Univariate and multivariate analyses of cadaver kidney graft survival data.

1. Clinical data on 6632 first cadaver transplants performed since January 1983 were analyzed by multivariate (Cox regression) and univariate (life-table) methods. 2. Cox regression analysis showed blood transfusion, recipient's race, HLA mismatch, highest antibody, warm ischemia time, and cyclosporine treatment as significant factors affecting graft survival. 3. A comparison of survival curves predicted by Cox regression analysis and that by the life-table method showed close agreement between the two methods. 4. Cyclosporine was one of the most significant variables affecting graft survival. There was a 13% overall increase in one-year graft survival due to cyclosporine treatment. However, other factors were still significant; they have not become obsolete in the cyclosporine era.