Rapid learning curve assessment in an ex vivo training system for microincisional glaucoma surgery

Purpose: Microincisional glaucoma surgeons operate in a highly confined space, making it difficult to learn by observation or assistance alone. We hypothesized that an ex vivo model would allow for better refinement of technique, quantification of progress, and computation of a learning curve.

Methods: Seven trainees without angle surgery experience performed nine ab interno trabeculectomies in pig eyes (n=63) after preparing with training slides and videos. Trainees placed the eyes on a tiltable mannequin head, visualized the trabecular meshwork gonioscopically through an ophthalmic microscope, and removed it by trabectome-mediated ablation. An expert surgeon observed, guided, and rated the procedure using an Operating Room Score (ORS). The extent of accessed outflow beds was estimated with canalograms using fluorescent microspheres. Data was fit using mixed effect models.

Results: ORS reached a half-maximum on an asymptote after only 2.5 eyes. Surgical time decreased by 0.9 minutes per eye in linear fashion. The ablation arc followed an asymptotic function with a half-maximum inflection point after 5.3 eyes. The mean ablation arc improved from 73 to 135 degrees. Canalograms revealed that this progress did not correlate well with improvement in outflow instead suggesting that about 30 eyes are needed for true mastery.

Conclusion: This inexpensive pig eye model provides a safe and effective training model for ab interno trabeculectomy and allows for quantification of outcomes. Trainees without prior angle surgery experience improved quickly. Actual outflow improvements progressed at a slower rate, which serves as a reminder to remain humbly committed to training.