Performance and cognitive workload during haptic virtual reality-based implant osteotomy training: a prospective repeated-measures study

Haptic virtual reality may provide a standardized environment for repeated implant osteotomy practice and objective performance assessment; however, evidence on performance and cognitive workload among surgical residents remains limited. This study evaluated changes during haptic virtual reality–based implant osteotomy practice and whether they differed by residency training level. This single-center, prospective repeated-measures study included 20 residents (12 novice and 8 experienced) who completed six implant osteotomy sessions on the same standardized case at approximately 48-hour intervals. Procedure time and angular, vertical, coronal, and apical deviations were recorded automatically. Cognitive workload was assessed using a Turkish-language administration of the Raw NASA Task Load Index (Raw NASA-TLX) after the first and sixth sessions; the specific Turkish-language form used had not undergone formal cross-cultural validation. Linear mixed-effects models with a random participant intercept and Kenward–Roger Type III tests were used, with Holm correction across primary outcomes. All participants completed all sessions without missing data. After Holm correction, session showed a significant main effect on all five objective outcomes. From session 1 to session 6, estimated mean changes were − 30.73 s for procedure time, − 4.34° for angular deviation, − 0.53 mm for vertical deviation, − 0.82 mm for coronal deviation, and − 1.32 mm for apical deviation; however, the reduction in vertical deviation observed in the primary model was not consistently confirmed across the sensitivity analyses. There was insufficient evidence for experience-group or session × experience-group effects for any objective outcome. Raw NASA-TLX scores decreased by 22.18 points (95% CI, − 29.42 to − 14.94; p < 0.001). Experienced participants had 9.47 points higher overall workload than novices across both assessment time points (95% CI, 2.23–16.71; p = 0.013), whereas there was insufficient evidence that workload reduction differed between groups ( p = 0.715). Repeated haptic virtual reality practice was associated with reductions in procedure time and angular, coronal, and apical deviations and with reduced self-reported cognitive workload. Vertical deviation also decreased in the primary analysis; however, this finding was not consistently confirmed across the sensitivity analyses and should therefore be interpreted cautiously. There was insufficient evidence that changes in objective performance or cognitive workload differed between novice and experienced participants. Haptic virtual reality may provide a standardized complementary environment for monitoring simulator-derived performance and perceived cognitive workload. Because post-training retention and clinical transfer were not assessed, these findings should not be interpreted as evidence of clinical competence or clinical skill transfer.