Charting the future of patient care: A strategic leadership guide to harnessing the potential of artificial intelligence

Abstract

Artificial Intelligence (AI) applications have the potential to revolutionize conventional healthcare practices, creating a more efficient and patient-centred approach with improved outcomes. This guide discuses eighteen AI-based applications in clinical decision-making, precision medicine, operational efficiency, and predictive analytics, including a real-world example of AI’s role in public health during the early stages of the COVID-19 pandemic. Additionally, we address ethical questions, transparency, data privacy, bias, consent, accountability, and liability, and the strategic measures that must be taken to align AI with ethical principles, legal frameworks, legacy information technology systems, and employee skills and knowledge. We emphasize the importance of informed and strategic approaches to harness AI’s potential and manage its challenges. Moreover, this guide underscores the importance of evaluating and integrating new skills and competencies to navigate and use AI-based technologies in healthcare management, such as technological literacy, long-term strategic vision, change management skills, ethical decision-making, and alignment with patient needs.

Introduction

By harnessing the ability to replicate human cognitive functions, Artificial Intelligence (AI) is reshaping industries and opening new horizons. A pivotal moment in this technological revolution occurred at the Dartmouth Conference of 1956¹ where visionaries John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon together coined the term “Artificial Intelligence.” AI research was initially characterized by high expectations and optimism, but that enthusiasm eventually waned, leading to the “AI winter” from the 1980s to the 2000s. Slower-than-anticipated progress and reduced funding characterized this period.

There was, however, a “summer of artificial intelligence” that emerged in the 2020s with an unprecedented increase in publications, patent applications, and funding during this period. Google CEO Sundar Pichai² predicted AI would be the most important technology of the future. The launch of OpenAI’s AI chatbot, ChatGPT in 2022 confirmed this prediction. Notably, as of November 2023, ChatGPT boasts over 100 million weekly active users.³

Methods and findings

The rationale was to provide a descriptive guide for health managers, practitioners, and patients new to this technology. Methodology involved selecting only recently published research and policy literature in order to provide an up-to-date description of this fast moving research area. The opinions, assertions and conclusions expressed in this guide are based on these selected studies.

AI-driven applications in healthcare

A key element of AI integration in healthcare is advanced computational algorithms and machine learning aimed at improving various aspects of the healthcare ecosystem. This convergence of technology and medicine extends beyond conventional approaches and represents a paradigm shift in the clinical decision-making process, operational optimization, assessment of professional competence, and the overall quality of patient care.

A recent Clinician of the Future report commissioned by Elsevier⁴ collected the views of over 2,600 clinicians from 116 countries around the world and analyzed perspectives on artificial intelligence technology as a facilitator of streamlined patient care, forecasting that 73% of physicians will require proficiency in digital health by the year 2025.

The following section discuses AI-based clinical decision-making, precision medicine, operational efficiency, and predictive analytics with a focus on patient care.

Clinical decision-making

By assimilating and interpreting diverse data sources, AI improves diagnostic accuracy and treatment planning, revealing complex patterns and correlations within medical data sets, thereby enabling a more personalized approach to patient care.

A brief description of five major AI-based clinical decision-making applications is listed below.

1. Diagnostic imaging: AI algorithms analyze medical images with high precision, aiding in the early detection of abnormalities across diagnostic imaging modalities such as radiology and pathology.

2. Aggregating patient data: AI systems are designed to gather and analyze diverse patient data, encompassing medical history, diagnostic test results, and genomic information. As a result of this comprehensive approach, AI can provide a holistic overview, providing valuable support for clinical decision-making.

3. Integration with Electronic Health Records (EHRs): AI can integrate with EHR systems, providing healthcare providers with rapid and comprehensive access to a patient’s medical history. This integration allows for more efficient and streamlined access to relevant patient information, facilitating better-informed decision-making and improved continuity of care.

4. Clinical documentation: Natural Language Processing (NLP) converts unstructured clinical notes into structured data, making it easier for healthcare providers to extract relevant information for effective decision-making.

5. Voice-to-text transcription: NLP technologies can transcribe spoken conversations, enabling healthcare professionals to capture and record information from patient consultations, interviews, or other interactions in a textual format.

Precision (personalized) medicine

AI has the potential to contribute significantly to more individually tailored and effective healthcare interventions, from early detection to personalized treatment based on individual patient profiles. The combination of AI-based remote patient monitoring, real-time data analysis, and predictive modelling enhances the continuum of care, promoting better health outcomes.

