As health-care education has become more inclusive of digital resources, a pattern is emerging: virtual simulation platforms often fail to meaningfully reflect patient diversity, which may be shaped by underlying biases and have implications for clinical outcomes.
As one emergency medicine resident noted on the diagnosis of myocardial infarction (MI) and patient gender: “I have begun to notice ways in which my training has changed the patterns I look for and the symptoms I consider significant.” The resident specifically highlighted the missed symptoms in women, potentially stemming from “the under-representation of women in the interconnecting systems that comprise medical education, including in the systems that use simulation in health-care training.”
Reflecting on the training, the resident shared, “a masculine face had worked its way into the constellation of signs I looked for to identify MI patients, as out of the total of nine virtual patient simulations, not a single one featured a woman.”
This personal account underscores a broader issue in medical education, specifically the need for meaningful, inclusive, and representative simulation-based training. This training can be defined as “a technique – not a technology – to replace or amplify real experiences with guided experiences that evoke or replicate substantial aspects of the real world in a fully interactive manner,” helping learners develop critical skills and make decisions without risk to actual patients.
While in-person simulations offer valuable hands-on experience, they are also resource-intensive. As a result, many programs are turning to virtual simulations – computer-based platforms that recreate realistic clinical scenarios, allowing users to practice clinical activities, make decisions and interact with virtual patients or team members.
Unfortunately, many simulations still lack meaningful social identity representation, an important part of training that can impact patient care. Although in-person simulations (e.g., through using mannequins or having standardized patients) are common, many lack representation (e.g., mannequins that are white male) and purchasing diverse mannequins can be quite costly. Virtual simulations done properly can expose learners to a wider range of complex cases and patient identities (including diversity in sex, gender and race). The value lies in showing how identity can influence clinical presentation and decision-making, and why that matters. When designed thoughtfully, virtual simulations can support clinical skill development while also helping learners meaningfully recognize and mitigate potential implicit biases in health care.
To better understand the current landscape, we conducted an environmental scan of virtual medical simulations, focusing on the types of identity representation present and the scenarios in which they appear.
The value lies in showing how identity can influence clinical presentation and decision-making, and why that matters.
A total of 58 virtual simulation platforms listed on the Simulation Canada website were explored. Of these, nine platforms were not freely accessible, while 23 provided access only to their landing pages, without the ability to view or interact with individual simulations without subscription. Among the 49 platforms that could be accessed, 19 were designed primarily for medicine, 12 for nursing education, 10 for allied health or technical training, three for mental health and communication, and five were multidisciplinary in scope.
Upon analysis, a wide variation in the diversity of avatars across platforms was observed. Of the simulations that could be accessed, 14 appeared to intentionally embed diversity by using inclusive imagery and clearly articulating their commitment to representing a range of races, genders and cultural backgrounds in their avatars and videos. Another 18 platforms demonstrated what appeared to be unintentional or incidental diversity, depicting avatars from different backgrounds but without explicit mention of inclusion or their pedagogical goals. The remaining 17 platforms showed little to no diversity, either due to neutral or monochromatic design choices or the absence of visible avatars altogether.
Additionally, analysis revealed discipline-specific trends in representation. Nursing-focused simulations demonstrated the most intentional and consistent use of diverse avatars, integrating racial, gender and cultural variation directly into patient encounters. These platforms tied diversity to learning objectives related to holistic patient-centred care, reflecting nursing education’s emphasis on cultural competence and health equity. In contrast, medical and interprofessional simulations tended to prioritize technical accuracy, diagnostic reasoning and procedural management. Although diversity was present in these platforms, it was often incidental rather than pedagogically driven. Lastly, technical and allied health simulations displayed little to no diversity, often using ‘neutral’ avatars. These findings highlight how the integration of identity representation varies not only by platform design but also by the educational priorities of different health disciplines.
While a few virtual simulations demonstrate intentional and meaningful representation, most platforms fail to embed diversity in ways that allow learners to engage with realistic, context-specific patient profiles. This gap could diminish the educational value of simulation-based training by limiting exposure to the full range of patient identities encountered in practice. Notably, one study found that when nursing simulations intentionally incorporated representation related to sexual orientation and gender identity, participants reported both deeper learning and greater awareness of their own biases.
Moving forward, as virtual simulation continues to expand within health education, it is essential that representation becomes both intentional and contextually relevant.
As seen in several of the nursing simulations, representation is most impactful when it is directly connected to the learning objectives of the case, rather than simply changing the visual appearance of a patient avatar. It is also important to recognize that diversity extends beyond skin tone or outward characteristics; identity should be woven meaningfully into the case narrative and followed by structured debriefing that explores how identity influences communication, symptom expression or clinical decision-making.
To achieve this, developers and educators should collaborate closely with patients, caregivers, interprofessional teams and community partners to ensure accuracy, authenticity and inclusivity in simulation design.
Ultimately, this approach aligns representation with clinical relevance, allowing learners to move beyond associating certain conditions with a single type of patient and instead recognize the full diversity of individuals who experience illness.
Conflicts of Interest: This project is supported in part by funding from the Physician’s Services Incorporated (PSI) Foundation (PI: Dr. Quang N. Ngo, co-PI: Dr. Elif Bilgic). Data collection and extraction was done by Simran Dhami; however, the team used ChatGPT to summarize the extracted data.
