SURVIVAL ANALYSIS AND VISUAL OUTCOME IN A LARGE SERIES OF CORNEAL TRANSPLANTS IN INDIA

Aim/background-The public health significance of corneal transplantati on in dealing with corneal blindness in the developing world would dep end upon the survival rate of transplants. This study was done to anal yse the survival rate of corneal transplants in a large series in Indi a, and to evaluate the influence of various risk factors on transplant survival. Methods-The records of a series of 1725 cases of corneal tr ansplants carried out during 1987-95 at a tertiary eye care institutio n in India were reviewed. The Kaplan-Meier method was used to determin e 5 year survival rates of corneal transplants performed for the vario us categories of preoperative diagnosis. Multivariate Cox proportional hazards regression was used to assess how preoperative diagnosis, soc ioeconomic status, age, sex, vascularisation of host cornea, quality o f donor cornea, and training status of surgeon influenced transplant s urvival. The effect of these variables on visual outcome was assessed using multiple logistic regression. Results-The survival rates at 1, 2 , and 5 years for all corneal transplants performed for the first time in 1389 cases were 79.6% (95% confidence interval = 77.3-81.9%), 68.7 % (65.7-71.7%) and 46.5% (41.7-51.3%). The 5 year survival rate was hi ghest if the corneal transplant was done for keratoconus (95.1% (84.8- 100%)) and lowest if carried out for previous transplant failure (21.2 % (13.8-28.6%)). The relative risk of transplant failure was higher if the preoperative diagnosis was previous transplant failure (2.04 (1.6 2-2.55)), aphakic bullous keratopathy (1.78 (1.38-2.28)), corneal clou ding due to miscellaneous causes including congenital conditions and g laucoma (1.63 (1.21-2.19)), or adherent leucoma (1.11 (0.81-1.51)) tha n for the other preoperative diagnoses. Patients with lower socioecono mic status had higher relative risk of transplant failure (1.28 (1.16- 1.42)), as did patients <10 years of age (1.42 (1.23-1.64)). Higher re lative risk of transplant failure was associated with vascularisation of the host cornea before transplantation (1.15 (1.04-1.27)), and with the use of fair quality donor cornea for transplantation compared wit h excellent, very good, or good quality donor cornea (1.26 (1.06-1.52) ). Before corneal transplant 80.2% of the eyes were blind (visual acui ty <3/60), whereas at last follow up 41.8% eyes were blind. The odds o f having visual acuity >6/18 were higher if the transplant was done fo r keratoconus (9.99 (6.10-16.36)) or corneal dystrophies (1.77 (1.21-2 .58)) than for the other preoperative diagnoses. Conclusion-Reasonable success with corneal transplantation is possible in the developing wo rld if data from this part of the world regarding the different surviv al rates for the various preoperative diagnoses and the influence of r isk factors on transplant survival and visual outcome are taken into a ccount while determining priority for transplant cases in the present situation of limited availability of donor corneas.