Five major AI-driven applications in precision medicine are described below.

1. Patient data analysis: AI demonstrates an unparalleled ability to analyze large and diverse patient data sets including medical histories, genetic information, and lifestyle factors. In this way, unique patient characteristics can be rapidly identified, and effective tailored treatment plans developed.

2. Pathology and biomarker analysis: By combining AI analysis with conventional pathology methods, it becomes possible to reveal new molecular profiles and patterns associated with diseases. This integration can contribute to more accurate and nuanced diagnostic insights, potentially leading to improved disease understanding, early detection, and personalized treatment strategies in healthcare.

3. Longitudinal profiling: AI reveals unique patient characteristics by integrating the longitudinal profile of the illness including symptom progression with information from genetic and other biological data sets. This results in early diagnosis and effective individually tailored therapy.

4. Adaptive medication management: AI enables specificity and speed in optimizing drug treatment regimens based on patient responsivity in real-time. This includes adjusting dosage, switching medication, and/or early identification of potential adverse reactions based on individual patient characteristics.

5. Patient remote monitoring: Through continuous monitoring of physiological parameters, such as by wearables or remote monitoring tools, AI systems can detect subtle changes or patterns thereby enabling proactive therapeutic interventions.

Operational efficiency

Automation of routine tasks represents a strategic boon for healthcare, offering a pathway to efficient resource allocation and optimization of organizational performance. This is accomplished using Robotic Process Automation (RPA) which automates repetitive, rule-based tasks without requiring real-time human input. By streamlining routine processes, healthcare organizations can enhance productivity, reduce operational costs, and focus human resources on more complex and value-added tasks,

An overview of five AI-based operational efficiency applications is provided below.

1. Hospital resource management: By predicting patient admission rates AI optimizes hospital resource allocation, staff management and the effective deployment of clinical facilities in real-time.

2. Cost prediction: AI-driven predictive analytics can integrate trends in utilization, patient demographics, and reimbursement patterns to facilitate optimal just-in-time allocation of healthcare budgets and resources.

3. Supply chain management: Healthcare inventories can be effectively managed in real-time by applying AI-driven analytics to predict patient demand for medications and medical supplies.

4. Appointment scheduling: Personalized and optimized scheduling can be improved with AI-driven appointment scheduling that considers individual patient preferences, medical histories, and urgency of care.

5. Reducing hospital readmissions: AI-driven analytics proactively addresses the risk factors associated with hospital readmission by enabling early intervention, monitoring follow-up care, and engaging patients through personalized education and advice, thus preventing avoidable return visits.

Predictive analytics

In predictive analytics, artificial intelligence-driven applications enable just-in-time resource allocation based on industry demands. This allows healthcare systems to predict, plan, and act on emerging health trends using data-driven insights. The integration of AI in predictive analytics enhances the ability to proactively allocate resources, respond to changing demands, and improve overall healthcare system efficiency.

Listed below are three major applications of AI for predictive analytics.

1. High-risk population identification: By integrating historical and real-time health data sets, AI-driven predictive analytics quickly identifies those patients/populations at higher risk.

2. Emergency preparedness: Predictive analytics driven by artificial intelligence enhance emergency preparedness by allocating resources efficiently and promptly during periods of unexpectedly high emergency admissions.

3. Monitoring of epidemiological trends: Assimilation and interpretation of diverse data sources, including the ability to reveal complex patterns and correlations within very large public health data sets, enables AI-driven predictive analytics to speed up and improve decision-making, for example, during the COVID-19 outbreak (Table 1).

Table 1.

Summary table showing an evidence-based, real-world example of AI-driven database search which sourced four early findings on COVID-19. These findings were sourced over a thirteen week period in the early phase of the pandemic and were rapidly integrated into the worldwide public health management of the disease.

Month/day/year	Study title	Major finding
03/20/2020	Virus shedding patterns in nasopharyngeal and faecal specimens in COVID-19 patients^a	Virus excreted in the stool
03/28/2020	Prolonged viral shedding in faeces of paediatric patients with coronavirus disease 2019^b	Virus also found in children
05/11/2020	Association insulin resistance marker TyG index with the severity and mortality of COVID-19^c	Diabetes is risk factor leading to increased symptom severity
05/23/2020	Severe obesity as an independent risk factor for COVID-19 mortality in hospitalized patients younger than 50^d	Obesity and high blood are risk factors leading to increased symptom severity

^aZhang N, Gong Y, Meng F et al. Virus shedding patterns in nasopharyngeal and faecal specimens of COVID-19 patients. MedRxiv. March 30, 2020. doi: 10.1101/2020.03.28.20043059.

