*Purpose: Desensitization protocols for HLA-incompatible living donor kidney transplantation (ILDKT) vary across centers. The impact of these variations, as well as other practice variations, on ILDKT outcomes remain unknown. We sought to quantify center-level variation in patient and graft survival following ILDKT.

*Methods: Using a 25-center cohort augmented by linkage to SRTR, we studied 1,358 ILDKT recipients transplanted between 1997-2016 and compared patient and graft survival across centers. Between-center differences in patient and graft survival were compared with a multi-level Cox model using shared frailty, allowing each center to have a different underlying hazard of mortality and graft loss. Between-center variation was quantified using the median hazard ratio (MHR), which describes the median increase in the risk of mortality or graft loss when comparing a patient at a center with higher risk to a center with lower risk.

*Results: Only 5 centers (20%) had significantly better outcomes than average, and only 1 center (4%) had worse patient survival, and 2 centers (8%) had worse graft survival, than average (Figure). After adjusting for patient-level characteristics, only 9.9% of the differences in patient survival (p<0.01), and 9.2% of the differences in graft survival (p<0.01), were attributable to between-center variation. These translated to a MHR of 1.36 for patient mortality, and 1.34 for graft loss. For perspective, the magnitude of these between-center differences was similar to the effect of a 10-year increase in recipient age on mortality (adjusted hazard ratio: _1.281.37_1.46, p<0.001). There was no relationship between the following center-level characteristics and improved outcomes: transplanting a higher proportion of highly sensitized recipients, prior transplant recipients, pre-emptive recipients, racial minorities, or performing a higher IKT volume.

*Conclusions: Unlike most aspects of transplantation which have very high center-level variation, we found almost no center-level variation in ILDKT outcomes. Importantly, we found no center volume effect. Our findings support the continued practice of ILDKT across these diverse centers.

« Back to 2019 American Transplant Congress