METHODS: This is a retrospective analysis of the United Network for Organ Sharing registry data of LT recipients from January 1, 2000, to December 31, 2021. Outcomes analysis was performed using Cox proportional model for all-cause mortality and graft failure. Confounding was reduced by coarsened exact matching causal inference analysis.

METHODS: MEDLINE, Cochrane, and Embase databases were searched. Data on demographic information and transplant outcomes were extracted from included studies. Meta-analyses were performed, and risk ratios (RR) were estimated to compare transplant outcomes from uDCD to cDCD.

RESULTS: Nine cohort studies were included, from 2178 uDCD kidney transplants. There was a moderate degree of bias, as 4 studies did not account for potential confounding factors. The median incidence of primary nonfunction in uDCD was 12.3% versus 5.7% for cDCD (RR, 1.85; 95% confidence intervals, 1.06-3.23; P = 0.03, I2 = 75). The median rate of delayed graft function was 65.1% for uDCD and 52.0% for cDCD. The median 1-y graft survival for uDCD was 82.7% compared with 87.5% for cDCD (RR, 1.43; 95% confidence intervals, 1.02-2.01; P = 0.04; I2 = 71%). The median 5-y graft survival for uDCD and cDCD was 70% each. Notably, the use of normothermic regional perfusion improved primary nonfunction rates in uDCD grafts.