Usability of virtual-reality simulation training in obstetric ultrasonography: a prospective cohort study

To assess the usability of virtual-reality (VR) simulation for obstetric ultrasound trainees.

Twenty-six participants were recruited: 18 obstetric ultrasound trainees (with little formal ultrasonography training) and eight certified experts. All performed five sequential VR-simulated crown-rump length (CRL) scans in a single session and three repetitions of biparietal diameter (BPD), occipitofrontal diameter (OFD) and femur length (FL) measurements. Outcome measures included mean percentage deviation from target for all measurements. Time taken to perform each type of scan was recorded.

The mean percentage difference for the first scan was significantly greater for the trainee group than for the expert group for BPD (P = 0.035), OFD (P = 0.010) and FL (P = 0.008) and for time taken for the first CRL (P < 0.001) and fetal biometry (including BPD, OFD and FL measurements) scan (P < 0.001), demonstrating that trainees were initially significantly less accurate and less efficient. Over subsequent scans, the trainees became more accurate for all measurements with a significant improvement shown for OFD and FL (P < 0.05). The time taken for trainees to complete CRL and fetal biometry scans decreased significantly (all P < 0.05) with repetition, to near-expert efficiency.

All participants were able to use the simulator and produce clinically meaningful biometry results. With repetition, beginners quickly approached near-expert levels of accuracy and speed. These data demonstrate that obstetricians with minimal experience can improve their ultrasonographic skills with short-phase VR-simulation training. The speed of improvement suggests that VR simulation might be useful as a warm-up exercise before clinical training sessions in order to reduce their impact on clinical service.