INTRODUCTION
Technology is playing an increasing role in the delivery of healthcare. The appropriate implementation of new technologies is a delicate balance of managing risk and meeting the needs of the population. Circumstances sur- rounding healthcare delivery are evolving through technological advances, demographic changes, and increases in the expectations of patients and soci- ety (Seitzinger et al. 2021). With the ever-increasing need for more accurate and precise healthcare delivery, the integration of Artificial Intelligence (AI) presents a promising opportunity to strengthen healthcare services.
© 2022. Published by AHFE Open Access. All rights reserved. 138
Implications and Consequences of Artificial Intelligence 139
Research into AI in medicine has grown exponentially in the past several decades. AI refers to the capacity of a machine to imitate intelligent behavi- our. AI can be divided into several major subfields such as machine learning and deep learning, each with unique capabilities and potential applications in healthcare (Davenport and Kalakota 2019; Amisha et al. 2019). Mach- ine learning refers to programs that use algorithms to modify themselves and extract data from unlabeled data. Unlike most modern machines that rely on predefined rules, machine learning can utilize an algorithm to learn new rules from large data sets and executive functions from the learned rules (Kalinin et al. 2018). Deep Learning refers to neural networks consisting of layered sets of algorithms that recognize patterns, modifying their algorithms independently.
The introduction of AI into healthcare services may be considered an innovative disruption, which with careful implementation, can strengthen healthcare quality. Transitioning into an era of AI-assisted healthcare servi- ces will have significant implications on the efficiency, quality, and justice of the healthcare system. Therefore, the implementation of AI warrants care- ful deliberation to ensure that impacts are considered and consequences are mitigated.
CURRENT STATE OF AI IN HEALTHCARE Any path in healthcare development involves inherent advantages and dra- wbacks. Maintaining the status quo of healthcare delivery systems is no exception. Currently, healthcare relies primarily on humans to recognize patterns and diagnose disease. Our previous studies have shown an ove- rall discordance rate of 12.7% between clinical and autopsy diagnoses (Kalra et al. 2020). More importantly, in 10.2% of cases, knowledge of the diagnosis post-mortem would have changed treatment to prolong disease or cure disease (Kalra et al. 2020). Previous studies have shown that 90% of diagnostic errors are attributable to cognitive factors, including faulty information synthesis and premature closure (Graber et al. 2005). Due to competing demands on physicians’ time, the duration of physicians patient encounters is dwindling, contributing to poor patient outcomes and clim- bing rates of burnout among healthcare providers (Seitzinger et al. 2021). With increasingly complex healthcare systems, medical error and disclosure have become a considerable issue around the globe (Kalra et al. 2020). These findings suggest a need for innovative diagnostic tools in healthcare quality.
AI systems can increase workflow without compromising the accurate identification of abnormalities (Paiva and Prevedello 2017). The limitations of fatigue can be mitigated by using AI (Pesapane et al. 2018). Computerized systems are unbiased and are not affected by fatigue like the human body. Automated AI programs can continue to work 24 hours a day and 7 days a week, increasing the output of results and minimizing interpretation bias. These systems have shown promise in serving as a co-pilot for clinicians. By predicting and identifying potential medical errors, clinicians can be alerted to reevaluate and mitigate errors before they occur. By taking over some more mundane healthcare tasks, time can be freed up to allow clinicians to focus on
140 Kalra and Seitzinger
humanistic aspects of medical care, which may help reduce burnout (Kalra et al. 2021). As alluring as the potential benefits of AI may be, they also present logistical and ethical implications.
The advent of AI in healthcare carries many ethical and accountability issues. There can often be a conflict between an AI and the ethical principles of medicine including autonomy, justice, beneficence, and non-maleficence. AI systems incorporate and often amplify existing patterns of practice, inclu- ding societal biases and inequitable healthcare practices. The momentum created by such innovations can lead to premature optimism and unintentio- nal consequences. AI systems incorporate and often amplify existing patterns of practice, including societal biases and inequitable healthcare practices.
A growing body of literature has been exploring the issue of moral accoun- tability and AI (Habli et al. 2020; Oshana 2004; Beil et al. 2019). The complexity of AI systems limits the ability of patients to provide informed consent. AI systems generating a result are unable to explain their process, limiting a physician’s ability to detect any error that might have occurred. This can make accountability difficult in the case of any potential or unwan- ted outcomes (Felländer-Tsai 2020). AI systems are programmed to function within pre-set parameters, leading to different conclusions for each patient. As a decision-making tool, AI is only as accurate as the data with which it is provided. Transparency in healthcare practice plays a significant role in pati- ent decision-making, and the lack thereof can make decision-making more complicated.
