Update: center effects

Abstract

1. From a multivariate log-linear analysis of 30,274 renal transplants between 1985 and 1989, center effects accounted for nearly 30% of all assignable variation in 1-year outcome, dominating 15 other factors analyzed. In order of accountability in 1-year graft outcome, the other important factors were donor relationship, CsA usage, donor age, graft number, highest pretransplant antibody, recipient race, HLA-A, B, and DR mismatches, cold ischemia time, donor sex, donor race, and pretransplant transfusions. Original disease, transplant year, recipient sex and age were not significant factors influencing 1-year graft survival. 2. The additive nature of these transplantation factors on the logit scale was confirmed in this analysis, implying that univariate analyses in renal transplantation are not necessarily improper. 3. There was no correlation between center effect and center size as measured by the number of renal transplants per year. Therefore, renal transplantation at small centers will not necessarily produce poorer graft function. 4. Similarly, we found only weak correlation between a center's 1-year graft survival and their patients' graft half-lives beyond 1-year. This suggests that factors determining center success in early graft survival differ from those that influence long-term success. 5. Graft survival rates among patients with and without rejection episodes prior to hospital discharge vary substantially among centers. Centers with poor 1-year graft survival demonstrated a significantly larger variation in graft survival between their patients with and without rejection than did excellent, good, and fair centers. This indicates that posttransplant patient maintenance is another factor influencing center effects on renal graft function.