccine adverse events. PLOS One. 2024;19(3):Article e0300919. https://doi.org/10.1371/journal.pone.0300919DOIGOOGLE SCHOLAR
  38. 38.Liu, H.; Crespo, R. G.; Martínez, O. S.. Enhancing privacy and data security across healthcare applications using blockchain and distributed ledger concepts. In Healthcare (Vol. 8, No. 3, p. 243). MDPI. 2020;8(3). https://doi.org/10.3390/healthcare8030243DOIGOOGLE SCHOLAR
  39. 39.Lv, H.; Pan, X.; Wang, Y.; Liang, H.; Yu, H.. Barriers to healthcare workers reporting adverse events following immunization in Zhejiang Province, China. Human Vaccines and Immunotherapeutics. 2022;18(5):Article 2083865. https://doi.org/10.1080/21645515.2022.2083865DOIGOOGLE SCHOLAR
  40. 40.McBee, M. P.; Wilcox, C.. Blockchain technology: Principles and applications in medical imaging. Journal of Digital Imaging. 2020;33(3):726–734. https://doi.org/10.1007/s10278-019-00310-3DOIGOOGLE SCHOLAR
  41. 41.Mesfin, Y. M.; Cheng, A.; Lawrie, J.; Buttery, J.. Use of routinely collected electronic healthcare data for postlicensure vaccine safety signal detection: A systematic review. BMJ Global Health. 2019;4(4):Article e001065. https://doi.org/10.1136/bmjgh-2018-001065DOIGOOGLE SCHOLAR
  42. 42.Mesfin, Y. M.; Cheng, A. C.; Enticott, J.; Lawrie, J.; Buttery, J. P.. Use of telephone helpline data for syndromic surveillance of adverse events following immunization in Australia: A retrospective study, 2009 to 2017. Vaccine. 2020;38(34):5525–5531. https://doi.org/10.1016/j.vaccine.2020.05.078DOIGOOGLE SCHOLAR
  43. 43.Nguyen, M. T. H.; Krause, G.; Keller-Stanislawski, B.; Glöckner, S.; Mentzer, D.; Ott, J. J.. Postmarketing safety monitoring after influenza vaccination using a mobile health app: Prospective longitudinal feasibility study. JMIR mHealth and uHealth. 2021;9(5):Article e26289. https://doi.org/10.2196/26289DOIGOOGLE SCHOLAR
  44. 44.Nyambayo, P. P.; Manyevere, R.; Chirinda, L.; Zifamba, E. N.; Marekera, S.; Nyamandi, T., and et al. Descriptive study of the adverse events following immunization (AEFIs) surveillance system in Zimbabwe. 2022.GOOGLE SCHOLAR
  45. 45.Ozonoff, A.; Milliren, C. E.; Fournier, K.; Welcher, J.; Landschaft, A.; Samnaliev, M. et al. Electronic surveillance of patient safety events using natural language processing. Health Informatics Journal. 2022;28(4). https://doi.org/10.1177/14604582221132429DOIGOOGLE SCHOLAR
  46. 46.Pattanayak, S.. Pro deep learning with TensorFlow 2.0: A mathematical approach to advanced artificial intelligence in Python. 2023;109–197. https://doi.org/10.1007/978-1-4842-8931-0DOIGOOGLE SCHOLAR
  47. 47.Poh Teo, S. P.. Arthritis as an adverse event of special interest post COVID-19 vaccine implementation. Aging Pathobiology and Therapeutics. 2021;3(1):10–11. https://doi.org/10.31491/APT.2021.03.050DOIGOOGLE SCHOLAR
  48. 48.Psihogios, A.; Brianne Bota, A. B.; Mithani, S. S.; Greyson, D.; Zhu, D. T.; Fung, S. G. et al. A scoping review of active, participant-centred, digital adverse events following immunization (AEFI) surveillance: A Canadian immunization research network study. Vaccine. 2022;40(31):4065–4080. https://doi.org/10.1016/j.vaccine.2022.04.103DOIGOOGLE SCHOLAR
  49. 49.Roberts, J. D.; DeFranco, J. F.; Kuhn, D. R.. Data block matrix and hyperledger implementation: Extending distributed ledger technology for privacy requirements. Distributed Ledger Technologies: Research and Practice. 2023;2(2):1–11. https://doi.org/10.1145/3585539DOIGOOGLE SCHOLAR
  50. 50.Rogers, W. A.; Draper, H.; Carter, S. M.. Evaluation of artificial intelligence clinical applications: Detailed case analyses show value of healthcare ethics approach in identifying patient care issues. Bioethics. 2021;35(7):623–633. https://doi.org/10.1111/bioe.12885DOIGOOGLE SCHOLAR
  51. 51.Salman, O.; Topf, K.; Chandler, R.; Conklin, L.. Progress in immunization safety monitoring-Worldwide, 2010–2019. MMWR. Morbidity and Mortality Weekly Report. 2021;70(15):547–551. https://doi.org/10.15585/mmwr.mm7015a2DOIGOOGLE SCHOLAR
  52. 52.Schumacher, Z.; Bourquin, C.; Heininger, U.. Surveillance for adverse events following immunization (AEFI) in Switzerland-1991-2001. Vaccine. 2010;28(24):4059–4064. https://doi.org/10.1016/j.vaccine.2010.04.002DOIGOOGLE SCHOLAR
