Seeing Anxiety: Ecological Video Ethnography and Simulation to Understand Anxiety and Decision-Making in Dental Care

Dental anxiety is a widespread barrier to care, often contributing to appointment avoidance, treatment disruption, and emotional strain for both patients and clinicians. While extensively studied as a psychological trait, less attention has been paid to how anxiety manifests situationally and relationally, through non-verbal, affective, and spatial cues within clinical encounters. Moreover, in certain contexts, such affective signals can intersect with decision-making processes, shaping how patients interpret, engage with, or withdraw from clinical choices.

To explore how anxiety manifests within the micro-choreography of dental encounters, develop a taxonomy of 'anxious tells', and assess the utility of ecological video ethnography and simulation-based camera testing in supporting affect-aware training and potential AI integration.

We conducted a two-phase qualitative study. Phase one involved re-analysis of over 70 h of ethnographic video from dental clinics, guided by threshold theory, affect theory and ecological-relational analysis. Phase two consisted of simulation-based testing of multi-angle camera setups in a custom-built dental lab designed to optimise audiovisual capture of non-verbal cues.

Anxiety was expressed through posture shifts, facial micro-expressions, breath modulation, eye movement, and interactional hesitancy. These behaviours were often relational and environmentally cued, and in some cases intersected with moments of decision-making such as consent, treatment planning, or pain management. We developed a context-sensitive taxonomy of 'anxious tells', identified optimal camera protocols, and revealed interpretive blind spots through interdisciplinary co-viewing. Key challenges included managing large AV data, synchronisation, and ethical clarity in consent.

This study demonstrates how video can function not only as an observational device but as a collaborative instrument for clinical reflection, training, and research co-production. It is useful to conceptualise anxiety in dentistry not as a fixed trait, but as a relational and affective phenomenon. The newly developed taxonomy of 'anxious tells' in dental practice supports the identification and effective clinical care of patients suffering from such anxiety. Ecological video ethnography offers a powerful lens for revealing these dynamics and supporting situated, affect-based training. Our framework and toolkit lay the groundwork for future human-in-the-loop AI systems that recognise, contextualise, and respond to affective cues supporting safer, more empathetic care.

As part of an EPSRC IAA grant, internal strategic investments, and an MPS grant, we conducted a series of patient and public engagement activities that shaped both the research question and the study design. We also ran sessions with dentists. The PPIE process directly informed the consent form, as participants highlighted that some individuals might want their data to be made available while others may not. A key divide emerged: while some were comfortable with filming, others were strongly opposed. Contrary to our expectations, most dentists were supportive of the idea. During the sessions, we also learned that some dentists routinely record patient interactions for potential legal purposes, which highlighted an additional potential use case for our toolkit.