From a multivariate log-linear analysis of 35,625 renal transplants between 1988 and 1991, center effects accounted for 28%, 45%, and 27% of all assignable variation in 3-month, 1-year, and 2-year outcomes, respectively. Although center variation dominated 22 other variables, most factors were relatively independent of transplant center (ie, a percent of factor variation due center less than 10%). Recipient race and health status were notable exceptions; both highly influenced by center affiliation. Centers also differed in the age mix of recipients and racial mix of donors in some epochs. Again, we found only extremely weak correlations among a center's 3-month, 1-year, and 2-year graft survival rates. 2. In order of 3-month accountability, the other important factors were PRA, donor age, recipient working status, year of transplant, HLA-A,B mismatching, previous transplant, donor's death, donor relationship, recipient race, body mass, recipient age, cold ischemia time, donor race, donor kidney mode (ie, left/right kidney), original disease, and HLA-DR mismatching. Regarding 1-year outcome, the important factors were recipient race, donor age, donor's death, donor relationship, HLA-A,B mismatching, previous transplant, and recipient sex. Finally at 2 years, the important factors were recipient race, donor age, year of transplant, donor relationship, recipient sex, working status, donor's death, recipient age, CMV status, body mass, and donor sex. 3. Body mass, donor kidney mode, and CMV status were novel factors in our own multifactorial analyses of the UNOS Registry file. An elevated body-mass index (> 30 kg/m2) had a negative impact on short- and long-term graft survival. Recipients receiving left kidneys had nominal improvement in 3-month graft survival, but no impact thereafter. Survival rates over the 4 combinations of donor/recipient CMV statuses, suggest that this covariate was principally long-term and donor related. 4. It is noteworthy that graft failures in the 2 most recent transplant years, 1990 and 1991, have shown both short- and long-term declines, breaking stationary patterns previously reported in this series on clinical transplants. 5. The transitory nature of most transplantation factors was confirmed in this study, implying that future multifactorial studies in renal transplantation must include some mechanism for varying risks.(ABSTRACT TRUNCATED AT 400 WORDS)
