© 2024. Published by AHFE Open Access. All rights reserved. 193
INTRODUCTION
Early identification of clusters of health-related events is a key component of effective public health interventions. Traditional surveillance methods have relied on manual data collection and analysis often resulting in delays. The frequency and magnitude of emerging health threats and global pan- demics has created a pressing need to find innovative solutions to surveillance challenges. The introduction of Artificial Intelligence (AI) presents an oppor- tunity to enhance health surveillance by providing timely insights into a wide array of health conditions (Yoon et al., 2017). AI technologies, driven by machine learning algorithms, have demonstrated the capacity to process vast amounts of data in real-time (Yoon et al., 2017). The applicability AI is best highlighted through case-use examples, showcasing how real- time insights facilitate timely intervention, resource allocation, and inform public health messaging. The purpose of this paper is to explore emerg- ing applications of AI in detecting clusters of health conditions and to assess the potential impact of AI on proactive intervention and prevention strategies.
INFECTIOUS DISEASE SURVEILLANCE AI systems have demonstrated increasing capabilities in syndromic surveil- lance for infectious diseases. Case-use examples serve to illustrate the tangible advantages of integrating AI into infectious disease surveillance systems. AI systems have detected potential outbreaks before laboratory confirmation by analysing data from healthcare facilities, laboratories, and self-reported symptoms (Lucero-Obusan et al., 2017). AI algorithms have detected pat- terns of various emerging health threats and facilitated early intervention (Nolan et al., 2017). These early warning systems are being increasingly utilized to curb the rapid spread of infectious diseases (Lucero-Obusan et al., 2017). AI- augmented analysis of emergency department visits has provided insights into the spread and treatment compliance of zoonotic infections such as rabies (Bemis et al., 2017). Through early detection and warning systems, AI systems have given health authorities actionable data to inform targeting vaccination campaign to the populations and areas at highest risk (Lucero-Obusan et al., 2017). This targeted resource allocation is essential in strengthening the capacity of healthcare systems to address the needs of the communities they serve in times of crisis (Nolan et al., 2017). By providing real-time insights into the dynamics of an outbreak, AI informs tailored and up-to-date public health messages (Nolan et al., 2017). The spread of vector-borne diseases is closely related to environ- mental factors. AI systems have been leveraged to analyse disease inci- dence, historical data, and weather reports, and to detect patterns related to the spread of vector-borne diseases (Pley et al., 2021). By integrating changes in temperature and humidity into the analysis, AI enhances our capacity to predict spread of vector borne diseases in high-risk areas (Pley et al., 2021). As we navigate the emergence and spread of infectious dis- eases, AI will serve as a key tool to safeguard health at a regional and global level.
Artificial Intelligence for Cluster Detection and Targeted Intervention in Healthcare 195
MENTAL HEALTH DISORDERS AI augmented analysis has been applied to identify clusters of mental health concerns. Sources of input data that have informed these algorithms include social media posts, emergency hotline calls, and hospital admissions (Thieme et al., 2020). Social media platforms provide valuable data about language patterns, sentiment analysis, and behavioural cues, that can be provide insights into early signs of mental health distress in particular com- munities and subgroups of the population (Graham et al., 2019). Analysing the frequency and nature of hotline calls enables AI to identify trends in men- tal health and responses to particular events and stressors (Graham et al., 2019). AI algorithms analyse social media posts, emergency hotline calls, and hospital admissions to identify clusters related to issues such as overdoses, suicides, and burnout, enabling timely and targeted intervention strategies (Thieme et al., 2020). These insights are instrumental in planning community outreach and additional resources where they are needed most. By analysing mental health related hospital admissions, AI can assess the capacity and needs of different health regions (Nolan et al., 2017). Through the insightful interpretation of these diverse datasets, AI offers a nuanced understanding of the evolving mental health landscape.
CHRONIC DISEASE AND PHARMACOVIGILANCE Chronic diseases pose a growing healthcare challenge. Subtle trends in chronic diseases often require the analysis of tremendous amounts of detailed data. AI systems have demonstrated the capability to efficiently detect these subtle patters to often go unrecognized through traditional health surveillance methods (Thiébaut and Cossin 2019). Chronic disease man- agement often involves various pharmaceutical combinations and sched- ules (Thiébaut and Cossin 2019). Close monitoring of these combinations is essential for safety monitoring., making it imperative to monitor and understand the safety profiles of various medications (Thiébaut and Cossin 2019). By promptly recognizing Adverse Drug Reactions patterns, AI can assist in optimizing patient care, minimizing risks, and maximizing therapeu- tic benefits (Liang et al., 2022). These insights will allow for more informed decision making for healthcare-providers, regulatory bodies, and researchers (Liang et al., 2022). By harnessing the power of AI to better understand intri- cate patterns in chronic disease data and identify gaps in pharmacovigilance, healthcare systems are empowered to implement proactive and data-driven interventions (Thiébaut and Cossin 2019).
ENVIRONMENTAL HEALTH RISKS In the field of environmental health, AI offers capabilities to better under- stand correlations between environmental factors and public health outcomes (Fan et al., 2023). By cross-referencing health data with environmental fac- tors, AI can identify associations between exposure to certain pollutants or environmental conditions and specific health issues (Yoon et al., 2017). Air-quality monitors provide a real-time data on pollutants and particulate
matter concentrations in the atmosphere (Yoon et al., 2017). AI can provide predictions of prevalence of respiratory conditions during periods of poor air quality and inform health system surge planning (Fan et al., 2023). In these ways, AI is expected to be a valuable tool in safeguarding public health in evolving environmental challenges.
