Are the baseline chances of survival comparable between the candidates for kidney transplantation who actually receive a graft and those who never get one?

ABSTRACT

BACKGROUND: The superiority of kidney transplantation over dialysis for patient survival often is assessed by comparing the survival rate of candidates who get a graft to that of those on the waiting list who do not. This study tries to ascertain if the two groups are comparable in terms of their chances of surviving. METHODS: Of the 187 non-diabetic patients who entered our waiting list during 1998 and 1999 for first cadaveric kidney transplants, 81 received a graft and 106 did not. We compared the two groups for factors which could affect survival and that were present at the moment of acceptance on the list. As one of those factors was the clinical score quantifying health status, as given by the transplant team and rated from 1 (high risk) to 4 (very good), we assessed its reliability by evaluating the survival of the patients we transplanted between 1988 and 1996, grouped according to that score. RESULTS: Transplanted patients had been immunized less frequently (2 vs 13%; P=0.02), had a lower dialytic age (16.9+/-2.1 vs 22.9+/-2.1 months; P<0.05), and better clinical scores (2.9+/-0.1 vs 2.6+/-0.1; P<0.05). The two groups did not differ in age, gender, or the presence of single specific diseases. Logistic regression analysis confirmed the results of univariate analysis. The clinical score was a very strong predictor of patient survival, as the survival of patients transplanted from 1988 to 1996 progressively improved with better scores (P<0.0001). CONCLUSIONS: Transplanted patients actually differ from non-transplanted candidates with respect to various factors potentially affecting survival. The difference is highly relevant clinically, yet it is not easily detected when considering mainly the presence or absence of specific diseases. A global quantitative clinical parameter based on a thorough medical evaluation is required to identify differences.