Univariate Analyses of the Impact of Donor and Recipient Related Parameters on Early Outcome of Kidney Transplantation in Germany

Abstract

Introduction: The aim of kidney transplantation is to provide renal replacement therapy by functioning grafts at low recipient mortality. Thereby the rate of delayed graft function should be low (DGF definition: post-operative dialysis>1). In this analysis, the impact of several donor (D) and recipient (R) related parameters on in-hospital mortality (M), death-censored graft survival (GS) and the incidence of DGF was investigated. Methods: Data concerning kidney donation (provided by Deutsche Stiftung Organtransplantation) and kidney transplantation (provided by BQS Institute for Quality and Patient Safety) were merged. In the resulting database, 4471 anonymized datasets of deceased D kidneys donated and transplanted in Germany into adult R during 2006 to 2008 are available. R with combined transplantations were excluded. To obtain independent observation units necessary for statistical analyses, for each R only the first (if more than one) transplantation in the study period as well as only one randomly selected kidney per D were considered. For later model building and validation purposes, data were randomly subdivided into a learning (n = 1621, 66%) and a validation sample (n = 810, 33%). Based on the learning sample, the impact of several D and R related parameters on M, GS and DGF was analysed by means of univariate logistic regression at a significance level of 5% (OR; p-value). Results: In the learning sample 1621 R were analysed with a mean age of 53.98y (SD=12.58), (62.6% m, 37.4% f). The mean age of D was 52.59y (SD=16.74), (52.6% m, 47.4% f), mean value of donor Serum-creatinine (S-creaa) on admission and before organ retrieval (S-crear) was 83.55mmol/l (SD=42.71) and 96.47mmol/l (SD=68.73), respectively. In-hospital mortality (M) was 1.4%, the incidence of delayed graft function (DGF) 28.1% and death-censored graft survival (GS) 94.3%. In-hospital mortality is significantly influenced by R age (1.069; 0.004) and to a lesser degree by D age (1.032; 0,040). DGF was determined by D age (1.010; 0.005) and D S-crea, particularly by S-crear before organ retrieval (1.003; < 0.001), but also by cold ischemia time (1.001; < 0.001). Concerning recipient-related factors, immunised patients (1.006; 0.007) seem to have a higher rate of DGF. Graft survival (GS), however, is influenced by D age (0.984; 0.019) and D S-crea (0.998; 0.049). Interestingly, the use of arterenol (1.676; 0.034) and steroids (1.739; 0.030) in the D before explantation of the organs had a positive impact on GS. Conclusion: In-hospital mortality after kidney transplantation was low and related to R and D age. The effect of D age might disappear in a multiple logistic regression model, as due to the European-Senior-Programme in Germany a strong relation between R and D age exists. Graft survival was determined by donor factors, such as D age and S-crea. Moreover our results indicate, that the incidence of DGF may be reduced by shortening cold ischemia time. These preliminary results are based on univariate analyses. Multivariate analysis considering the above described and several other risk factors simultaneously is currently performed and needed for further conclusions.