*Purpose: Tools for indicating which patients are most at risk for complications and death after transplant listing would allow for a rational approach to targeted waitlist management.

*Methods: Following IRB approval, demographics, frailty metrics, pedometer, treadmill ability (METS), troponin, and BNP were collected on 314 patients listed for kidney transplant between July 2015 and March 2020 and three subsequent listing update visits. Frailty metrics included handgrip, chair sit/stand, 8 foot up-and-go, sit/reach, and exhaustion questions. Calibrated pedometer activity was recorded over several days and measured as steps/time. After patients with missing data were eliminated, univariate and multivariate logistic regression models were fit to assess the association of measures with delisting for death or medical complications in 283 remaining patients. Model fits of the predictor subsets for each outcome were compared using the Akaike information criterion (AIC). Additional logistic modeling was conducted on 84 patients with multiple visits and included change in metrics over time to test if progression and time between visits was predictive of waitlist outcomes. AIC was used to identify optimal covariate sets considered in our modeling. We assessed the significance of individual predictors at the alpha = 0.05 level.

*Results: Mean age was 59.9 years, 95% male, 57% White, 30% Black, 9.5% Hispanic. Mean BMI was 29.6. Diabetes was the primary cause of ESRD (45%), followed by HTN (15%), FSGS (9%), and ADPKD (7.6%). Reasons for delisting were medical (DLM, n=43), death on list (DOL, n=20) and other (n=9). The mean (SD) time between first and last follow-up was 2.09 (0.84) years. BNP was the most consistent univariate predictor of DLM and DOL (both p<0.0001). Factors at initial listing significantly associated with DOL on multivariate analysis included higher BNP (p<0.0001) and lower steps/time (p=0.015). Lower METS and steps/time, and high BNP were associated with delisting for medical reasons on multivariate analysis (p<0.05). Considering change in metrics over time, declining steps/time was associated with DLM (p<0.0001) and DOL (p=0.0127). Worse performance on treadmill (METS) and chair sit/stand were associated with DLM (p=0.0086) and DOL (p=0.048), respectively. Increase in BNP values consistently predicted DLM (p<0.0001) and death at any time after listing (p<0.0001).

*Conclusions: BNP and pedometer activity at time of listing and changes in BNP over time were significantly associated with kidney transplant waitlist outcomes. Functional metrics and BNP screening can identify patients in need of more intense waitlist management.

« Back to 2021 American Transplant Congress