Validation in a Single-Center Cohort of Existing Predictive Models for Delayed Graft Function After Kidney Transplantation

• DOI: 10.12659/aot.894034
• Publication Date: 2015-09-15T10:18:57Z
• Authors: Alexander Decruye...; Philippe Decruyen...; Patrick Peeters; Frank Vermassen
• Keywords: Single Center

BACKGROUND:Kidney transplantation is the preferred treatment for patients with end-stage renal disease. Delayed graft function (DGF) a common complication and associated short- long-term outcomes. Several predictive models DGF have been developed. MATERIAL AND METHODS:497 kidney transplantations from deceased donors at our center between 2005–2011 are included. Firstly, accuracy of existing proposed by Irish et al. (M1), Jeldres (M2), Chapal (M3), Zaza (M4) was assessed. Secondly, were aggregated into meta-model (MM) using stacked regressions. Finally, association 47 risk factors studied in ­cohort-fitted model (CFM) logistic regression. The all assessed area under receiver operating characteristic curve (AUROC) Hosmer-Lemeshow test. RESULTS:M1, M2, M3, M4, MM, CFM AUROCs 0.78, 0.65, 0.59, 0.67, 0.82, respectively. M1 (P=0.018), M2 (P<0.001), M3 M4 (P<0.001) overestimate risk. MM (P=0.255) (P=0.836) well calibrated. Donor subtype recipient cardiac donor serum creatinine age (P=0.006), duration dialysis (P=0.02), BMI (P=0.008), (P=0.041), preoperative diastolic blood pressure (P=0.049) CFM. CONCLUSIONS:Four cohort low incidence DGF. We identified 2 parameters that not included previous models: pressure.