Background:Debate continues regarding the utility of including additional clinical variables to improve SRTR center-specific predicted outcomes. In this study we used a cohort of kidney recipients transplanted in our program. [bold}Results: There were 586 deceased adult kidney transplant procedures performed at our center between 1/1/2002 and 6/30/2008. All covariates included in the Cox proportional hazards model for 1 year patient suvival reported by SRTR on 1/13/2012 were included in our initial Cox model (SRTR Model). We then performed Cox models with each of 12 additional covariates and finally with all 12 additional covariates. Each Cox model predicted 1 year patient survival and number of graft losses, and Log likelihood are reported in the table below. Also, the Likelihood ratio test (tests whether the Log likelihood is significantly different than that for the SRTR Model), Harrell’s C discrimination, and the Goodness of fit test were performed for each Cox model and reported in the table below.

« Back to 2013 American Transplant Congress