Predictive and construct validity of virtual reality cataract surgery simulators

Author: Nayer ZH, Murdock B, Dharia IP, Belyea DA.

Geographical coverage: North and South America, Europe, Asia and Oceania

Sector: Cataract surgery

Sub-sector: Virtual reality

Equity focus: Not reported

Study population: Patients with cataracts

Review type: Effectiveness review

Quantitative synthesis method: Narrative synthesis

Qualitative synthesis method: Not applicable

Background

Cataract surgery is one of the most frequently performed operations worldwide and demands exceptional precision as well as extensive training. Traditional residency training—operating-room instruction, animal models and synthetic eyes—can be time-consuming, costly, variable in quality and lacking objectivity. Early trainees often prolong procedures and experience higher complication rates, increasing healthcare costs. Virtual-reality (VR) simulators such as the EyeSi® now offer immersive, three-dimensional environments with hand–eye-coordination tracking and haptic feedback, enabling standardised and objective assessment of surgical skill. High purchase costs and limited dissemination, however, have slowed widespread adoption.

Objective

To systematically review evidence on VR simulators for cataract-surgery training, assessing both construct validity (ability to discriminate between novice and experienced surgeons) and predictive validity (translation of simulator skill to the operating theatre).

Main findings

Twenty studies published between 2009 and 2019 were included. Investigations came from the United States, Brazil, Canada, Denmark, Italy, Malaysia, New Zealand, Saudi Arabia, Sweden, Turkey and the United Kingdom. Eleven studies examined construct validity; nine assessed predictive validity. The EyeSi simulator dominated the evidence base, although MicroVisTouch and an in-house platform were also evaluated.

• Construct validity
  • Most studies demonstrated that the EyeSi reliably distinguished between novices and experienced surgeons on tasks such as capsulorhexis, phacoemulsification, forceps handling and anti-tremor exercises. Only one study failed to detect skill-level differences across all tested modules.
  • Limited but positive construct-validity data were reported for MicroVisTouch and the custom platform.
• Predictive validity
  • Simulator-trained residents generally achieved higher Objective Structured Assessment of Cataract Surgical Skills (OSACSS) scores, shorter phacoemulsification times and fewer complications (e.g. posterior-capsule tears) than simulator-naïve peers.
  • A small number of studies produced mixed results; one found no significant advantage over wet-lab training for capsulorhexis performance.

Overall, the balance of evidence supports both construct and predictive validity for the EyeSi simulator, with encouraging preliminary findings for alternative VR systems.

Methodology

Searches of PubMed and Scopus up to 1 May 2019 identified English-language studies evaluating trainee performance on VR cataract-surgery simulators and reporting construct or predictive validity outcomes. Two reviewers independently screened records against prespecified criteria, extracted data and synthesised findings narratively.

Applicability / external validity

Studies spanned a variety of countries and training settings, suggesting broad applicability; however, most evidence concerns the EyeSi, so generalisation to other simulators is limited. Further research on alternative VR platforms would strengthen external validity.

Geographic focus

No geographical restrictions were applied; included studies originated from North and South America, Europe, Asia and Oceania.

Summary of quality assessment

Confidence in the review’s conclusions is low. Searches were restricted to two databases and to English publications; a list of excluded studies was not provided; reference-list and expert searches were not reported. The number of reviewers involved in data extraction was unspecified, and risk-of-bias assessment of included studies was not described.