RESUMO
Objective The aim of this study was to evaluate of adilty of two acute renal failure-specific scoring systenms (the classification by Bellome et al and the AKIN criteria) for predicting hospital mortality after cardiac surgery in adult patients. Methods Between October 1 st 2006 to Decemjber 31 st 2006, 509 adult patients who ungerwent coronary artery bypass grafting (CABG) and/ or valve operation were enrolled in this study. The medical data collection included gender, age, types of operation, perioperative he- modynamic parameters, urine output, biochemical parameters and outcome. Renal function was assessed daily according to the classi- ficatinn by Bellomo and the AKIN criteria, respectively. As references, Acure Physiology and Chronic Health Evaluation(APACHE) Ⅱ and Sepsis-related Organ Failure Assessment (SOFA) score were also calculated. Resuits Three hundred and forty-one patients were male (67.0%), and 168 were female (33.0%), mean age was (56.2±12.0) years old. Tnree hundred and nine patieats un- derwent CABG, 182 underwent valve operation and 18 underwent CABG plus valve operation, Mean duration of ventilation support was (20.4±17.7) houra, and the ICU stay was (1.4±1.0) days. Postoperative hospital stay was (13.8±9.1) days. According to the classification by Bellomo., the highest in-hospital mortality was 52.9% in ARFS group. Mahiplicatinn of in-hospital morality rate was abserved (X2 for trend, P<0.01) in 0.4% (non-ARF), 1.2% (stage 1), 12.0% (stal～ 2) and 32.4% (stage 3) of pa- tients based on the AKIN criteria. By applying the area under the receiver operating characteristic ourve, the classification by Bellomo and the AKIN criteria had good discriminative power. Furthering, multivariate logistic regression analysis verified that the Odds Ratio of the AKIN criteria was 5.478 (P =0.028, 95% Confidence Interval 1.027- 24.856), after adjusting for gender and age. Con- clusion Analytical data confinned good discriminative power of both the AKIN criteria and the classification by Bellomo for predicting hospital mortality of adult postoperative patient with ARF.