Across 2024/2025, I led implementation of a large-scale educational innovation using SimConverse, an AI-driven virtual patient (VP) platform, to enhance teaching of advanced communication skills in the Medicine program and addressing a persistent challenge in medical education: the provision of consistent, high-quality experiential learning in communication across large and geographically-distributed student cohorts. Collaborating with colleagues in Nexus, Educational Development and Teaching and Rural Support teams, as well as Phase and Term Convenors (as critical subject-matter experts), and Campus Coordinators we created opportunities for repeated practice and feedback on these important skills.
Curricular Design / Pedagogical Alignment: Learning design was grounded in experiential and mastery-learning theory. Activities follow a structured workflow: (1) pre-briefing/orientation to ensure psychological safety; (2) first interaction with an AI-driven VP; (3) automated, personalised feedback generated from analysis of the transcript; (4) guided student reflection; and (5) a second interaction with the same VP to enable deliberate practice and consolidation. This design operationalises the Kolbian learning cycle and embeds reflection as an explicit phase of skill development, aligning with Phase 1 skills development.
Scenarios mapped to six advanced Communication Skills Learning Outcomes (CSLOs) from the Phase 2 Clinical Skills Guide: communicating with anxious/depressed patients; taking a sexual history; conducting motivational interviewing; cross-cultural communication; taking a social history; and discussing care in late-stage cancer. Each scenario was co-designed with clinical academics to ensure authenticity of content, accuracy of medical context, and alignment with Course and Phase Learning Outcomes.
Scalability and Educational Equity: The design prioritised scalability and equity of access. SimConverse integrates directly with Moodle via single-sign-on, allowing seamless student access across all campuses. Each student completes three AI-VP sessions for each CSLO, enabling repeated, standardised practice irrespective of clinical placement variability. This structure ensures every student receives equivalent experiential learning opportunities - previously unachievable within existing resourcing.
Assessment Integration: A key design goal was to strengthen the connection between learning activities and assessment. Scenarios were mapped to communication criteria in the Phase 2 Integrated Clinical Examination (ICE). This enabled, for the first time, the creation of dedicated communication stations within the ICE which, following that assessment, will strengthen evaluation of the platform outcomes.
Evaluation embedded in the design process: Quantitative data are being collected on student engagement, and correlations between examiner ratings in the ICE this year as compared with prior cohorts. Complementary qualitative data are gathered through integrated student surveys and feedback from course convenors. Preliminary findings indicate high levels of engagement and perceived value as a safe space for practice, plus valuable feedback for development of skills.
Educational Significance: This project demonstrates how emerging generative-AI technologies can be harnessed to create authentic, personalised and scalable learning opportunities that preserve and extend on my guiding pedagogical principles of experiential and reflective practice, building on earlier teaching. By integrating SimConverse in Year 3, we provided 300 students with 18 feedback-rich opportunities to practise complex communication in a psychologically safe, reproducible format, enhancing learning outcomes and contributing to their readiness for assessment in communication skills in high-stakes clinical examinations.