To determine the incremental gains in graft and patient survival under a risk-based, deceased donor kidney allocation compared with the current Australian algorithm.Methods
Risk-based matching algorithms were applied to first graft, kidney only recipients (n = 7513) transplanted in Australia between 1994 and 2013. Probabilistic models were used to compare the waiting time, life, and QALYs and graft years between the 8 risk-based allocation strategies against current practice.Results
Compared with current practice, Kidney Donor Risk Index-Estimated Posttransplant Survival matching of the lowest 20% of scores reduced median waiting time by 0.64 years (95% confidence interval [CI], 0.52-0.73) for recipients aged 30 years or younger, but increased waiting time by 0.94 years (95% CI, 0.79-1.09) for recipients older than 60 years. Among all age groups, the greatest gains occurred if Kidney Donor Risk Index-Estimated Posttransplant Survival matching of the lowest 30% of scores was used, incurring a median overall gain of 0.63 (95% CI, 0.03-1.25) life years and 0.78 (95% CI, 0.30-1.26) graft years compared with the current practice. A median gain in survival of 1.91 years for younger recipients (aged 30-45 years) was offset by a median reduction in survival (by 0.95 life years) among the older recipients. Prioritization of lower-quality donor kidneys for older candidates reduced the waiting time for recipients older than 45 years, but no changes in graft and patient survivals were observed.Conclusions
Risk-based matching engendered a moderate, overall increase in graft and patient survivals, accrued through benefits for recipients 45 years or younger but disadvantage to recipients older than 60 years.