Incidence of presumed iatrogenic graft failure in Descemet stripping automated endothelial keratoplasty.

Purpose: To evaluate presumed iatrogenic graft failure (PIGF) in Descemet stripping automated endothelial keratoplasty (DSAEK).

Methods: Deidentified data were collected retrospectively from the Illinois Eye Bank between April 2007 and May 2010. PIGF was defined as cases in which a repeat corneal transplant was performed <8 weeks after an initial DSAEK. Data collected for each case included days between initial DSAEK and regraft, regraft type (DSAEK vs. penetrating keratoplasty), precut versus surgeon-cut tissue, and number of DSAEKs distributed to individual surgeons. Ninety-three cases of PIGF were identified for a group of 46 surgeons who received 2504 corneas. Nine PIGF cases from 4 surgeons in cornea fellowships were analyzed separately. Individual surgeon failure rate and overall failure rate were calculated. Analysis of the overall failure rate was conducted without cases performed by surgeons who train fellows. Effect of surgical experience on PIGF was analyzed.

Results: Overall presumed iatrogenic DSAEK failure rate was 3.66% (84 PIGF cases divided by 2294 corneas). Within the group of 4 surgeons that trained fellows, the failure rate was 4.29% (9 cases PIGF divided by 210 corneas). Average time to regraft was 27.5 ± 17.4 days. Initial diagnoses included Fuchs dystrophy, pseudophakic bullous keratopathy, edema, and other disease states. Average donor and recipient ages were 54.3 ± 13.0 and 71.7 ± 11.3 years, respectively. Use of precut versus surgeon-cut tissue did not affect overall failure rates. Failure rates of less experienced surgeons (21.8% ± 10.8%) were higher than more experienced surgeons (1.5% ± 1.4%).

Conclusions: Overall, DSAEK failure rates improve as surgeons gain experience. Failure rates in cornea fellowship programs are not excessive.