AI-Driven Clinical Surveillancein Practice : A Mixed-Methods Case Study on Staff Readiness andEpidemiological Monitoring at FMP Hanoi

Sammanfattning

The integration of artificial intelligence (AI) in healthcare presents significant opportunities to improve operational efficiency and clinical decision-making. This thesis examines the feasibility and practical implementation of AI applications within Family Medical Practice (FMP) in Hanoi, a multilingual private healthcare provider. A mixed-methods approach was used, combining a quantitative staff survey based on the Technology Acceptance Model (TAM3) with qualitative stakeholder insights. The findings reveala strong preference for augmented intelligence solutions that support, rather than replace, clinical staff, as well as a substantial trust gap toward fully autonomous systems.Based on these results, an AI-driven prototype, the Clinic Surveillance Hub, was developed. The system applies statistical anomaly detection using Z-scores and a hybrid forecasting model to identify disease trends and support epidemiological monitoring. To improve usability, an AI-based interpretation layer translates complex analytical outputs into structured clinical reports. The results demonstrate that the system can detect significant deviations in disease patterns and provide actionable insights for clinical planning, even when limited to anonymized diagnosticdata. The study shows that technical performance alone is not enough—staff trust and system transparency are equally important for adoption. This work contributes to the development of scalable, low-risk AI solutions for healthcaresettings by combining data-driven analytics with user-centered system design.