Thesis (M.Sc.)--Chulalongkorn University, 2006

Objective: Acute renal failure (ARF) in patients undergoing coronary angiography diagnosed by daily serum creatinine monitoring is inconvenient and under-investigated. It is interesting whether ARF can be easily diagnosed by urinary cystatin C to creatinine ratio; UCCR. Design: Prospective cross-sectional study (Diagnostic test) Setting: Department of Medicine, Phramongkutklao Hospital, Bangkok, Thailand Method: Patients with GFR 15-59 ml/min/1.73 sq.m. scheduled for coronary angiography were enrolled. All eligible patients were studied for baseline characteristics, serum creatinine, serum cystatin C, urine creatinine and urine cystatin C at baseline, 24 and 48 hours after procedure. ARF was defined as serum creatinine rising [is more than or equal to] 0.5 mg/dl or >= 25% from baseline. Results: One hundred and twenty two patients were enrolled, 115 patients completed data for analysis. ARF developed in 12 patients (10.4%). From ROC, the best diagnostic test was the maximum value of UCCR within 48 hours post procedure (AUC=0.63; 95%CI 0.46-0.80) with sensitivity of 92% and specificity of 28% at cut-off value >= 0.07*10[superscript-3]. Likelihood ratio of UCCR at level of 0-0.3, 0.31-0.5, > 0.5 (*10[superscript-3]) were 0.73, 1.34, 1.98 respectively. If urine cystatin C alone was evaluated as a diagnostic tool, the best diagnostic test was the percent change of urine cystatin C at 24 hours post procedure from baseline (AUC=0.81; 95%CI 0.67-0.95) with sensitivity of 70% and specificity of 67% at cut-off value >= 3%. Likelihood ratio of this test at level of [is less than or equal to] 0, 0.1-100, 101-200, > 200% were 0.45, 0.82, 3.64, 9.10 respectively. Conclusion: Urine cystatin C and urinary cystatin C to creatinine ratio have only fair usefulness for diagnosis of acute renal failure in patients undergoing coronary angiography.