Kidney transplant mortality relationships.
1. Patient mortality for adults was described by the sum of 3 components of risk: age-dependent risk; transplant-dependent risk, depending on time after first transplant; and failure-dependent risk, depending on time after failure. 2. The age-dependent risk increased exponentially at a rate about two-thirds the rate of increase in the general population. This risk was about twice as large for recipients of cadaver kidneys as for those who received grafts from HLA-identical siblings; the risk for recipients of parent donor grafts was intermediate. The risk for transplant recipients was much larger than that for the general population over most of the age range. 3. The transplant-dependent component decreased exponentially with the time after transplant and was reduced by half after about 2.25 years. The component was about 4 times as large for recipients of cadaveric kidneys as from HLA-identical sibling donors. Since patients studied were limited to those who survived at least 3 months after transplant, the transplant-dependent component of risk, as defined, does not represent the larger and more rapidly decreasing loss that occurs within a month or two after transplant. 4. The failure-dependent component was large and decreased exponentially with time after graft failure; the risk was reduced by half in about 7.5 months. For recipients of cadaveric organs, "excess" of death in the first year after failure was about 10%--slightly more than percent excess death as in the first year after transplant. In the case of diabetics, the excess deaths in the year after failure was about 25%. 5. Life expectancy at various ages at transplant was calculated from the estimates of risk of death. The maximum loss for males transplanted from HLA-identical siblings was about 10 years. The maximum occurred at age-at-transplant 20 years; life expectancy for patients transplanted at that age was about 42 years. Perhaps 5% to 10% of patients transplanted at age 20 from cadaveric donors might expect to survive into their 80's. 6. Survival analysis with allocation to graft failure of patient-death-with-functioning-graft fractionally (according to probability that the death was transplant related) provided a method of adjusting for patient age in studying graft survival. In younger patients, death was most likely attributable to the transplant, whereas in older patients, death was likely not attributable to the transplant. 7. Appropriate attribution of patient death to graft failure is important to the study of long-term graft survival, particularly among older patients.(ABSTRACT TRUNCATED AT 400 WORDS)