AI-Enhanced Administrative system : Automating Diagnosis Coding- in Healthcare-At Linköping University Hospital (Östergötland Region)

Sammanfattning

Artificial Intelligence (AI) continues to reshape the healthcare landscape, offering improvements in clinical accuracy, treatment personalisation, and patient safety. However, the administrative side of healthcare remains heavily manual and time-consuming, with tasks such as diagnosis coding, medical record retrieval, billing, and scheduling contributing to increased workloads and reduced efficiency among healthcare professionals. The study employs a Design Science Research (DSR) methodology to design and develop an AI-Enhanced Administrative System tailored for Linköping University Hospital. The artefact focuses on automating diagnosis coding, aiming to alleviate the manual workload and improve administrative accuracy and speed. Through iterative development and stakeholder feedback, the research identifies key functional requirements for successful integration and highlights challenges in data quality, system interoperability, and user trust. Findings suggest that AI driven administrative tools can significantly streamline workflows, reduce human error, and allow healthcare professionals to focus more on patient care. However, successful implementation depends on contextual adaptation, continuous evaluation, and active collaboration between developers and end users. Future research should explore broader deployment strategies and address ethical considerations around data privacy and accountability in AI-supported administrative systems.