Many who oppose the implementation of AI argue that the system cannot provide holistic care due to the lack of emotion and ability to detect contex- tual cues needed in clinical situations (Loder and Nicholas 2018; Parks 2010). This can make it difficult for patients receiving care as it often lacks empa- thy, which is necessary when delivering specific outcomes. If implemented incorrectly, these technologies may exacerbate health disparities, disempower patients, and reduce the humanity of medical practice (Kalra et al. 2021).
MEDICAL EDUCATION Healthcare educators are tasked with preparing learners across all healthcare disciplines to function in an increasingly technological and rapidly evolving field of practice. This entails instilling learners with digital literacy to leverage new tools and acknowledge limitations. The Institute of Medicine report in 2001 indicated that to increase the quality of healthcare provided, health clinicians required the ability to utilize technology in healthcare settings (Weber 2006).
Many healthcare providers have not been educated on the use of adva- nced technology, which adds to their hesitation when using complex systems like AI to deliver patient care. Much of the current medical education is outdated and lacks the training necessary for the ever-evolving changes in healthcare practices (Paranjape et al. 2019). The curriculums at medical sch- ools should be reevaluated and structured to allow students to gain a better grasp in terms of digital literacy. Many healthcare providers are reluctant to adopt AI systems in the workplace due to a lack of knowledge and fear of imperfections in AI (Chan and Zary 2019).
Implications and Consequences of Artificial Intelligence 141
The rapid advancement of technology in healthcare systems requires an educational model that can be accessed by current healthcare providers. Empowering healthcare providers with knowledge will improve efficiency in the workplace and increase the quality of care delivered to patients (Holt et al. 2020). Traditionally, medical trainees would require hands-on experience with patients to gain the necessary competencies. With growing technolo- gical advances in virtual reality and augmented reality, new opportunities have been presented to allow learners to practice skills in simulated clini- cal environments. Incorporating AI into virtual reality provides real-time medical tutoring to guide learners through situations that mimic real-life cli- nical scenarios. Studies have found that AI-assisted virtual reality teaching technologies improve the competencies of medical trainees and accurately evaluate their skills levels (Winkler-Schwartz et al. 2019). A modern educa- tional curriculum is necessary to keep up with the advancing technology in the healthcare field.
A STRATEGIC PATH FORWARD The potential of AI makes it an essential component of healthcare in the future. Healthcare providers must be ready for the change in practice that AI will bring before it arrives. The next step in the AI system will be to incorpo- rate more human distinctions such as ethics, morality, and emotion to better mimic human healthcare providers (Noorbakhsh-Sabet et al. 2019).
The goal of AI in the future should be centred on addressing the patient’s needs while maintaining the best quality of care practice. We suggest that using AI correctly has the potential to enhance the efficiency and quality of healthcare delivery. The future of AI should aim at creating a system that can work beyond the pre-determined data and carve an independent path when completing tasks. To uncover the true potential of AI, current practices should incorporate AI into the industry to better assess and understand the system.
Although missteps cannot be eliminated from health processes, medical training programs worldwide must remain diligent and persistent in impro- ving the quality of education in a context relevant to current and future healthcare practice. Healthcare providers will need to find a mutually bene- ficial balance between the use of AI and humans in the healthcare industry. For AI to be appropriately utilized, the legal, ethical, and accountability issues must be addressed by implementing regulations at both the level of practice and governing oversight. A framework for medical colleges to prepare lear- ners to practice in an era of AI-assisted medicine will determine the quality of education that they deliver and the sustainability of their practices. Continu- ous revaluation of the current state and directions of medical training quality will allow current systems to grow and adapt while creating opportunities for excellence and innovation.
CONCLUSION
Technology is playing an increasing role in the delivery of healthcare. The appropriate implementation of new technologies is a delicate balance of
142 Kalra and Seitzinger
managing risk and meeting the emerging needs of the population. Health- care providers will need to be prepared for the age of AI and find a mutually beneficial balance between the use of AI and the use of humans in the health- care industry. If implemented incorrectly, these technologies may exacerbate health disparities, disempower patients, and reduce the humanity of medical practice.
