Year: 2026
Type: dissertation
Abstract: Clinical documentation is a central yet increasingly burdening part of physicians’ work. Electronic health records (EHRs) have expanded documentation requirements beyond clinical reasoning to include legal, administrative, and organisational purposes, which has contributed to information overload, fragmented workflows, and reduced time available for patient care. In response, artificial intelligence (AI)-based documentation tools are being promoted as a way to reduce the documentation burden and support clinical work. However, there is still limited qualitative understanding of how physicians themselves experience current documentation practices, what they expect from AI-assisted documentation, and under what conditions such tools are perceived useful, trustworthy, and professionally acceptable. This thesis examines Finnish physicians’ experiences of their current documentation workflows and their anticipatory evaluations of AI-assisted documentation. The study employs a qualitative, user-centered, and socio-technical approach that draws on semi-structured interviews conducted with seven physicians working in different specialised care contexts. The results show that physicians experience documentation as a multi-stage workflow, where the primary burden is not on writing the encounter notes but on navigating, reviewing, and synthesising fragmented and redundant patient information. AI-assisted documentation is widely envisioned as a supporting background technology rather than a workflow driver. The strongest expectations focused on intelligent information synthesis, context-aware search, and automation of repetitive action-oriented tasks. Anticipated risks centre on verification effort, loss of narrative control, and professional responsibility. Trust is found to be conditional and task-specific. It is shaped by perceived reliability, workflow fit, and respect for professional autonomy rather than abstract performance claims. By situating anticipated user experience (UX) within physicians’ lived practices, this thesis contributes empirical insight into how AI-assisted documentation is evaluated at an early phase of adoption in the Finnish healthcare context and cautions against solutions that may replicate or reinforce existing challenges.
Source: Aaltodoc (Aalto University)
Author: Klemetti, Louna
Institution: Aalto University
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Keywords: AI-assisted documentation, clinical documentation, electronic health records, physician experience, documentation burden, trust in AI, professional autonomy, clinical workflow, information synthesis, narrative control, professional responsibility, user experience, AI adoption in healthcare
Open Access status: green
OpenAlex ID: W7139304156
