BackgroundThe main objective of this study was to determine whether virtual reality (VR) reduces procedure-related pain and other outcomes for epidural steroid injections (ESI). Randomized controlled study was undertaken in academic centers in the U.S. and Thailand where 146 patients with lumbosacral radiculopathy undergoing an ESI were allocated to immersive VR with local anesthetic, sedation with midazolam and fentanyl + local anesthetic, or local anesthetic alone.The primary outcome was procedure-related pain recorded on a 0-10 scale. Other immediate-term outcome measures were pain from a standardized subcutaneous skin wheal, procedure-related anxiety, ability to communicate, satisfaction, and time to discharge. Intermediate-term outcome measures at 4 weeks included back and leg pain scores, function, and success defined as a >2-point decrease in average leg pain coupled with a score >5/7 on a Patient Global Impression of Change scale. FindingsProcedure-related pain scores with both VR (mean 3·7 (SD 2·5)) and sedation (mean 3·2 (SD 3·0)) were lower compared to control (mean 5·2 (SD 3·1); mean differences -1.5 (-2.7, -0.4) and -2.1 (-3.3, -0.9), respectively), but VR and sedation scores did not significantly differ (mean difference 0.5 (-0.6, 1.7)). Among secondary outcomes, communication was decreased in the sedation group (mean 3·7 (SD 0·9)) compared to the VR group (mean 4·1 (SD 0·5); mean difference 0.4 (0.1, 0.6)), but neither VR nor sedation was different than control. The trends favoring sedation and VR over control for procedure-related anxiety and satisfaction were not statistically significant. Post-procedural recovery time was longer for the sedation group compared to both VR and control groups. There were no meaningful intermediate-term differences between groups except that medication reduction was lowest in the control group. InterpretationVR provides comparable benefit to sedation for procedure-related pain, anxiety and satisfaction, but with fewer side effects, superior communication and a shorter recovery period. FundingFunded in part by grants from MIRROR, Uniformed Services University of the Health Sciences, U.S. Dept. of Defense, grant # HU00011920011. Equipment was provided by Harvard MedTech, Las Vegas, NV