Learning curve of two common Descemet membrane endothelial keratoplasty graft preparation techniques

Objective: To compare the learning curve of two Descemet membrane endothelial keratoplast (DMEK) graft preparation techniques. Design: Experimental study. Participants: Twenty paired donor corneoscleral rims. Methods: The corneas were randomized to DMEK peeling using the peripheral blunt dissection technique (n = 10) or the modified submerged cornea using backgrounds away (mSCUBA) technique (n = 10). Outcome measures included graft peeling time, surgeon's peeling difficulty grading (on a scale of 1 to 10, 1 being the easiest and 10 the hardest), number of tears, and percentage tissue loss. Results: Average graft peeling time using the peripheral blunt dissection technique and the mSCUBA technique was 15.75 +/- 4.01 minutes and 8.43 +/- 3.26 minutes, respectively (p < 0.0005). The first 3 grafts' average peeling time was longer than the last 7 grafts: 19.14 +/- 2.40 versus 14.21 +/- 3.50 minutes in the peripheral blunt dissection technique (p = 0.06) and 12.36 +/- 3.76 versus 6.67 +/- 0.49 minutes in the mSCUBA technique (p = 0.016). In the latter, there were significantly fewer radial tears compared to the former: 1.5 +/- 1.0 and 3.1 +/- 1.9, respectively (p = 0.049). No tissue loss was noted in the mSCUBA group compared to one (10%) in the peripheral blunt dissection group. The average difficulty grading for the mSCUBA was significantly lower than the peripheral blunt dissection technique: 3.3 +/- 1.9 and 5.8 +/- 1.6, respectively (p = 0.024). Conclusions: Our study suggests a shorter learning curve with the mSCUBA technique for DMEK graft preparation, with shorter peeling time and fewer complications in comparison to the peripheral blunt dissection technique.