CHALLENGES AND RECOMMENDATIONS Despite the far-reaching potential applications of AI in health surveillance, various challenges exist. The sensitive nature of health data privacy must be considered when balancing the risks and benefits of utilizing health infor- mation for public health purposes (Yoon et al., 2017). The effectiveness of AI algorithms is dependent on both the quantity and the quality of avail- able data (Kalra and Seitzinger 2022). Accessing data from various sources presents logistical challenges and introduces the potential for misleading and biased results. The collection of health data is often healthcare systems (Yoon et al., 2017). Addressing these challenges requires collaborative efforts to standardize data collection. Establishing effective communication channels between AI systems and public health authorities is an ongoing challenge that necessitates interdisciplinary collaboration and continuous refinement. The balance between public health objectives and individual privacy may be found in anonymizing health data (Yoon et al., 2017). This would pro- tect the confidentiality of personal information while still allowing for the identification of clusters and health needs at a population level.
CONCLUSION
The integration of AI into an all-risks approach to syndromic surveillance represents an important step in a proactive approach to addressing clusters of health-related events. This approach has diverse applications in infectious diseases, vector-borne diseases, mental health disorders, chronic diseases, and environmental health risks. These applications will have important implica- tions on shaping policies, public health campaigns, and regulatory measures. It is crucial to navigate challenges surrounding data privacy and quality to ensure the highest ethical standards privacy considerations are maintained while also protecting the health of the population. Continued research and collaboration are required to appropriately realize the potential applications of AI in shaping public health strategy.
REFERENCES
Bemis, K., Frias, M., Patel, M. T. and Christiansen, D. (2017). Using an Emergency Department Syndromic Surveillance System to Evaluate Reporting of Potential Rabies Exposures, Illinois, 2013–2015. Public Health Reports [online], 132(1), pp. 59S–64S. /pmc/articles/PMC5676512/ Fan, Z., Yan, Z. and Wen, S. (2023). Deep Learning and Artificial Intelligence in
Sustainability: A Review of SDGs, Renewable Energy, and Environmental Health. Sustainability 2023, Vol. 15, Page 13493 [online], 15(18), p. 13493. https://www. mdpi.com/2071-1050/15/18/13493/htm
Artificial Intelligence for Cluster Detection and Targeted Intervention in Healthcare 197
Graham, S., Depp, C., Lee, E. E., Nebeker, C., Tu, X., Kim, H. C. and Jeste, D. V. (2019). Artificial Intelligence for Mental Health and Mental Illnesses: An Overview. Current psychiatry reports [online], 21(11), p. 116. /pmc/arti- cles/PMC7274446/ Kalra, J. and Seitzinger, P. (2022). Implications and Consequences of Artificial Intel-
ligence in Healthcare Quality and Medical Training. Healthcare and Medical Devices [online], 51(51). https://openaccess.cms-conferences.org/publications/bo ok/978-1-958651-27-8/article/978-1-958651-27-8_18 Liang, L., Hu, J., Sun, G., Hong, N., Wu, G., He, Y., Li, Y., Hao, T., Liu, L. and Gong, M. (2022). Artificial Intelligence-Based Pharmacovigilance in the Setting of Limited Resources. Drug Safety [online], 45(5), pp. 511–519. /pmc/ar- ticles/PMC9112260/ Lucero-Obusan, C., Winston, C. A., Schirmer, P. L., Oda, G. and Holodniy,
M. (2017). Enhanced Influenza Surveillance Using Telephone Triage and Elec- tronic Syndromic Surveillance in the Department of Veterans Affairs, 2011-2015. Public Health Reports [online], 132(1), pp. 16S–22S. /pmc/articles/PMC5676515/ Nolan, M. L., Kunins, H. V., Lall, R. and Paone, D. (2017). Developing Syndromic Surveillance to Monitor and Respond to Adverse Health Events Related to Psychoactive Substance Use: Methods and Applications. Public Health Reports [online], 132(1), pp. 65S–72S. /pmc/articles/PMC5676520/ Pley, C., Evans, M., Lowe, R., Montgomery, H. and Yacoub, S. (2021). Digital and
technological innovation in vector-borne disease surveillance to predict, detect, and control climate-driven outbreaks. The Lancet Planetary Health [online], 5(10), pp. e739–e745. https://www.thelancet.com/article/S2542519621001418/ fulltext Thiébaut, R. and Cossin, S. (2019). Artificial Intelligence for Surveillance in Public Health. Yearbook of medical informatics [online], 28(1), pp. 232–234. /pmc/arti- cles/PMC6697516/ Thieme, A., Belgrave, D. and Doherty, G. (2020). Machine Learning in Mental
Health: A systematic review of the HCI literature to support the development of effective and implementable ML Systems. ACM Transactions on Computer- Human Interaction [online], 27(5), p. 34. https://doi.org/10.1145/3398069 Yoon, P. W., Ising, A. I. and Gunn, J. E. (2017). Using Syndromic Surveillance for All- Hazards Public Health Surveillance: Successes, Challenges, and the Future. Public Health Reports, 132(1), pp. 3S–6S.