REFERENCES
Amisha, Malik, P., Pathania, M. and Rathaur, V. (2019), “Overview of Artificial Intelligence in Medicine: A Necessary Step in Healthcare Improvement.” Journal of Family Medicine and Primary Care [online], 8(7), 2328. Beil, M., Proft, I., van Heerden, D., Sviri, S. and van Heerden, P.V. (2019), “Ethical
considerations about artificial intelligence for prognostication in intensive care.” Intensive Care Medicine Experimental [online], 7(1), 70. Chan, K.S. and Zary, N. (2019), “Applications and Challenges of Implementing Artificial Intelligence in Medical Education: Integrative Review. JMIR Medical Education [online], 5(1), 13930. Davenport, T. and Kalakota, R. (2019), “The Potential for Artificial Intelligence in
Healthcare.” Future Healthcare Journal, 6(2), 94–98. Felländer-Tsai, L. (2020), “AI Ethics, Accountability, and Sustainability: Revisiting the Hippocratic Oath.” Acta Orthopaedica [online], 91(1), 1–2. Graber, M.L., Franklin, N. and Gordon, R. (2005), “Diagnostic error in internal
medicine.” Archives of Internal Medicine [online], 165(13), 1493–1499. Habli, I., Lawton, T. and Porter, Z. (2020), “Artificial intelligence in health care: Accountability and safety.” Bulletin of the World Health Organization [online], 98(4), 251–256. Holt, K.A., Overgaard, D., Engel, L.V. and Kayser, L. (2020), “Health Literacy,
Digital Literacy and eHealth Literacy in Danish Nursing Students at Entry and Graduate level: A Cross Sectional Study.” BMC Nursing [online], 19(1), 22. Kalinin, A.A., Higgins, G.A., Reamaroon, N., Soroushmehr, S., Allyn-Feuer, A., Dinov, I.D., Najarian, K. and Athey, B.D. (2018), “Deep Learning in Pharma- cogenomics: From Gene Regulation to Patient Stratification. Pharmacogenomics [online], 19(7), 629–650. Kalra, J. (Jay), Rafid-Hamed, Z. and Seitzinger, P. (2020), “Disclosure of Medical
Error: A Necessary Step in Healthcare Improvement.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergonomics in Healthcare and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 11–16. Kalra, J. (Jay), Rafid-Hamed, Z. and Seitzinger, P. (2021), “Artificial Intelligence and Humanistic Medicine: A Symbiosis.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergonomics in Healthcare and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 3–8. Kalra, J., Markewich, D. and Seitzinger, P. (2020), “Quality assessment and mana-
gement: An overview of concordance and discordance rates between clinical and autopsy diagnoses.” In: Lightner N., Kalra J. (eds). Advances in Human Factors and Ergonomics in Healthcare and Medical Devices. AHFE 2019. Advances in Intelligent Systems and Computing. Cham: Springer, 857. 45–54. Loder, J. and Nicholas, L. (2018), “Confronting Dr Robot Creating a peoplepowered future for AI in health. 1–38.
Implications and Consequences of Artificial Intelligence 143
Noorbakhsh-Sabet, N., Zand, R., Zhang, Y. and Abedi, V. (2019), “Artificial Intelligence Transforms the Future of Health Care.” American Journal of Medicine, 132(7), 795–801. Oshana, M. (2004), “Moral Accountability”. Philos. Top. 32, 255–274. Paiva, O.A. and Prevedello, L.M. (2017), “The potential impact of artificial intelli-
gence in radiology.” Radiologia Brasileira [online], 50(5), 5–6. Paranjape, K., Schinkel, M., Nannan Panday, R., Car, J. and Nanayakkara, P. (2019), “Introducing Artificial Intelligence Training in Medical Education.” JMIR medical education [online], 5(2), 16048. Parks, J.A. (2010), “Lifting the Burden of Women’s Care Work: Should Robots
Replace the “Human Touch”?” Hypatia [online], 25(1), 100–120. Pesapane, F., Codari, M. and Sardanelli, F. (2018), “Artificial Intelligence in Medical Imaging: Threat or Opportunity? Radiologists Again at the Forefront of Innovation in Medicine.” European Radiology Experimental [online], 2(1), 35. Seitzinger, P., Rafid-Hamed, Z. and Kalra, J. (Jay), “(2021), “Global Trends in Clini-
cal Practice and Healthcare Delivery: Opportunities for Growth and Innovation.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergono- mics in Healthcare and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 9–15. Weber, J.M.K. (2006), “Core Competencies for Health Care Professionals.” Journal of Allied Health [online], 35(3), 185. Winkler-Schwartz, A., Yilmaz, R., Mirchi, N., Bissonnette, V., Ledwos, N., Siyar,
S., Azarnoush, H., Karlik, B. and Del Maestro, R. (2019), “Machine Lear- ning Identification of Surgical and Operative Factors Associated with Surgical Expertise in Virtual Reality Simulation.” JAMA Network Open [online], 2(8), 198363–198363.