Multifactorial analysis of the outcome of 6430 cadaver kidney grafts

A total of 6430 cadaver kidney grafts performed within the network of France-Transplant between 1 January 1978 and 1 January 1989 were analyzed. Each case was examined comprehensively in regard to 12 variables. A multifactorial analysis (Cox regression) was used to determine the degree of association between each covariate and the outcome of the graft. The results were evaluated by calculating relative risks of graft failure for each variable. A total of seven covariates appeared to influence graft survival significantly: the period of transplantation (P = 10(-8)), retransplantations (P = 0.003), age and sex of the donor (P = 0.003 and 0.009 respectively), duration of pretransplant dialysis (P = 0.03), pretransplant sensitization to HLA antigens (P = 0.05), and matching for HLA-A, -B, and -DR loci (P = 0.03). This last parameter has previously been reported as influencing the outcome of the graft in seven out of eight international studies carried out using similar methodology.