^bXing YH, Ni W, Wu Q et al. Prolonged viral shedding in faeces of paediatric patients with coronavirus disease 2019. Journal of Microbiology, Immunology and Infection. 2020 53 (3):473-480. doi: 10.1016/j.jmii.2020.03.021.

^cRen H., Yang Y., Wang F. et al. (2020) Association of the insulin resistance marker TyG index with the severity and mortality of COVID-19. Cardiovascular Diabetology.19, 58. doi: 10.1186/s12933-020-01035-2.

^dKlang E, Kassim G, Soffer S et al. Severe obesity as an independent risk factor for COVID-19 mortality in hospitalized patients younger than 50. Obesity (Silver Spring). 2020 28 (9):1595-1599. doi: 10.1002/oby.22913.

The challenge of implementing AI in healthcare

While AI-driven applications in healthcare show promise, they also present challenges. As AI becomes increasingly important in decision-making, issues of accountability, transparency, and ethics become paramount. In this overview, we discuss challenges such as transparency, data privacy, bias, consent and autonomy, accountability, liability, and legacy IT capabilities.

Ethical considerations

Kluge⁵ emphasizes the fiduciary nature of physician-patient relationships and the need for AI systems to adhere to ethical standards. Indeed, the responsibilities go beyond medical proficiency to encompass ethical, legal, and substitute decision-making considerations.

Transparency

Since AI systems are powered by algorithms and machine learning, they often function as black boxes, making their decisions difficult to understand.⁶ In light of this perceived lack of transparency in AI operations, questions have been raised about the rationale and communication of AI-driven healthcare decisions in the future. For both generating patient trust as well as verifying compliance with regulatory requirements and professional competence, it is essential to address this concern.

Data privacy

AI systems operate by leveraging extensive data sets, which may comprise sensitive patient information. The efficacy and reliability of these systems are contingent on the quality and quantity of the data sets analyzed. With such extensive use of health information comes a host of important concerns, which require well-established trust between patients, healthcare providers, and AI technologies, as well as advanced encryption technologies like fingerprint and retinal scanning.

Data bias

The ethical and effective integration of AI systems depends on fairness and equity. Ensuring diversity and representation within training datasets is critical for mitigating bias in the development of predictive algorithms; however, caution should be exercised when incorporating sensitive attributes like ethnicity, gender, or race to avoid reinforcing stereotypes, while AI models must be continuously evaluated to maintain optimal performance and reduce inequality.

Autonomy and consent

Instead of replacing human agency, ethical leadership should underscore AI’s collaboration with and augmentation of existing healthcare expertise. Equally important is the consideration of how AI recommendations influence patients’ decision-making processes. Furnishing patients with informed consent about how AI will be utilized in their care and enabling them to actively participate in decisions is a fundamental element of ethical healthcare.

Accountability and liability

In contrast to traditional decision-making processes, AI algorithms can create ambiguity about ultimate responsibility. The perceived loss of agency among current healthcare professionals poses an obstacle to the seamless integration of AI. Clearly defining responsibilities, ensuring accountability, and resolving uncertainties related to the roles of humans and AI in decision-making processes are crucial to addressing this challenge.

Lack of evidence

The widespread adoption of AI in healthcare is impeded by the absence of robust empirical evidence. Practitioners, regulators, and patients rightly seek real-life instances of successful and secure AI-driven solutions before embracing AI technologies. Bridging this evidentiary gap is imperative for cultivating trust and alleviating scepticism regarding the effective integration of AI into real-world healthcare settings (Table 1).

Legacy IT systems

Integrating AI into existing IT systems is challenging due to the often-insufficient capacity of current frameworks to handle AI applications’ data processing and analysis demands. The challenge is further compounded by interoperability issues between different computer systems or software, hindering efficient data exchange across healthcare silos. Adapting AI seamlessly requires healthcare organizations to upgrade or replace legacy systems, a costly process. Additionally, transitioning to an AI-friendly IT infrastructure requires strict adherence to or adaptation of security and privacy standards to maintain data integrity and protect patient information.

