Medical errors remain a critical issue in healthcare systems worldwide, contributing to millions of adverse events, fatalities, and significant economic burdens annually, as reported by the World Health Organization (WHO, 2018). In the context of Kermanshah University of Medical Sciences hospitals, the situation is exacerbated by systemic challenges, including a shortage of specialized healthcare professionals, excessive workloads, inadequate infrastructure, and a deficient culture of error reporting. These factors collectively elevate the likelihood of diagnostic, pharmaceutical, and surgical errors, compromising patient safety and healthcare efficiency. To address this, Artificial Intelligence (AI) emerges as a transformative solution due to its robust data analysis and decision-support capabilities. This study was designed to develop a strategic model for reducing medical errors through AI integration in Kermanshah hospitals, with a focus on identifying error patterns, leveraging AI’s potential, addressing implementation barriers, and proposing actionable strategies to enhance patient safety, reduce costs, and improve operational efficiency.
The research adopted a qualitative, applied approach grounded in Strauss and Corbin’s (1998) grounded theory methodology. The study population included physicians, pharmacists, and managerial staff from Kermanshah University of Medical Sciences in 2025. Using purposeful and snowball sampling, 15 participants—comprising 8 physicians, 4 pharmacists, and 3 managerial staff—were selected based on theoretical saturation. Data collection involved an extensive literature review and in-depth semi-structured interviews, each lasting 45 to 60 minutes. The data were analyzed using MAXQDA 11 software through a three-stage coding process: open, axial, and selective. To ensure validity, member checking and triangulation were employed, while reliability was verified through recoding, achieving a 77% agreement coefficient.
The analysis produced a paradigmatic model centered on the strategic reduction of medical errors through AI. Causal conditions driving errors were categorized into early-stage issues, such as inaccurate predictions, and late-stage issues, including human errors. Contextual conditions influencing AI adoption included managerial factors like standardized processes, cultural factors such as fostering a robust reporting culture, and human factors like comprehensive staff training. Intervening conditions that pose challenges to implementation were identified as inadequate supervision, weak executive structures, legal restrictions, and the inherent complexities of AI technologies. To overcome these, the model proposes strategies such as developing advanced technological infrastructure, providing targeted training for healthcare professionals, formulating supportive policies, and deploying AI-driven tools like decision support systems. The anticipated outcomes of these strategies include a significant reduction in medical errors, improved operational efficiency, enhanced quality of healthcare services, and greater patient satisfaction.
The proposed model serves as a comprehensive guide for hospital administrators to leverage AI in managing clinical risks effectively. By addressing structural and systemic challenges, this approach transforms barriers into opportunities for innovation, ultimately elevating the standard of care. To facilitate AI integration, policymakers are urged to invest in infrastructure development, staff training, and the establishment of clear legal frameworks. These steps will help align Iran’s healthcare system with global benchmarks, ensuring safer and more efficient care delivery. Furthermore, this study lays a robust foundation for future research exploring the intersection of AI and patient safety, encouraging further advancements in healthcare innovation.