Background: Relative survival estimates, used in population-based monitoring of survival of patients with cancer, accounts for differences in the background mortality (obtained from general population life tables) of patient populations to improve the comparability of survival estimates. We apply and demonstrate the benefits of this measure in solid organ transplantation for the first time.
Methods: Data of deceased and living donor kidney transplant recipients aged 0 to 17, 18 to 39, 40 to 49, 50 to 59, and 60+ registered in the United Network for Organ Sharing/Organ Procurement and Transplantation Network data set and age, sex, calendar period, and race-specific US life table for 2006 were used. We calculated absolute 5- and 10-year patient survival by period analysis to obtain up-to-date estimates, whereas the expected survival of the exactly matched group of the general population was derived from life table data. Relative survival was calculated as the ratio of observed and expected survival.
Results: For 5-year absolute survival, an age gradient of 23.7% and 14.0% units between the youngest and the oldest age group was found for recipients of kidneys from deceased and living donors, respectively. For relative survival, the age gradient was decreased substantially, to 15.0% and 5.0% units, respectively. For 10-year survival, the nominal effect of accounting for background mortality was even larger.
Conclusions: Absolute survival estimates fail to account for potential differences in background mortality of various patient groups in general and may substantially overestimate the survival gap of older patients. Relative survival may be a useful additional measure in registry-based transplant patient outcome monitoring.