Findings Favor Haptics Feedback in Virtual Simulation Surgical Education: An Updated Systematic and Scoping Review

Background Virtual simulations (VSs) enhance clinical competencies and skills. However, a previous systematic review of 9 RCT studies highlighted a paucity of literature on the effects of haptic feedback in surgical VSs. An updated systematic and scoping review was conducted to encompass more studies and a broader range of study methodologies. Methods A systematic literature search was conducted on July 31, 2023, in MEDLINE, Embase, and Cochrane. English language studies comparing haptic vs non-haptic conditions and using VSs were included. Studies were evaluated and reported using PRISMA-ScR guidelines. Results Out of 2782 initial studies, 51 were included in the review. Most studies used RCT (21) or crossover (23) methodologies with medical residents, students, and attending physicians. Most used post-intervention metrics, while some used pre- and post-intervention metrics. Overall, 34 performance results from studies favored haptics, 3 favored non-haptics, and the rest showed mixed or equal results. Conclusion This updated review highlights the diverse application of haptic technology in surgical VSs. Haptics generally enhances performance, complements traditional teaching methods, and offers personalized learning with adequate simulator validation. However, a sparsity of orienting to the simulator, pre-/post-study designs, and small sample sizes poses concerns with the validity of the results. We underscore the urgent need for standardized protocols, large-scale studies, and nuanced understanding of haptic feedback integration. We also accentuate the significance of simulator validation, personalized learning potential, and the need for researcher, educator, and manufacturer collaboration. This review is a guidepost for navigating the complexities and advancements in haptic-enhanced surgical VSs.


PRISMA flow diagram
For the more difficult cutting, the time to complete the tasks was shorter with haptics, at all levels of difficulty (p < 0.05).Grasping tension was shorter at level 1 (11.2 ± 3.3 versus 7.9 ± 1.8).Data on instrument path length and errors trended towards improve efficiency from haptics at all difficulty levels but did not achieve statistical significance.Haptics reduced the time taken for manual reduction procedure (p < 0.0001).Haptics in the manual reduction group led to better outcomes in terms of varus/valgus deformity (p = 0.034).Haptics in the manual reduction group resulted in reduced ante-/recurvature deformity (p < 0.01).Haptics did not contribute significantly to reducing malrotation in the manual reduction group (p < 0.001).Haptics in the robot-assisted reduction group resulted in improved image control, requiring fewer camera changes.Haptics in the robot-assisted reduction group led to reduced distraction within the fracture gap (p < 0.001).No significant difference in visual spatial ability (BASIQ and MRT-A tests).Flow questionnaire and Borg CR10 showed no significant differences.After 1 hour of training, no significant differences in MD and PD tasks.After 2 hours of training, haptics showed better performance in both MD and PD tasks (p = 0.01, p < 0.05).Only haptics-first group showed improvement during the second session in both MD and PD tasks (p < 0.001 and < 0.01).Virtual haptic-simulated repair did not significantly differ from surgical repair (p = 0.42).Virtual CAD-simulated repair (non-haptics) differed from surgical repair (p = 0.02).Mean measurements between haptic-and CAD-(non-haptics) differed (p = 0.006).Surgical values were greater than haptics (p = 0.03).More consistency between users for haptics than CAD-(p ≤ 0.001).Self-reported quality of repair and intuitiveness was higher for haptics than CAD (p = 0.04 and p = 0.003).No difference in the number of views used between CAD and haptics (p = 0.6).There was a significant improvement after 1 h of training in the Procedicus MIST regarding total score (p < 0.001) and in the Procedicus KSA regarding time, movement economy, collisions with the probe, and total score (p < 0.001 The peg transfer drill showed no difference in performance between the haptic and non-haptic simulations for all metrics at all three levels of difficulty.Task completion time, instrument path length, and error quantification demonstrated similar values in the haptic and non-haptic environments at all levels of difficulty.In more difficult cutting, time was shorter with haptics, at all levels of difficulty (p < 0.05).Grasping tension was shorter at level 1. Tactile feedback resulted in more alleviation of deafferentation pain (p = 0.047).Average reduction rates were different from 0 under tactile feedback (p = 0.02) but not under no tactile feedback (p = 0.078).Ownership and agency scores did not significantly differ between the two conditions (p = 0.13 and p = 0.69, respectively), and their correlation with pain reduction rates was weak and not significant (p = 0.67).

