Background Orthopaedic care is limited in rural U.S. regions and globally. While telemedicine has improved access, barriers remain when clinicians must perform orthopaedic procedures without prior experience or in-person guidance. Off-grid augmented reality (AR) offers real-time visual feedback and support, potentially expanding orthopaedic care in underserved areas. Methods This study evaluated AR’s effectiveness using three simulated orthopaedic scenarios commonly faced by medical professionals in our rural health system: (1) outpatient suture removal, (2) inpatient infected knee aspiration, and (3) urgent care posterior slab splint application. Twenty-four first-year medical students with no prior experience were randomized into an AR-assisted study group (n=12) and a control group (n=12). The AR group received real-time guidance from a remote orthopaedic surgeon via an ARSpectra headset and TeleAssist program over a satellite internet connection. The control group used identical headsets but without AR guidance from the participating surgeon. Both groups received identical AI-generated prompts detailing each scenario and goals. Performances were video-recorded and scored on binary step completion and task duration. In the suture removal task, binary data included sterile technique and successful removal of various suture types (nylon, arthroscopy, Monocryl, Prolene). For the knee aspiration task, data included number of attempts, volume aspirated, and adherence to sterile/no-touch technique. For the splinting task, binary scoring assessed steps such as skin assessment, soft roll placement, metacarpophalangeal (MCP) coverage, fiberglass layering, plaster length, and splint security. Results AR assisted participants had higher success rates across all scenarios. In the suture removal task, participants successfully removed more suture and correctly identified and removed specific types of sutures such as Monocryl (p< 0.001) and Prolene (p<0.001) compared to the control group. In the knee aspiration scenario, the study group aspirated a greater mean volume of fluid compared to the control group (study group mean 41.67 mL vs control group mean 12.67 mL, p=0.002) and demonstrated better adherence to a sterile no-touch technique (p<0.001). In the long-arm splint scenario, the study group outperformed the control group in critical procedural steps such as skin assessment (p<0.001), MCP coverage (p<0.001), and soft roll placement (p=0.037). Task completion time was longer for the AR assisted study group across all simulated scenarios; however, they demonstrated more accurate overall performance. Conclusion AR significantly improved performance of novice medical learners across three orthopaedic procedures, demonstrating its value in telemedicine and remote procedural support. Although task times were longer, higher success and adherence to protocol support AR’s effectiveness. Importantly, off-grid AR programs, including satellite internet and power, functioned reliably in this study. These findings support further exploration of AR to expand orthopaedic care in rural and remote settings.
Bockern et al. (Mon,) studied this question.