Simulation and AI reshape how health care students rehearse hard decisions safely

Before a hospital administrator ever signs off on a staffing plan, a budget cut, or a patient safety intervention, the consequences of that decision can ripple through an entire institution. Yet the people trained to make those calls have historically learned them the way most professionals do: by reading about them, discussing them in a seminar, and then hoping the lessons hold when reality arrives. A new review from George Mason University suggests that this old model is being rapidly overtaken by something far more immersive, in which students rehearse high-stakes leadership decisions inside simulations, mobile platforms, and even AI-generated scenarios long before any real patient or budget is on the line.

The study, a scoping review led by John Cantiello, a professor in the Department of Health Administration, Policy, and Informatics at George Mason’s College of Public Health, with co-author Renee Geschke, pulls together 83 articles published between 2017 and 2025. The work appears in the Journal of Health Administration Education as part of a new special issue devoted to case studies, simulations, and gamification. Its central question is deceptively simple: how has case-based learning, the long-standing practice of teaching through realistic problems, evolved in the age of simulation technology and generative artificial intelligence, and what conditions determine whether these innovations actually improve learning?

Case-based learning, often abbreviated as CBL, asks students to apply what they know to realistic problems, weigh competing options, and make decisions before the consequences are real. The review makes clear that the approach has traveled a long way from the traditional image of students reading a printed case and debating it in class. In one of the studies captured by the review, students on a pediatric orthopedic rotation worked through clinical cases together on WeChat, the Chinese messaging platform, turning a mobile app into a collaborative learning space. In another, medical students tested their judgment in high-pressure scenarios using lifelike mannequins that could simulate physiological deterioration. And in health administration classrooms, educators have begun experimenting with ChatGPT to generate cases that put management concepts into practice.

That last example illustrates both the promise and the peril that run through the review’s findings. Generative AI can produce case material quickly and tailor scenarios to specific learning objectives, but the authors caution that newer does not necessarily mean better. The reviewed literature documents cases in which generative AI omitted important information or produced inaccurate answers, and in which students came to rely on the technology too heavily. The review’s response to this is a principle rather than a prohibition: technology should be chosen to support specific learning goals, not treated as the innovation itself. A virtual reality headset, a chatbot, or a high-fidelity mannequin is only as valuable as the pedagogical design wrapped around it.

Cantiello summarized the recurring theme across the 83 studies in blunt terms. Technology, he noted, is opening up new possibilities for case-based learning, including realistic simulations and feedback tailored to individual students. But across the literature, what mattered most was using those tools within carefully designed cases that asked students to work through realistic problems, make decisions, and apply what they were learning. Faculty guidance, discussion, and feedback remained central to that process. In other words, the tools have changed dramatically; the fundamentals of good teaching have not.

The review distills several practical lessons for educators. The first is to use technology for a reason. Simulations and virtual or augmented reality can place students inside a realistic scenario where they must act, not just analyze. Mobile platforms can make cases easier to use outside the classroom, extending learning into clinical rotations and workplaces. AI can help generate cases or provide feedback at a scale no single instructor could match. Each of these capabilities maps onto a distinct educational purpose, and the review argues that educators should start with the purpose rather than the tool.

The second lesson is to make students do something with the problem. The review highlights flipped classrooms, peer instruction, team-based cases, and interdisciplinary exercises, all of which require students to make decisions and explain their reasoning, often alongside classmates who approach the same problem from a different professional angle. Across the reviewed studies, these more active approaches were linked to stronger critical thinking, better problem-solving, and higher engagement. The contrast with passive case discussion is stark: a scenario that students merely read about produces far less durable learning than one they must navigate, defend, and revise in real time.

The third and fourth lessons address the scaffolding around the case itself. A complex scenario alone does not guarantee useful learning, the review emphasizes. Instead, effective implementations set clear learning goals, present structured cases, provide timely instructor feedback, and guide discussion so that it stays focused directly on the goals of the lesson. Without that structure, an elaborate simulation can become an expensive distraction. And because developing successful cases takes time and skill, the review calls for institutional support: faculty training, technical support, and deliberate curriculum planning are necessary if these approaches are to last beyond a handful of enthusiastic early adopters.

The review is especially significant for health administration, a field in which only limited research has focused specifically on case-based learning. To fill that gap, Cantiello and Geschke looked to adjacent disciplines, including nursing, public health, medicine, and business, for approaches that could transfer to the education of future health care administrators. The logic is pragmatic: the decisions administrators face, from resource allocation to quality improvement to crisis response, share structural features with the dilemmas already being simulated in clinical and business education. Borrowing and adapting proven practices from those fields offers a faster route to a mature pedagogy than starting from scratch.

What comes next, according to Cantiello, is more research aimed directly at health administration rather than borrowed evidence. More discipline-specific work is needed, he said, particularly around outcomes such as leadership, decision-making, implementation at scale, and how the effectiveness of case-based learning should be measured. That measurement question is not trivial. If programs invest in simulations, AI-generated cases, and faculty development, they will need rigorous evidence that these investments produce administrators who decide better, not just students who report enjoying class more. The special issue also features related work by Phillip Zane and Deborah Goldberg on game-based learning in health economics, signaling a broader institutional push to treat play, simulation, and case work as serious instruments of professional education. For a generation of future health care leaders, the message of the review is quietly transformative: the hardest decisions of their careers can now be rehearsed, refined, and learned from in the classroom, where the only thing at stake is the lesson itself.

Subject of Research: Case-based learning in health care education using simulation and artificial intelligence

Article Title: How future health care leaders practice tough decisions before the stakes are real

Article References: How future health care leaders practice tough decisions before the stakes are real. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: case-based learning, health administration education, simulation, artificial intelligence, generative AI, scoping review, George Mason University, medical education, virtual reality, flipped classroom, faculty development, health care leadership