+
- Evaluation task was an endoscopic sinus surgery Performance metrics -quality, efficiency, safety, calculated learning, curve for each task Endoscopy task: Haptic feedback improved task completion time compared to the non-haptic condition (p < 0.001).There was improvement in task completion time between the first and third attempts, which was sustained during the eighth attempt with haptic feedback (p = 0.05 for the first vs third attempt, p = 0.001 for the first vs eighth attempt).The variance between trainees narrowed with successive practice attempts using haptic feedback.
Polypectomy task: No significant difference in task performance between the haptic and non-haptic conditions.
Evaluation task: Haptics trended towards better performance, although improvement scores did not reach statistical significance (p = 0.09 for first vs seventh attempt, p = 0.09 for first vs eighth attempt).The haptics group achieved a smaller plunge gap distance (p = 0.012), longer mean drilling time (p = 0.027), higher combined OSATS scores (p = 0.0006), higher ratings in the procedural flow domain (p = 0.029), higher scores in the instrument handling domain (p = 0.053), and more participants in the haptics group achieved a learning plateau compared to the non-haptics group (p = 0.01).

Surgical Innovation
through both cortices of the bone while minimizing drill plunge depth beyond the far side of the bone to prevent tissue damage.
The haptic group received haptic feedback upon use and manipulation of the drill, while the non-haptic group relied on visual and auditory feedback alone.
Objective structured assessment of technical skills (OSATS) where expert consultant surgeons rated video performances based on time, motion, instrument handling, procedural flow, overall performance, safe drill (achieving plunge gap of <5 mm in at least 2/3 attempts), learning curve plateau, and participant feedback questionnaire.
The haptics group showed a significant difference with 40% of participants achieving a safe drill depth in 2/3 attempts, while none in the nonhaptics group achieved this (p = 0.02).Both groups had similar ratings in the time/motion and overall performance domains (p = 0.053).In the participant questionnaire, more participants in the haptics group reported that instruments felt and sounded more realistic than the non-haptics group (p = 0.006 and p = 0.03, respectively), both groups reported high ratings for simulation enjoyment and perceived education value, both groups agreed that the simulation improved theoretical knowledge and instrument handling, the haptics group rated the role of simulation in recognizing when to stop drilling significantly higher than the non-haptic group (p = 0.039, and both groups expressed positive views on the use of simulation in surgical training and its potential to improve clinical outcomes.The Wilcoxon test detected statistically significant decreased time taken to gain access to the surgical area (p = 0.033) and the time taken to attach the surgical port, perform a facetectomy, and use the powered burr, as well as the total procedural time (p = 0.017) in the haptics group.Statistical significance was observed of time spent using the surgical probe and the powered burr (p = 0/017).The median score allocated to 94% of the questions was above the defined limit of sufficient validity (≥3/5) in the questionnaire.Distance to ideals for the middle inferior, superior anterior, and superior posterior aspects of the femoral neck, and to the articular surface of the femoral head and the distance from the lower border of the lesser trochanter to the most inferior pin.Primary evaluator evaluation of postprocedure CT scans and final AP and lateral fluoroscopic images to determine whether triangle construct was achieved.

Scott
The haptics trained group was significantly better at achieving an inverted triangle construct compared to the untrained group (p = 0.05).
No statistically significant differences were observed between the haptics trained and untrained groups regarding the number of AP or lateral view images obtained or fluoroscopic time.No statistically significant differences were found between the haptics trained group and untrained groups related to the distance from the tip of the wire to the femoral head articular surface as shown by the CT in the axial and coronal planes.There were statistically significant differences favouring the haptics trained group for the angle between the inferior and superior posterior pin (p < 0.01), angle between the inferior and superior anterior pin (p = 0.02), and superior anterior pin distance to mid-neck (p < 0.01).

