Virtual-reality simulation to assess performance in hip fracture surgery

BACKGROUND AND PURPOSE: Internal fixation of hip fractures is a common and important procedure that orthopedic surgeons must master early in their career. Virtual-reality training could improve initial skills, and a simulation-based test would make it possible to ensure basic competency of junior surgeons before they proceed to supervised practice on patients. The aim of this study was to develop a reliable and valid test with credible pass/fail standards. METHODS: 20 physicians (10 untrained novices and 10 experienced orthopedic surgeons) each performed 3 internal fixation procedures of an undisplaced femoral neck fracture: 2 hook-pins, 2 screws, and a sliding hip screw. All procedures were preformed on a trauma simulator. Performance scores for each procedure were obtained from the predefined metrics of the simulator. The inter-case reliability of the simulator metrics was explored by calculation of intra-class correlation coefficient. Validity was explored by comparison between novices' and experts' scores using independent-samples t-test. A pass/fail standard was set by the contrasting-groups method and the consequences were explored. RESULTS: The percentage of maximum combined score (PM score) showed an inter-case reliability of 0.83 (95% CI: 0.65-0.93) between the 3 procedures. The mean PM score was 30% (CI: 7-53) for the novices and 76% (CI: 68-83) for the experienced surgeons. The pass/fail standard was set at 58%, resulting in none of the novices passing the test and a single experienced surgeon failing the test. INTERPRETATION: The simulation-based test was reliable and valid in our setting, and the pass/fail standard could discriminate between novices and experienced surgeons. Potentially, training and testing of future junior surgeons on a virtual-reality simulator could ensure basic competency before proceeding to supervised practice on patients.