
Contributing to our BMERG blog series on building community, our BMERG Journal Club lead Dr Claire Hudson reflects on the discussions at a recent journal club about AI-generated clinical narratives.

Artificial intelligence (AI)-generated image of Selena Gomez singing with Justin Bieber. Taken from Bland (2025) doi:10.2196/63865
The BMERG journal club recently met to discuss the following paper:
Bland, Tyler. “Enhancing Medical Student Engagement Through Cinematic Clinical Narratives: Multimodal Generative AI-Based Mixed Methods Study.” JMIR medical education vol. 11 e63865. 6 Jan. 2025, doi:10.2196/63865
This paper sparked my attention; it’s not often you see mention of Justin Bieber and Selena Gomez in an academic research paper! I was interested to find out whether their cinematic clinical narrative was a valid educational approach, or whether it was all just a gimmick…. I was sceptical.
Publication overview
This study, based in one US Medical School, explored a creative way to teach complex pharmacology to medical students using Generative AI (GenAI). Instead of a traditional case-based learning (CBL) scenario, the educators transformed a clinical case into what they called a cinematic clinical narrative, a multimedia story titled Shattered Slippers. It featured an AI-generated plot, narration, images and even a theme song, blending clinical content with an emotionally engaging storyline.
The approach was grounded in multiple educational theories, including Constructivist Learning Theory, Cinemeducation, Mayer’s Cognitive Theory of Multimedia Learning, and the Technological Pedagogical Content Knowledge (TPACK) framework. They hoped to improve student engagement and recall.
What did the researchers find out?
The majority of the students preferred this AI-enhanced case to traditional teaching methods. They scored it highly on a ‘Situational Interest Survey’ and performed well on related exam questions (although these weren’t compared with performance in other topics!). Students reported liking the narrative style and the pop culture references. The authors suggest that multimedia storytelling, supported by GenAI, could enhance engagement, comprehension and retention, and could even help mitigate early medical-school burnout driven by information overload.
Why this study resonated
This paper generated one of our liveliest journal discussions. It tapped into current debates about the role of AI, narratives and emotion in medical and health sciences education.
We recognised similarities to CBL materials already used in Bristol Medical School (BMS) teaching. The phrase cinematic clinical narrative suggests a Netflix-style experience, however, this was essentially still a patient case, albeit within a more elaborate fictional storyline. The story was presented using Powerpoint slides with AI-generated narration, included two AI-generated ‘cinematic-style’ images of ‘Selena’ and ‘Justin’, plus an AI-generated song. There were clinical questions posed to the students within the narrative, akin to the ‘pause points’ or facilitator questions used within CBL in BMS.
What did we think?
During our discussion, we used the phrase “style over substance”, as we didn’t think the images, song or AI-narration particularly enhanced the story. However, we also recognised that our average age was significantly above that of an undergraduate medical student, so perhaps we simply weren’t the target audience! We didn’t like the AI-generated voice-over, but given how quickly it was produced, we could see the advantage over recording narration ourselves. Within our own CBL sessions, we prefer students to read cases aloud, and believe it allows pause for questions, discussion and deeper engagement with the content.
The authenticity was a positive aspect of this cinematic clinical narrative. The case was based on lupus, a condition experienced by actor and singer Selena Gomez in real life, which helps humanise the topic and connect students emotionally with the material. We already value this approach; for example, our curriculum includes videos of real patient stories, which resonate strongly with students and enhance their empathy.
The methodology and reporting were also strong. For example, they included CHERRIES (Checklist for Reporting Results of Internet E-Surveys) to report reliability of the Situational Interest Survey; listed all AI tools used (ChatGPT-4, Leonardo.ai, Eleven Labs and Suno); and shared all prompts used and generated media. The authors report that AI generated the storyline and media quickly, while the original case took approximately one day to write. It is reassuring that AI did not generate the clinical content per se, suggesting existing case material could be ‘fleshed out’ using this approach.
What ideas did we have, and what can we take away?
The idea discussed mostly relate to medical CBL, but could be adapted to other contexts.
Use a range of media: We already embed patient videos, YouTube clips, and numerous clinical images within CBL resources; it is clear from this research and the wider literature that using a range of media in teaching helps keep students engaged.
Use storytelling: Adding stronger narrative arcs or character development could make CBL cases more distinctive and memorable, may help students connect emotionally with the material and visualise more diverse patient experiences.
AI-assisted voices: Using AI to generate patient voices in different accents or tones, helping students become familiar with diverse pronunciations and communication styles. Some parts of the case could be read by an AI-patient, perhaps communicating patient experience following treatment.
Students as producers: Students could create their own ‘cinematic clinical narratives’ which could be judged at an informal ‘Oscars’-style showcase. Expert review would ensure medical accuracy while giving students creative ownership.
Flipped-learning resources: Cinematic narratives could serve as pre-session materials, freeing live teaching time for deeper discussion.
Final Thoughts
This study encouraged us to think more creatively about multimedia teaching materials and prompted some vibrant discussion. Overall, we were supportive of our existing CBL approach, which places greater emphasis on group activities and student-led discussion than this ‘cinematic clinical narrative’ appeared to. However, we realise that delivering information to students using a variety of media is important for sustaining engagement and interest.
If you already use storytelling in your teaching, using GenAI or not, we’d love to hear from you!
Author Biography
Dr Claire Hudson is a Lecturer on the Teaching and Scholarship Pathway within the Bristol Medical School. Claire’s early research career was in biomedical sciences, with a recent transition to pedagogic research. She has a special interest in self-regulated learning and the use of reflective practice in developing academic and feedback literacy skills.



















