Video Grading of Descemet Membrane Endothelial Keratoplasty Surgery to Identify Surgeon Risk Factors for Graft Detachment and Rebubbling: A Post Hoc Observational Analysis of the Advanced Visualization In Corneal Surgery Evaluation Trial

The aim of this study was to explore video-graded intraoperative risk factors for graft detachment (GD) and rebubbling in Descemet membrane endothelial keratoplasty surgery. METHODS: A post hoc analysis of 65 eyes of 65 pseudophakic subjects with Fuchs endothelial dystrophy that underwent Descemet membrane endothelial keratoplasty surgery as part of the Advanced Visualization In Corneal Surgery Evaluation trial. All surgical recordings were assessed by 2 graders using a structured assessment form. A multinominal regression was performed to estimate the independent effect of video-graded intraoperative factors on the incidence of GD and rebubbling. Secondary outcomes are corrected distance visual acuity and endothelial cell density. RESULTS: In total, 33 GDs were recorded, of which 17 required rebubbling. No significant predictors for GD or rebubbling were identified. However, the results revealed 2 clinically relevant patterns. An unfavorable graft configuration (ie, wrinkled, tight scroll, or taco-shaped) and a gas-bubble size smaller than the graft diameter were associated with an increased risk of GD [odds ratio (OR) 2.5 and OR 2.26, respectively] and rebubbling (OR 2.0 and OR 2.60, respectively). Inversely, a larger gas-bubble size was associated with a reduced risk of GD (OR 0.37) and rebubbling (OR 0.36). At 3 and 6 months postoperatively, corrected distance visual acuity was poorer in subjects requiring a rebubbling and endothelial cell density loss was higher in subjects with a partial GD. CONCLUSIONS: Our analysis revealed that the gas-bubble size and graft shape/geometry seem to be relevant clinical factors for GD and rebubbling, whereas descemetorhexis difficulty, degree of graft manipulation, graft overlap, and surgical iridectomy were not associated with an increased risk.