We read with interest the article “Postoperative Risk Factors Influencing Corneal Graft Survival in the Singapore Corneal Transplant Study.” The authors provide valuable data on an Asian population with insight into various factors affecting the outcome of penetrating keratoplasty, especially the interaction between preoperative, intraoperative, and postoperative risk factors.
The authors studied a heterogeneous group of transplantation recipients that included a majority of optical transplantations as well as therapeutic and tectonic grafts. It is well known that the survival of therapeutic and tectonic grafts is poorer, and hence, more valuable information would have been obtained from a homogenous group of optical transplantations.
Because the mean follow-up period was not mentioned in this study, we presume that it was the same as the earlier study on the same population, 36.8 ± 35.5 months. The median and range ought to have been provided because the standard deviation is quite large. In addition, the Kaplan-Meier survival analysis does not depict the number of patients still in the study at different time points; hence, the number of patients with more than 10 years of follow-up is not known. This would be useful in determining the risk factors affecting graft survival over a prolonged period. The Kaplan-Meier survival analysis plots for risk factors other than allograft rejection and glaucoma have not been included.
Repeated grafts do adversely affect the graft survival ; thus, selection of the last graft in eyes with multiple grafts would have given more information than randomly selecting 1 graft per patient.
Although the authors found that preoperative inflammation, perforation, and recurrence of primary disease were significant risk factors for graft survival, no data were provided in the multivariate Cox regression analysis final model.
A closer look at Table 3 reveals that aphakic and pseudophakic bullous keratopathy has a very high hazard ratio predictive of corneal graft failure, but the authors have not elaborated on this. Because the duration of follow-up has not been provided, the hazard ratio has limited value because it is dependent on the time of follow-up.
In conclusion, although the authors have described valuable data on an Asian population, nonavailability of important data like duration of follow-up, raw data in various subgroups, and more survival analysis data involving risk factors other than glaucoma and allograft rejection leaves many questions unanswered.