© 2023. Published by AHFE Open Access. All rights reserved. 7
8 Kalra and Seitzinger
INTRODUCTION
Healthcare delivery is a complex and ever-evolving process. Various factors including advancing technologies, aging populations, time pressure, and dia- gnostic errors pose challenges to ensuring the best possible care is provided to all patients in all circumstances (Seitzinger et al. 2021). Healthcare systems are tasked with finding innovative ways to address modern healthcare quality challenges while ensuring patient safety in a sustainable and adaptable man- ner. As the complexity and strain on healthcare systems increases, traditional healthcare quality assurance practices require constant re-evaluation and modernization. Each health encounter, near miss and suboptimal outcome, is an opportunity to recognize deficits and improve practice strategies. Staffing shortages and competing priorities often result in problematic patterns going unrecognized. The amount of healthcare data available is insurmountable given the traditional methods of healthcare quality assurance and impro- vement. As a result, healthcare systems are left struggling to react to the downstream effects of systemic factors without the time or resources to act proactively to address route causes. AI can be described as the self-learning computer algorithms which carry out processes characteristically performed by humans. The capabilities demonstrated by modern AI systems make it a promising endeavour to alleviate many of the data management and inter- pretation challenges faced by healthcare systems and to enhance the calibre of patient care around the world.
CHALLENGES IN MODERN HEALTHCARE DELIVERY Modern healthcare delivery systems can be characterized by complex challen- ges with various contributing and interacting factors such as time pressure, burnout, inadequate medical error disclosure and suboptimal patient outco- mes (Kalra et al. 2020, Kalra et al. 2019). The average duration of a primary care patient encounter in the United States ranges between 9.2- 21.2 minutes and in Canada ranges between 9.0 and 15.5 minutes (Irving et al. 2017). These are likely to be overestimates as evidence has shown that physicians overestimate time spent with patients by 175-200% (Gottschalk and Flocke 2005, Becker et al. 2010). During patient-physician interacti- ons, patients are interrupted while sharing their health concerns within an average of 23.1 seconds (Marvel et al. 1999). The time available for face- to-face interactions with patients and families is dwindling (Seitzinger et al. 2021a). Instead, considerable time, stress and resources are being allocated to tasks such as documentation, synthesizing information, administration and trying to remain up to date on the plethora of new evidence being published on a daily basis. Staffing shortages are becoming an increasing concern in North America and around the world. Among clinicians that stay in practice, 72% report emotional exhaustion, 63.2% report depersonalization and 72% report overall burnout (Rotenstein et al. 2018). Even among medical trainees 30-50% report feeling burnout (Rodrigues et al. 2018). The reconceptuali- zation of healthcare quality and delivery strategies is not only timely but is imminently necessary to ensure a sustainable efficacious path forward for the field of medical care.
Expanding Our Grasp: Artificial Intelligence as the Next Leap Forward in Healthcare Quality 9
A SPRINGBOARD FOR HEALTHCARE IMPROVEMENT The advantages of AI in healthcare are growing at an exponential rate, with new applications being realized daily. The advantages of AI include efficiency, accuracy, prediction/modelling, and standardization of care (Seitzinger et al. 2021b). These systems are immune to fatigue and have the potential to take over many of the mundane and less desirable aspects of clinicians’ daily tasks that contribute to burnout (Kalra et al. 2021). The self-learning nature of these systems allows them to persistently improve accuracy while adapting to new data and applying the newest evidence-based guidelines. Examples of AI systems currently being utilized to improve the efficiency of healthcare tasks include AI transcription technology widely used to increase the efficiency of clinical documentation (Seitzinger and Kalra 2022). AI is being implemen- ted to detect sepsis earlier in the clinical course and alert clinicians to cases that may otherwise be missed (Schinkel et al. 2019). In Laboratory medicine and pathology, AI systems are being used for the histological identification of tumours, with promising precision and efficiency (Chang et al. 2019). Alth- ough the introduction of AI systems into healthcare systems is a relatively novel phenomenon, the applications implemented thus far have demonstra- ted levels of accuracy comparable to that of human clinicians (Seitzinger and Kalra, 2022).
A CAUTIOUS AND CONSCIENTIOUS LEAP FORWARD As with all leaps forward in healthcare quality and delivery, careful conside- ration and tactful implementation is required to ensure patients receive the best possible care at all points in time. As healthcare quality strategies utilize new tools to adapt to modern healthcare challenges, there will inevitably be growing pains in domains related to technical challenges, ethical concerns, and legislative clarifications. developmental costs, updating legislation, inte- gration issues, and a need for insufficient digital literacy. Technical aspects of AI will need to be addressed including development costs, transparency, data security, integration and increasing digital literacy.
From an ethical perspective, the implementation of AI in healthcare rai- ses concerns regarding informed consent, accountability, non-maleficence and autonomy (Loder and Nicholas 2018, Parks 2010). Patients should be afforded the opportunity to receive the best calibre of care possible, with the utilization of available tools and resources. They should also have the opportunity to understand the options available to them and how new tech- nologies may contribute to or jeopardize their care (Beil et al. 2019). Laws and regulations may require adaptation and revision to clarify issues around liability around clinicians’ decisions to use or disregard interpretations and suggestions provided by new technologies (Wachter et al. 2017, Cath 2018). Navigating these challenges while preventing undue delays in progressing medicine will require a shift in approaches to healthcare delivery and medical education. Clinicians and trainees require ongoing digital literacy training in order to appropriately utilize new technologies to enhance healthcare qua- lity (Weber 2006). Traditional methods of medical education lack the digital literacy training required to practice effectively in the upcoming AI-assisted
10 Kalra and Seitzinger
medical era (Paranjape et al. 2019). Medical schools, residency programs and continuing medical education will require continuous re-evaluation to ensure clinicians are equipped with adequate digital literacy and understanding to practice effectively and contribute to AI in healthcare quality.
