Feasibility and preliminary psychological effects of bedside immersive Virtual reality in prolonged bedridden...

Prolonged immobility is a common consequence of neurosurgical spinal pathology and is associated with significant psychological morbidity, including anxiety, low mood, and reduced quality of life.[1,2] Conventional psychological support is often impractical in high-volume neurosurgical ward settings. Virtual reality (VR) has shown promise as a non-pharmacological adjunct in several clinical contexts,[3-10] though its specific application in prolonged bedridden neurosurgical patients has not been systematically explored. To evaluate the feasibility, tolerability, and safety of delivering structured VR sessions to prolonged bedridden neurosurgical patients (primary objective), and to obtain exploratory estimates of any associated change in anxiety, affect, mood, and quality of life to inform a future controlled trial (secondary objective).

Ten patients (5F/5M; mean age 38.3 ± 15.73 years) with spinal pathologies at the All India Institute of Medical Sciences (AIIMS), New Delhi, bedridden ≥ 15 days, underwent seven VR sessions delivered every third day (5-30 min each). GAD-7, PANAS, and POMS were assessed pre- and post-each session; WHOQOL-BREF before Session 1 and after Session 7. Wilcoxon Signed-Rank, Friedman and Spearman tests were used, with a random-intercept linear mixed-effects model fitted to all 140 pre- and post-session observations per measure, estimating cumulative (across-session) and acute (within-session) effects separately.

The intervention was feasible and well tolerated: all ten patients completed all seven sessions (100 % completion) with no adverse events. Within-patient psychological measures all moved in the direction of improvement, providing preliminary effect-size estimates to inform a future controlled trial. GAD-7 declined by a median of 5.0 points (bootstrap 95 % CI 3.0 to 7.0; p = 0.002), shifting the cohort from the upper-moderate into the mild anxiety range. Positive affect improved (p = 0.006), negative affect declined (p = 0.002), and POMS-TMD fell (p = 0.008). WHOQOL-BREF domain scores rose in the physical (21.4 to 27.5), psychological (8.8 to 20.4) and environment (8.1 to 17.5) domains (all p = 0.016), with no change in the social relationships domain. Because all participants changed in the same direction, the corresponding effect-size r values reach their mathematical ceiling for n = 10 (≈0.89) and indicate consistency of direction rather than magnitude; all estimates are exploratory given the single-arm design. An exploratory association between bedridden duration and GAD-7 change (ρ = 0.854) is hypothesis-generating only.

These preliminary findings suggest that immersive VR is feasible, well-tolerated, and associated with preliminary improvements in psychological well-being in prolonged bedridden neurosurgical patients. The results support further evaluation through a randomized controlled trial.