AI and inclusion in simulation education and leadership: a global cross-sectional evaluation of diversity

Joana Berger-Estilita ^1,2, Mia Gisselbaek ^3,4,5, Arnout Devos ⁶

¹Institute for Medical Education, University of Bern, Bern, Switzerland.
²RISE-Health, Centre for Health Technology and Services Research, Faculty of Medicine, University of Porto, Porto, Portugal.
³Division of Anesthesiology, Department of Anesthesiology, Pharmacology, Intensive Care and Emergency Medicine, Geneva University Hospitals and Faculty of Medicine, Geneva, Switzerland. Mia.gisselbaek@gmail.com.
⁴Unit of Development and Research in Medical Education (UDREM), Faculty of Medicine, University of Geneva, Geneva, Switzerland. Mia.gisselbaek@gmail.com.
⁵Division of Anesthesiology, Department of Acute Care Medicine, Geneva University hospitals, Geneva, Switzerland. Mia.gisselbaek@gmail.com.
⁶ETH AI Center, Swiss Federal Institute of Technology Zurich (ETH Zurich), Zurich, Switzerland.
⁷Department of Anaesthesia, Pain and Perioperative Medicine, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, Hong Kong.
⁸Centro di Simulazione (CeSi), Centro Professionale Sociosanitario Medico-Tecnico, Lugano, Switzerland.
⁹Department of Anesthesia and Intensive Care, , Ankara University Faculty of Medicine, Ankara, Turkey.
¹⁰Ankara University Brain Research Center (BAUM), Ankara, Turkey.
¹¹Department of Anesthesia and Perioperative Care, University of California San Francisco, San Francisco, CA, USA.
¹²Division of Anesthesiology, Department of Anesthesiology, Pharmacology, Intensive Care and Emergency Medicine, Geneva University Hospitals and Faculty of Medicine, Geneva, Switzerland.
¹³Unit of Development and Research in Medical Education (UDREM), Faculty of Medicine, University of Geneva, Geneva, Switzerland.
¹⁴Division of Anesthesiology, Department of Acute Care Medicine, Geneva University hospitals, Geneva, Switzerland.
¹⁵Anesthesiology Department, Hospitais da Universidade de Coimbra, Coimbra, Portugal.
¹⁶Faculty of Medicine, University of Coimbra, Coimbra, Portugal.
¹⁷Clinical Academic Center of Coimbra, Coimbra, Portugal.
¹⁸Copenhagen Academy for Medical Education and Simulation (CAMES), Herlev, Copenhagen, Capital Region of Denmark, Denmark.
¹⁹Department of Quality and Health Technology, University in Stavanger, Stavanger, Norway.
²⁰Department of Public Health, Copenhagen University, Copenhagen, Denmark.
²¹Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark.
²²Department of Anesthesiology, Helora, Mons, Belgium.
²³Department of Surgery, Research Institute for Health Sciences and Technology, University of Mons (UMONS), Mons, Belgium.

⁰Institute for Medical Education, University of Bern, Bern, Switzerland.

Advances in Simulation (london, England) +

|

May 4, 2025

View abstract on PubMed

Summary

AI-generated profiles for simulation-based medical education leadership reveal platform biases. ChatGPT and Gemini show more diversity, while Claude favors older, White, male profiles, potentially perpetuating inequities.

Area Of Science

Medical Education Technology
Artificial Intelligence in Healthcare
Health Professions Education

Background

Simulation-based medical education (SBME) is vital for healthcare training, influencing skills, professional identity, and inclusivity.
Leadership demographics in SBME significantly impact program design and learner outcomes.
Artificial intelligence (AI) platforms can generate demographic data but may perpetuate existing inequities due to inherent biases.

Purpose Of The Study

To evaluate the demographic profiles of simulation instructors and simulation lab heads generated by three AI platforms: ChatGPT, Gemini, and Claude.
To identify potential biases in AI-generated demographic data relevant to healthcare education leadership.
To compare the representation of age, gender, race/ethnicity, and medical specialties across different AI models.

Main Methods

A global cross-sectional study utilized standardized English prompts to query ChatGPT, Gemini, and Claude.
Demographic data (age, gender, race/ethnicity, medical specialty) were collected for 2014 simulation instructors and 1880 lab heads.
Statistical analyses included ANOVA and chi-square tests with Bonferroni corrections (P < 0.05).

Main Results

Significant demographic differences were found among AI platforms.
Claude generated older lab heads (mean 57 years) with a male predominance (63.5%) and predominantly White profiles (47.8%), favoring anesthesiology and surgery.
ChatGPT and Gemini exhibited smaller age gaps, balanced gender representation, greater racial diversity (up to 24.4% Black, 20.6% Hispanic/Latin), and broader interdisciplinary specialty representation.

Conclusions

AI-generated demographic profiles for SBME leadership highlight biases that could reinforce healthcare education inequities.
ChatGPT and Gemini demonstrated more diverse demographic outputs compared to Claude, which skewed towards older, White, male profiles, especially in leadership.
Addressing AI biases through ethical development, AI literacy, and promoting diverse leadership is crucial for equitable SBME.

Keywords:

Artificial intelligence (AI) Bias Diversity and inclusion Ethical AI design Gender representation Healthcare education Simulation-based medical education