CONCLUSION
Healthcare systems have grown to a level of complexity that requires inno- vative solutions to ensure quality assurance and improvement. AI technology presents a timely resource to interpret and act on data sets that would oth- erwise be insurmountable. Using these new technologies to protect patient safety, reduce burnout and address concerning patterns in healthcare deli- very is timely and necessary. A framework to prepare healthcare providers to utilize AI technology to enhance the calibre of patient care is not only cru- cial to advance medicine, but also to alleviate many of the pressing concerns currently faced by healthcare systems. Even with unprecedented innovations in healthcare, we must utilize tried and true methods of healthcare assurance and improvement including identifying vulnerabilities, mitigating biases, and ensuring health equity.
REFERENCES
Becker, G., Kempf, D. E., Xander, C. J., et al. (2010) Four minutes for a patient, twenty seconds for a relative - An observational study at a university hospital. BMC Health Services Research, 10(1): 1–9. Beil, M., Proft, I., van Heerden, D., Sviri, S., et al. (2019) Ethical considerations
about artificial intelligence for prognostication in intensive care. Intensive Care Medicine Experimental [online], 7(1): 70. Cath, C. (2018) Governing artificial intelligence: Ethical, legal and technical opportunities and challenges. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences [online], 376(2133): 20180080. Chang, H. Y., Jung, C. K., Woo, J. I., et al. (2019) Artificial intelligence in pathology.
Journal of Pathology and Translational Medicine [online], 53(1): 1–12. Gottschalk, A. and Flocke, S. A. (2005) Time spent in face-to-face patient care and work outside the examination room. Annals of Family Medicine, 3(6): 488–493. Irving, G., Neves, A. L., Dambha-Miller, H., Oishi, A., et al. (2017) International
variations in primary care physician consultation time: A systematic review of 67 countries. BMJ Open, 7(10): e017902. Kalra, J. (Jay), Rafid-Hamed, Z. and Seitzinger, P. (2021) Artificial Intelligence and Humanistic Medicine: A Symbiosis. In: Lecture Notes in Networks and Systems. Springer, Cham, 263: 3–8. Kalra, J. (Jay), Rafid-Hamed, Z. and Seitzinger, P. (2020) Disclosure of Medical Error:
A Necessary Step in Healthcare Improvement. Advances in Intelligent Systems and Computing [online], 1205: 11–16. Kalra, J., Rafid-Hamed, Z. and Seitzinger, P. (2019) The Future of Diagnostic Laboratory Testing in Healthcare. American Journal of Biomedical Science & Research, 5(2): 89–91. Loder, J. and Nicholas, L. (2018) Confronting Dr Robot Creating a people- powered
future for AI in health [online]. Nestla Health Lab, 1: 1–38.
Expanding Our Grasp: Artificial Intelligence as the Next Leap Forward in Healthcare Quality 11
Marvel, M. K., Epstein, R. M., Flowers, K., et al. (1999) Soliciting the patient’s agenda: Have we improved? Journal of the American Medical Association [online], 281(3): 283–287. Paranjape, K., Schinkel, M., Nannan Panday, R., et al. (2019) Introducing Artifi-
cial Intelligence Training in Medical Education. JMIR medical education [online], 5(2): e16048. Parks, J. A. (2010) Lifting the Burden of Women’s Care Work: Should Robots Replace the “Human Touch”? Hypatia [online], 25(1): 100–120. Rodrigues, H., Cobucciid, R., Nio Oliveira, A., et al. (2018) Burnout syndrome
among medical residents: A systematic review and meta-analysis. PLOS ONE 13(11): 0206840. Rotenstein, L. S., Torre, M., Ramos, M. A., et al. (2018) Prevalence of burnout among physicians a systematic review. JAMA - Journal of the American Medical Association, 320(11): 1131–1150. Schinkel, M., Paranjape, K., Nannan Panday, R. S., et al. (2019) Clinical applications
of artificial intelligence in sepsis: A narrative review. Computers in Biology and Medicine, 115: 103488. Seitzinger, P. and Kalra, J. (2022) Drawing Connections: Artificial Intelligence to Address Complex Health Challenges. Healthcare and Medical Devices [online], 51: 151–156. Seitzinger, P., Rafid-Hamed, Z. and Kalra, J. (2021a) Global Trends in Clinical Pra-
ctice and Healthcare Delivery: Opportunities for Growth and Innovation. In: Lecture Notes in Networks and Systems. Springer, Cham, 263: 9–15. Seitzinger, P., Rafid-Hamed, Z. and Kalra, J. (Jay). (2021b). Healthcare Delivery: Leveraging Artificial Intelligence to Strengthen Healthcare Quality. In: Lecture Notes in Networks and Systems. Springer, Cham, 263: 16–21. Wachter, S., Mittelstadt, B. and Floridi, L. (2017) Why a Right to Explanation
of Automated Decision-Making Does Not Exist in the General Data Protection Regulation. International Data Privacy Law, 7(2): 76–99. Weber, J. M. K. (2006) Core Competencies for Health Care Professionals. Journal of Allied Health [online], 35(3): 85.