Predictive ability of severity scores and outcomes for mortality in kidney transplant recipients with coronavirus disease 2019 admitted to the intensive care unit: results from a Brazilian single-center cohort study

Abstract Background: the predictive ability of severity scores for mortality in patients admitted to intensive care units is not well-known among kidney transplanted (KT) patients, especially those diagnosed with coronavirus disease 2019 (COVID-19). The purpose of the present study was to evaluate the predictive ability of severity scores for mortality in KT recipients. Methods: 51 KT recipients with COVID-19 diagnosis were enrolled. The performance of the SOFA, SAPS 3, and APACHE IV tools in predicting mortality after COVID-19 was compared by the area under the ROC curve (AUC-ROC) and univariate Cox regression analysis was performed. Results: The 90-day cumulative incidence of death was 63.4%. Only APACHE IV score differed between survivors and nonsurvivors: 91.2±18.3 vs. 106.5±26.3, P = 0.03. The AUC- ROC of APACHE IV for predicting death was 0.706 (P = 0.04) and 0.656 (P = 0.06) at 7 and 90 days, respectively. Receiving a kidney from a deceased donor (HR = 3.16; P = 0.03), troponin levels at admission (HR for each ng/mL = 1.001; P = 0.03), APACHE IV score (HR for each 1 point = 1.02; P = 0.01), mechanical ventilation (MV) requirement (HR = 3.04; P = 0.002) and vasopressor use on the first day after ICU admission (HR = 3.85; P