Assessing the Learning Curve for DMEK Using Post-Procedural Clinical Outcomes : Comparison of Four Different Surgeons during Two Different Periods

Abstract

Purpose: Evaluating the learning curve of individual surgeons for Descemet Membrane Endothelial Keratoplasty (DMEK) and Triple-DMEK and assessing outcome with experience. Methods: The first 41 and the last 41 surgeries of each of the four surgeons were retrospectively included. Surgery duration and graft preparation time were recorded. Corrected distance visual acuity (CDVA, logMAR) and central corneal thickness (CCT, µm) were collected preoperatively after 6 and 12 months, as well as postoperative complications, e.g., re-bubbling or repeat penetrating keratoplasty. Results: Surgical duration for Triple-DMEK and DMEK decreased significantly by 21 min and 14 min between the two periods (p < 0.001; p < 0.001). Graft preparation time decreased significantly from 13.3 ± 5.2 min (95%CI 12.8–14.3) in period 1 to 10.7 ± 4.8 min (95%CI 10.2–11.4) in period 2 (p = 0.002). The postoperative changes in CDVA and CCT over both periods were not significant (p = 0.900; p = 0.263). The re-bubbling rate decreased significantly from 51.2% in period 1 to 26.2% in period 2 (p < 0.001). The repeat penetrating keratoplasty (PKP) was 7.3% in period 1 and 3.7% in period 2 (p = 0.146). Re-DMEK was necessary in 6.1% in period 1 and 4.9% in period 2 (p = 0.535). Several parameters showed significant differences between the surgeons in both periods (surgical duration: period 1: p < 0.001, period 2 p < 0.001; graft preparation: period 1: p < 0.001, period 2 p < 0.001). Conclusion: Significant decrease in surgery duration, graft preparation time, and the re-bubbling rate can be attributed to gained individual experience.