Staff capacity and expertise

There is currently a critical skills gap in the availability of professionals adept at developing, implementing, and maintaining AI-driven solutions, particularly in the healthcare sector. Additionally, successfully integrating AI in healthcare will require a workforce that understands how to integrate the technical intricacies of AI with the professional competencies of health professionals.

A strategy for effective integration

Implementing AI-driven applications in healthcare necessitates meticulous planning and a deep understanding of the associated challenges. An effective integration strategy should be grounded in knowledge and understanding. The following provides a concise overview of key considerations within the domains of collaboration and stakeholder engagement, education and training, regulatory compliance, data privacy and security, trust, and transparency, testing and validation, and incremental implementation.

Stakeholder collaboration and engagement

Optimizing AI integration in healthcare hinges on collaboration. Engaging key stakeholders, such as healthcare professionals, administrators, patients, data scientists, AI experts, IT specialists, and regulatory authorities, in the development and implementation process is essential. Simultaneously, academic and research institutions should establish partnerships to conduct independent evaluations and trials, gathering robust evidence on the efficacy of AI in real healthcare settings. This collaborative approach fosters a comprehensive understanding of the technology’s impact and ensures a well-rounded implementation.

Education and training

Investing in comprehensive education and training is essential to preparing a workforce that not only grasps the technical aspects of AI but also understands its ethical implications. This approach is instrumental in addressing the existing skills gap and laying the foundation for building a resilient healthcare workforce capable of fully leveraging the potential of AI. Establishing partnerships between academia and research institutions is necessary to develop new courses, modules, and degrees specifically designed to integrate AI into healthcare education. This collaborative effort ensures that professionals are equipped with the knowledge and ethical considerations necessary for the responsible application of AI in the healthcare domain.

Regulatory compliance

Given the inherently sensitive nature of patient data, healthcare organizations utilizing AI must prioritize compliance with privacy laws. In Canada, the Personal Information Protection and Electronic Documents Act (PIPEDA)⁷ serves as a federal privacy law governing the collection, use, and disclosure of personal information by private sector organizations engaged in commercial activities.

While PIPEDA applies across industries, including healthcare, Canada’s health sector is additionally subject to provincial and territorial regulations. Thus, healthcare organizations operating in Canada must comply with both federal and provincial/territorial legislation regarding the handling of health information. In addition to data protection, healthcare AI applications are often regulated by medical device and software regulatory bodies. Compliance with regulations from agencies like Health Canada⁸ is crucial to ensuring the safety, effectiveness, and reliability of AI solutions.

Data privacy and security

Due to the sensitive nature of healthcare data, robust security and privacy measures must be prioritized within the sector. Healthcare organizations are obligated to conduct regular audits and compliance checks to ensure adherence to data protection regulations. In the era of AI integration, these measures are essential to protect patient information, maintain compliance with privacy regulations, and ensure data integrity and confidentiality.

Transparency and trust

The bedrock of transparency and trust within healthcare systems rests on safeguarding patient privacy and confidentiality. Beyond regulatory compliance, this commitment encompasses transparent practices, robust security measures, and advanced anonymization techniques. To facilitate informed consent, AI systems must ensure that healthcare professionals and patients understand all factors involved. Using this comprehensive approach fosters trust, transparency, and patient confidentiality in the ethical use of AI in healthcare.

Testing and validation

Ensuring the effectiveness and adaptability of AI solutions across a range of real-world applications necessitates rigorous testing and validation of AI algorithms. This process should encompass diverse patient demographics, medical conditions, and healthcare settings to ensure comprehensive evaluation. To gain acceptance, AI solutions need to undergo iterative improvement processes guided by this evidence. Collaboration between professional, academic, and research institutions is imperative to ensure robust testing, validation, and continuous enhancement of AI algorithms in diverse healthcare scenarios.

Incremental implementation

Healthcare professionals and other users leveraging AI need to be adequately trained before implementing AI solutions in healthcare settings. Adopting an incremental approach to implementation minimizes disruptions, ensuring a smoother integration process. Ongoing training initiatives and continuous refinement can then be implemented based on valuable user feedback. This iterative approach not only enhances the adaptability of AI solutions but also allows for the incorporation of real-world insights, ultimately contributing to more effective and user-friendly applications in healthcare.

Essential leadership skills and competencies for navigating the AI landscape

AI will transform the healthcare industry, requiring leaders to hone new skills and competencies. The following is a brief description of the skills needed in technological literacy, long-term strategic vision, change management, and ethical decision-making.