Primary molar pulpotomy procedure
Performance metrics -procedural time, number of instructor prompts, and student perceptions which included questions on the realism of images displayed on the monitor, tactile sensation, and texture/hardness of dental tissue, the impact of psychomotor skills and confidence, and whether training can replace or supplement conventional pre-clinical training.
The haptics group had non-significant higher access outline scores (p = 0.67), non-significant higher deroofing skills (p = 0.37), non-significant different in access outline prompts (p = 1.0), nonsignificant deroofing prompts, non-significantly lower procedural time, and no difference in the number of instructor prompts.Student perceptions -72% of students agreed/strongly agreed that images of teeth, pulp chamber, and instruments displayed on the monitor looked realistic, mixed responses on the realism and tactile sensations and texture/ hardness of +/-

Surgical Innovation
dental tissues, 443% agreed/strongly agreed that the tactile force feedback felt realistic, 50% indicated that they could not differentiate between enamel and dentine texture/hardness on the simulator, 64% of students agreed/strongly agreed that training on the haptics device improved their fine motor dental skills and confidence in performing the pulpotomy procedure, a significant majority (86%) disagreed/strongly disagreed with the notion that haptics training could replace conventional pre-clinical training on typodont teeth for the pulpotomy procedure, 57.1% of students expressed interest in having more haptic session for preclinical pediatric pulp procedures, and student responses from open-ended questions indicated that the main benefits of haptic training was increased practice opportunities, better visualization, and improved fine motor skills.

Cavity inspection
Implantology-related drilling outcomes

Time Measurement 1
Subjective Assessment 9 Objective Assessment 2

Surgical Innovation
Table 3. Instructor assistance during follow-up Did not favour haptics The time needed for instructor assistance was significantly higher for the haptic group compared to the non-haptic group (p<0.001).

Surgical Innovation
Instructor assistance between intervention and follow-up Mixed The haptic group required significantly more feedback from the instructor for the followup test than the assistance needed during the intervention phase (p < 0.001; e.g., once haptic group transitioned to non-haptic they required more instructor assistance).In contrast, the non-haptic group used significantly less help from the instructor during the follow-up phase (p < 0.001). Number Figure 1.
their experience of hapticenabled test condition and haptic-disabled test condition Favoured haptics Time spent to reach the predefined proficiency level (

Table 1 .
Proficiency in Basic tasks 5 in the Haptic was achieved in MTST of 12:49 in 7.3 trials compared to 16:28 minute in 7.7 trials for non-haptic.Proficiency in Basic tasks 6 in the Haptic was achieved in MTST of 12:20 min in 7.2 trials compared to 19:22 minute in 7.2 trials for nonhaptic.Proficiency in Procedural tasks 3 in the Haptic was achieved in MTST of 26:42 in 5.3 compared to 59:19 minute in 12.4 for non-haptic.Proficiency in procedural tasks 4 in the Haptic was achieved in MTST of 27:40 in 5.2 trials compared to 1:05:25 minute in 8 trials for non-haptic.In full LC proficiency was achieved in MTST of 30:04 minute in 3.4 trials compared to 1:27:43 minute in 8.1 trials for non-haptic.
Participants found several mechanical properties of the physical bone model (non-haptics) to be more comparable to a cadaveric bone than the virtual model (haptics) -differences in cortical (p = 0.011) and trabecular (p = 0.004) osseous realism, vibrational properties (p = 0.001), and air system generation (p = 0.003).No differences found between virtual and physical bone model in anatomic features.Both virtual and physical bone model found to be productive resources for acquiring surgical skills.Physical bone model ranked more effective for learning cortical mastoidectomy (p = 0.015) and the posterior tympanotomy/facial recess approach to the middle ear (p = 0.023).Physical bone model considered the superior tool in 7/9 domains (p < 0.05).No difference in ease of use and use as a visual learning tool.

Table 4 .
of malfunctions and time spent solving them during the intervention and the follow-up testMERSQI Results

Table 5 .
CASP RCT Standard Checklist Results

Table 6 .
JBI Critical Appraisal Checklist for Quasi-experimental Studies Results