The Micro-Expression Decoder: A Preliminary Mixed-Methods Evaluation of Training in Subtle Facial and Non-Verbal Cue...

Background/Objectives: Subtle facial and interactional cues may prompt clinicians to explore possible anxiety, confusion or hesitation, but they do not establish a patient's internal state. The Micro-Expression Decoder (MED) links cue recognition with tentative interpretation, direct checking and communication adaptation. Methods: This preliminary, uncontrolled, single-group pre-post convergent mixed-methods evaluation involved 60 final-year medical students from three London medical schools. Ten groups of six completed a 270 min simulation-based programme comprising conceptual teaching, video observation, standardised-patient encounters, feedback and reflection. Study-specific self-ratings and a parallel-form recognition task were completed before teaching and immediately after the instructional and simulation components, before the final debrief. Standardised-patient ratings were obtained during the first and final simulated encounters. There was no control group or follow-up. Results: Recognition confidence increased from 2.7 to 4.3 (mean change 1.6, 95% CI 1.39-1.81; dz = 1.98), interpretation confidence from 3.0 to 4.5 (dz = 2.01), standardised-patient-rated attentiveness from 3.2 to 4.6 (dz = 2.02), and satisfaction from 3.5 to 4.8 (dz = 1.94). Anxiety recognition increased from 25/60 (41.7%) to 55/60 (91.7%), and confusion recognition from 29/60 (48.3%) to 52/60 (86.7%). Qualitative accounts described earlier noticing, clarification and checking, alongside difficulty distinguishing cues, distraction, cultural variation and over-interpretation. Nineteen students increased confidence without categorical improvement in anxiety or confusion recognition. Conclusions: MED was associated with immediate improvement across study-specific outcomes in simulation. The uncontrolled design, immediate assessment, investigator-developed measures, expectancy effects and ceiling in post-training satisfaction preclude conclusions about causality, retention, real-patient benefit, empathy or clinical competence.