Center effect in the UNOS Renal Transplant Registry.

Based upon univariate and multivariate analyses of transplant center effects: 1. Among the 115 centers selected for these analyses, there was no correlation between 6-month graft survival rates and half-lives projected for grafts surviving 6 months. 2. There was no significant center effect for living-related donor transplantations. 3. Centers that had the poorest short- and long-term survival performed more cadaveric transplants with patients in poor health at the time of transplant, more often transplanted a recipient other than the one originally identified for a particular kidney, and more often transplanted kidneys with prolonged cold ischemia, and kidneys from pediatric (age 0-11) donors, older (ages over 50) donors, Black donors, or donors who died with cerebrovascular accidents. 4. After adjusting for 17 potentially confounding variables, the difference between the best and worst center groups was the second most detrimental factor (following broad sensitization) in the first 6 months and was the factor ranked ninth in long-term survival. 5. The transplant year has emerged as a significant factor in long-term survival. This suggests that the late loss rate for transplants performed after 1988 may be diminishing. 6. The health status of the patient at the time of transplant was the dominant factor affecting long-term survival. 7. The choice of outcome variable and the selection criteria used in center classification affect the magnitude of the center effect and its relationship to other significant variables that influence graft outcome.