A randomized crossover trial with 30 fourth-year preclinical dental students compared verbal, visual, haptic, and no VR mirror guidance during simulated maxillary-molar access preparation. Each guidance mode reduced the outcome error score and increased visible operating-area time versus no guidance, while obstruction, posture, completion time, and usability did not differ significantly. The evidence supports short-term performance in a simulation laboratory, not skill retention, clinical transfer, or patient outcomes.
Key findings
- Guidance modality affected outcome error score (F(3,84)=20.34, p
- Visible operating-area time increased versus no guidance by 37.79 seconds for verbal (95% CI 11.15–64.43), 59.15 seconds for visual (32.51–85.79), and 31.52 seconds for haptic guidance (4.90–58.15).
- No significant differences were detected for obstruction, posture error, completion time, or usability. These were exploratory outcomes without multiplicity correction across all outcomes, so nonsignificance is not evidence of equivalence.
- Participant comments suggested that multiple visual elements could be complex and that haptic direction was not always clear, consistent with a device providing three-degree-of-freedom force but no rotational torque.
- Reuse of the same virtual tooth and the preclinical setting mean that the study measured immediate simulated performance, not skill retention or patient-care quality.
Why this matters globally
Automated feedback could standardise repeated dental practice and reduce the need for continuous instructor input, especially for indirect-vision skills that are difficult to verbalise. The crossover design compares several modalities within the same learners, but adoption requires evidence of retention, transfer across teeth and into clinics, accessibility for different visual needs, cost, and added value over standard instruction.
Thai researcher contribution
Unthitar Suranin, Siriwan Suebnukarn, and Panuroot Aguilar are affiliated with Thammasat University's Faculty of Dentistry; Kamon Budsaba with Thammasat's Department of Mathematics and Statistics; and Peter Haddawy and Myat Su Yin with Mahidol University. Siriwan Suebnukarn is corresponding author. The trial ran in Thammasat's VR laboratory and was supported by its Faculty of Dentistry Research Fund.
Limitations to consider
The trial enrolled 30 students from one year and one institution. It was powered for the primary outcome and may miss smaller secondary effects. Corrected visual acuity and refraction were not measured; repeated use of one virtual tooth may create familiarity; the haptic device lacked rotational torque; and each modality bundled several components, preventing component attribution. VRSUQ needs dental-context validation, and open comments were exploratory. Most importantly, short-term simulation performance does not establish clinical safety or effectiveness.
Verify the original sources
JMIR Serious GamesOriginal article in JMIR Serious Games↗DOI: 10.2196/100884