Files and links (1)
PDFOpen Access (License Unspecified), Open Access
Dissertation
Generative AI in healthcare: strategic, ethical, and operational readiness for clinical integration
Doctor of Philosophy (Ph.D.), Drexel University
Mar 2025
DOI:
https://doi.org/10.17918/00010929
Abstract
Generative artificial intelligence (AI) is poised to transform clinical care by addressing key challenges such as staffing shortages, provider burnout, population health management, and patient care capacity. However, the application of generative AI in healthcare remains in its early stages, with most organizations lacking structured processes for model implementation and governance. Unlike other industries, healthcare must operate with minimal errors and biases due to the direct impact of AI-driven decisions on patient outcomes. Deploying generative AI in clinical settings requires careful consideration of fairness, security, privacy, workflow integration, and value to patient care and outcomes. Additionally, unanticipated risks and unintended consequences necessitate a robust governance and monitoring framework. This dissertation examines the essential factors for generative AI integration in healthcare, focusing on data strategy, implementation processes, accountability mechanisms, and governance structures. Key concerns are ensuring fairness, minimizing bias, and aligning AI-driven insights with clinical workflows. Given the high stakes of AI-assisted decision-making, this research highlights the importance of transparency, continuous monitoring, and adherence to ethical and regulatory standards throughout the AI lifecycle. A multidisciplinary team approach is proposed to AI application, incorporating stakeholder engagement, performance evaluation, and best practices for responsible AI adoption. A case study on the deployment of ChatGPT for patient messaging at Penn Medicine will be utilized to illustrate the opportunities and challenges of large language models (LLMs) in clinical applications. Findings from this research contribute to AI implementation strategies by providing a structured framework for organizational readiness, governance principles, and key performance metrics. The dissertation underscores the need for patient-centered AI development and outlines strategies to mitigate risks related to data integrity, model transparency, and system integration. Future research will explore expanding AI data sources, enhancing interoperability, addressing scalability challenges, and incorporating a multidisciplinary approach and patient engagement framework into AI governance models. By fostering collaboration and rigorous oversight, healthcare organizations can harness AI's transformative potential while ensuring value, safety, trust, and accountability.
Metrics
90 File views/ downloads
108 Record Views
Details
- Title
- Generative AI in healthcare
- Creators
- Jon K. Pomeroy
- Contributors
- Christopher C. Yang (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Philosophy (Ph.D.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- 86 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- Information Science (Informatics) [Historical]; College of Computing and Informatics (2013-2026); Drexel University
- Other Identifier
- 991022052739904721