Abstract
Purpose: :
To develop and test a device that simulates the capsulorrhexis procedure on a computer system and to test user performance metrics to evaluate performance.
Methods: :
A prototype simulator was developed that consisted of visualization of a 3D image of the anterior segment of the eye including the cataractous lens overlaid on to a deformable membrane viewed through passive stereo goggles. A virtual forceps controlled by a robotic arm was used to initiate and propagate the capsular tear. Three performance based metrics (i.e., time to complete capsulorrehexis, number of hits on anterior lens surface, and number of forceps grabs of capsule) were evaluated in 28 residents at different training levels and 10 attending physicians. In addition, results from a feedback questionnaire were evaluated. Data analysis included comparison of the various metrics by year of training. In addition, Pearson correlation coefficients were calculated to estimate the relationships among the three performance metrics and self-report assessments of the procedure.
Results: :
Partially based on the small sample size, no significant group differences were noted for the three performance metrics when compared to the level of training. There seemed to be, however, moderate variability in performance across groups. Indeed, when collapsing the data between less experienced (PGY2 and PGY3) and more experienced (PGY4 and Attending) surgeons, there was a trend (p=.067) showing that less experienced surgeons took longer to complete the procedure. For the full sample, significant correlations (ranging from .61 to .70) were noted among all 3 performance metrics. The questionnaire feedback from this small group was mostly positive. The users confirmed that the simulator would be effective in developing surgical skills in residents, was easy to use, was better than porcine eyes, and the respondents felt that they would use the product in the future if available. No significant correlations were found between the questionnaire data and performance metrics.
Conclusions: :
Data from this pilot project suggest that simulator performance was variable and showed trends of better performance with experience. We believe that the moderate variability in metrics may have resulted from simulator training protocol and individual computer skills.
Keywords: cataract • learning