Technological literacy

Leaders in AI-driven healthcare should be technologically savvy, and understand algorithms, deep learning models, and natural language processing. This expertise should extend to comprehending the capabilities and limitations of AI in healthcare, foreseeing potential risks, and actively cultivating an innovation-oriented culture within the healthcare organization.

Long-term strategic vision

Kassam and Kassam⁹ argue that leaders in AI-driven healthcare must align technological advancements with organizational goals while also considering sustainability. Technology must be flexible to meet the system’s long-term needs, which requires consistent financial investment and AI infrastructure that is evergreen. Healthcare leaders must anticipate potential risks and create robust contingency plans, enabling effective risk management throughout AI adoption.

Change management

The adoption of AI will require healthcare leaders to demonstrate exceptional change management skills. Nurturing an environment that fosters continuous learning, innovation, and adaptability to lead an organization through technological transitions requires unwavering confidence, adaptability, and the ability to create a clear vision for the future. By creating a compelling vision, leaders will ensure staff understand the rationale behind AI adoption, address concerns, and communicate openly. Real-world examples and carefully selected case studies of AI effectiveness also reinforce the tangible benefits of this vision by showcasing early successes (Table 1). Additionally, celebrating milestones reinforces a shared vision, motivates staff, and acknowledges progress.

Ethical decision-making

The ability to make ethical decisions in an AI-driven healthcare era is a fundamental leadership competency. It has been highlighted above that healthcare leaders must navigate complex ethical considerations when implementing AI systems, including safeguarding patient privacy. Moreover, leaders must ensure that AI algorithms do not introduce bias into healthcare practices, and they should work actively to promote fairness and inclusivity when applying AI.

A patient-centred focus

A commitment to aligning care with patient needs and values is essential for healthcare leaders to effectively guide AI integration. Through a patient-centred approach, leaders can ensure that AI augments rather than replaces human agency in therapeutic alliances. In this regard, AI-driven applications should also consider patient preferences, including cultural considerations.

Conclusion

The imminent adoption of AI in healthcare offers unprecedented opportunities for enhancing decision-making, streamlining operational efficiency, and ultimately improving patient outcomes. Indeed it is clear that there are vast areas in healthcare where AI will be important for better decision-making and enhancing patient outcomes. Simultaneously, health leaders will encounter medico-legal challenges related to accountability, transparency, and ethics. Successfully navigating and integrating these AI-based technologies into healthcare management strategies necessitates a thoughtful consideration of a broad set of skills and competencies.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Institutional Review Board approval was not required.

ORCID iD

William T. O’Connor

References

McCarthy

Minsky

Rochester

Shannon

. A proposal for the dartmouth summer research project on artificial intelligence, August 31, 1955. AI Mag. 2006;27(4):12. doi:10.1609/aimag.v27i4.1904

World Economic Forum . Artificial Intelligence Could Be Our Savior, According to the CEO of Google. Retrieved from: www.weforum.org/agenda/2018/01/google-ceo-ai-will-be-bigger-than-electricity-or-fire. Accessed December 30, 2023.

ChatGPT Statistics . Detailed insights on users. 2024. Retrieved from: www.demandsage.com. Accessed December 30, 2023.

Elsevier . Clinician of the future report 2023. Retrieved from: www.elsevier.com/resources/clinician-of-the-future-2023. Accessed December 30, 2023.

Kluge

. Artificial intelligence in healthcare: ethical considerations. Healthc Manag Forum. 2019;33:47-49. doi:10.1177/0840470419850438

Jassar

Adams

Zarzeczny

Burbridge

. The future of artificial intelligence in medicine: medical-legal considerations for health leaders. Healthc Manag Forum. 2022;35(3):185-189. doi:10.1177/08404704221082069

Office of the Privacy Commissioner of Canada . PIPEDA Legislation. Retrieved from: www.priv.gc.ca/en/privacy-topics/privacy-laws-in-canada/the-personal-information-protection-and-electronic-documents-act-pipeda/. Accessed December 30, 2023.

Health Canada . Retrieved from https://www.canada.ca/en/health-canada.html. Accessed December 30, 2023.

Kassam

. Artificial intelligence in healthcare: a Canadian context. Healthc Manag Forum. 2019;33:5-9. doi:10.1177/0840470419874356