  53. 53.Sghir, N.; Adadi, A.; Lahmer, M.. Recent advances in Predictive Learning Analytics: A decade systematic review (2012-2022). Education and Information Technologies. 2023;28(7):8299–8333. https://doi.org/10.1007/s10639-022-11536-0DOIGOOGLE SCHOLAR
  54. 54.Sidey-Gibbons, J. A.; Sidey-Gibbons, C. J.. Machine learning in medicine: A practical introduction. BMC Medical Research Methodology. 2019;19:1–18.GOOGLE SCHOLAR
  55. 55.Solomonides, A. E.; Koski, E.; Atabaki, S. M.; Weinberg, S.; McGreevey, J. D.; Kannry, J. L. et al. Defining AMIA’s artificial intelligence principles. Journal of the American Medical Informatics Association. 2022;29(4):585–591. https://doi.org/10.1093/jamia/ocac006DOIGOOGLE SCHOLAR
  56. 56.Sun, Y. X.; Liu, Z. K.; Nie, X. L.; Zhan, S. Y.. Review of near real-time vaccine safety surveillance [Zhonghua liu Xing Bing xue za zhi=]. Zhonghua Liu Xing Bing Xue Za Zhi. 2021;42(2):351–356.GOOGLE SCHOLAR
  57. 57.Suragh, T. A.; Lamprianou, S.; MacDonald, N. E.; Loharikar, A. R.; Balakrishnan, M. R.; Benes, O. et al. Cluster anxiety-related adverse events following immunization (AEFI): An assessment of reports detected in social media and those identified using an online search engine. Vaccine. 2018;36(40):5949–5954. https://doi.org/10.1016/j.vaccine.2018.08.064DOIGOOGLE SCHOLAR
  58. 58.Suripeddi, M. K. S.; Purandare, P.. In Journal of Physics: Conference Series. IOP Publishing. 2021;1964(4):42005. https://doi.org/10.1088/1742-6596/1964/4/042005DOIGOOGLE SCHOLAR
  59. 59.Tahri Sqalli, M.; Aslonov, B.; Gafurov, M.; Nurmatov, S.. Humanizing AI in medical training: Ethical framework for responsible design. Frontiers in Artificial Intelligence. 2023;6:Article 1189914. https://doi.org/10.3389/frai.2023.1189914DOIGOOGLE SCHOLAR
  60. 60.Thakur, A.. A comprehensive study of the trends and analysis of distributed ledger technology and blockchain technology in the healthcare industry. Frontiers in Blockchain. 2022;5:Article 844834. https://doi.org/10.3389/fbloc.2022.844834DOIGOOGLE SCHOLAR
  61. 61.Thomas, R. A.; Rajan Joseph, M. R.; Castilloux, A.-M.; Moride, Y.. Understanding reporting practices and perceptions of barriers in adverse events following immunisation surveillance: A cross-sectional survey of paediatricians in Kerala, India. Vaccine. 2021;39(33):4678–4684. https://doi.org/10.1016/j.vaccine.2021.06.052DOIGOOGLE SCHOLAR
  62. 62.Westphal, D. W.; Williams, S. A.; Leeb, A.; Effler, P. V.. Continuous active surveillance of adverse events following immunisation using SMS technology. Vaccine. 2016;34(29):3350–3355. https://doi.org/10.1016/j.vaccine.2016.05.015DOIGOOGLE SCHOLAR
  63. 63.Xu, M.; Zhao, F.; Zou, Y.; Liu, C.; Cheng, X.; Dressler, F.. BLOWN: A blockchain protocol for single-hop wireless networks under adversarial SINR. IEEE Transactions on Mobile Computing. 2023;22(8):4530–4547. https://doi.org/10.1109/TMC.2022.3162117DOIGOOGLE SCHOLAR
  64. 64.Yamoah, P.; Mensah, K. B.; Padayachee, N.; Bangalee, V.; Oosthuizen, F.. Assessment of adherence to pre-vaccination precautions and AEFI reporting practices during BCG vaccination in 4 hospitals in Ghana. Human Vaccines and Immunotherapeutics. 2023;19(1):Article 2199654. https://doi.org/10.1080/21645515.2023.2199654DOIGOOGLE SCHOLAR
  65. 65.Young, I. J. B.; Luz, S.; Lone, N.. A systematic review of natural language processing for classification tasks in the field of incident reporting and adverse event analysis. International Journal of Medical Informatics. 2019;132:Article 103971. https://doi.org/10.1016/j.ijmedinf.2019.103971DOIGOOGLE SCHOLAR
  66. 66.Zeleke, E. D.; Yimer, G.; Lisanework, L.; Chen, R. T.; Huang, W.-T.; Wang, S.-H. et al. System and facility readiness assessment for conducting active surveillance of adverse events following immunization in Addis Ababa, Ethiopia. International Health. 2023;15(6):676–683. https://doi.org/10.1093/inthealth/ihac085DOIGOOGLE SCHOLAR
  67. 67.Zeng, D.; Cao, Z.; Neill, D. B.. Artificial intelligence-enabled public health surveillance-From local detection to global epidemic monitoring and control. In Artificial Intelligence in Medicine. Academic Press. 2021;437–453.GOOGLE SCHOLAR

Cite this article

SELECT FORMAT

Siddiqui, S., Ramesh, M., & Narayana, P. S. (2026). Digital Surveillance and Artificial Intelligence in Detection of AEFI: A New Frontier in Vaccine Safety Monitoring. International Journal of Pharmaceutical Investigation, 16(1), 51–57. https://doi.org/10.5530/ijpi.20260414