VA Pilot Shows Value of AI-Powered VR Training

An AI-powered virtual reality system is enabling clinicians with the Department of Veterans Affairs’ Southern Nevada Healthcare System to train and sharpen their skills remotely and on their own schedules.

The system can simulate patients with a variety of needs, while its virtual reality component enables training in simulated hospital environments, according to Bonnie Barti, a simulation technician at VA’s North Las Vegas VA Medical Center.

“We needed a way to deliver realistic, flexible training that didn’t interrupt patient care, and that we could scale across our VA,” Barti told GovCIO Media & Research.

The agency developed the AI system due to the North Las Vegas VA Medical Center lacking computer-based and virtual reality programs for staff education, Barti said. Existing training opportunities were limited by classroom availability and scheduling constraints because staff were pulled away from caring for patients.

Barti noted that the VA has a strong history of using virtual reality in patient care, creating an opportunity to expand its use. Working with experts at the facility, she developed alternative training methods with veterans and staff in mind. The system is designed to meet the needs of off-campus employees, night-shift personnel, weekend staff and others working nontraditional schedules.

The platform’s AI simulates a variety of patients with whom staff can have real-time conversations.

“It feels like speaking with an actual person,” Barti said.

Barti said the first scenario focuses on interacting with a disruptive patient. Additional courses cover women’s health, behavioral health, recognizing suicidal ideation, chronic disease management, supervisor feedback and crisis intervention. The courses are repeatable and available on demand. Staff often complete them multiple times to improve their scores and reinforce learning, she added.

The VR Platform

Barti’s team selected an immersive training platform that offers two primary features: AI-enabled virtual human patients and 360-degree video recordings.

The recordings allow the creation of lifelike clinical scenarios that staff can access anytime and from any location. Another key factor in selecting the platform was the facility’s ability to maintain full control over the content and capabilities.

“Other systems are more costly. They take months to create by offsite developers, and once they’re created, you can’t change or modify them,” Barti said. “The platform is based on licenses, not the amount of content that we create. This provides us with the opportunity to create meaningful content without concern of cost of creation.”

The platform also includes analytics capabilities that provide metrics for staff development, she added.

Pilot to Proof of Concept

The virtual training project is halfway through a two-year proof-of-concept pilot.

During the second year of the project, Barti said six training courses have been created using the platform’s 360-degree camera to support virtual reality applications in hospital settings. These include a mass-casualty training exercise involving much of the facility’s staff and other participants.

The team has also created 25 virtual humans representing a variety of patient types, allowing staff to practice managing and interacting with a range of individuals in clinical settings. Additionally, Barti said the system has been in use long enough to support broader training deployments, accelerate program development and collect performance and feedback data.

The AI-based training is deployed monthly to a variety of groups. The facility currently has 60 licenses that can be assigned and reassigned as needed. One of the platform’s strengths, Barti said, is the team’s ability to maintain complete control over the content it creates.

“It’s a great use of funding,” she said.

Positive Performance

The system has produced promising results, with participants demonstrating an average 12.7% increase in knowledge performance. The improvement reflects stronger retention and practical understanding, Barti said.

Staff also reported significant gains in confidence and skills in several areas, including safe and effective patient de-escalation, application of Prevention and Management of Disruptive Behavior principles, and adherence to protocols for managing disruptive behavior.

The training also has financial benefits. Barti said training costs fell from $130 per person per session to approximately $11.29 per person per session. Allowing staff to train asynchronously reduces disruptions to patient care and eliminates productivity losses, she added.

Feedback from staff has been positive. Participants described the training as “realistic, accessible and engaging,” according to Barti. The AI-powered patients provide a level of authenticity that helps build confidence and reinforce learning.

The platform’s ease of use also facilitates collaboration with subject-matter experts. Barti said the positive feedback has improved the speed, ease and fidelity of content creation.

“The longer we use it, the easier the system is to work with,” she said.

The team is now exploring expanding the scenario library. Because the platform can be used by a wide range of personnel, it supports greater interactivity and the development of additional training modules for more clinical roles.

“The goal is to make it a scalable long-term solution that continues to elevate workforce readiness across the VA,” Barti said.