A risk score as part of an evidence-based approach to the selection of corneal replacement surgery

Purpose: There is currently no standardized methodology that pen-nits preoperative prediction of the surgical technique or drug regimen for corneal grafting most likely to succeed for an individual patient. A score factor related to the chance of successful penetrating keratoplasty (PK) may assist in standardization of graft outcome reporting and evidence-based management.Methods: The Risk Score (RS) is based on factors shown to reduce the probability of 1-year survival of conventional PK in increments, using published Australian Corneal Graft Registry data. To asses its validity, Kaplan-Meier survival curves from 156 corneal replacement surgeries, stratified by preoperative RS, were used to derive the probability of survival at 1 year for donor PK, with or without systemic immunosuppression, and for synthetic PK with the AlphaCor (TM) artificial cornea. The probability of survival was then plotted against RS for each regimen, producing a family of curves.Results: The mean RS for immuno suppressed cases was higher than for nonimmunosuppressed cases. The AlphaCor group had the highest scores. AlphaCor survival is little affected by the factors that affect the risk of graft survival, whereas the probability of donor graft survival, immunosuppressed or not, is inversely related to RS.Conclusions: This study was designed to test a method that allows objective comparison of outcomes of different methods of corneal replacement for different risk categories. Analysis of this small dataset supports the validity of the concept and suggests that the Risk Score is a useful descriptive statistic. With a sufficient sample size, statistically significant evidence of the optimum management method for an individual patient with a particular risk profile could be determined such that a prospective patient's preoperative RS could be used to assist management decisions. In particular, data allowing separation of different immunosuppression regimes would be useful. Functional outcomes such as visual acuity would be more significant clinically than the survival end point used in this study and the system also could be adapted to different follow-up periods.