INTRODUCTION: Urine analysis is affected by the great biological variability for most analytes. Microalbuminuria determination usually requires an overnight or 24-hour urine collection, which can be inconvenient because of its heavy dependence on patient compliance and is frequently unreliable because of errors in collecting the required urine sample. OBJECTIVE: In this study, we compared the albumin urinary excretion in overnight urine collection and random samples. MATERIALS AND METHODS: Ninety-seven patients were evaluated; overnight and second morning samples were analyzed. Urinary albumin and creatinine concentrations were determined by nephelometry (BNII - Dade) and by modified Jaffé method (Hitachi 917 - Roche), respectively. Simple linear regression (least square method) was used to evaluate the correlation between both methods. The accuracy of the albumin to creatinine ratio (ACR) was assessed considering the decision level of 30mg/g. RESULTS: 33 patients had abnormal urinary albumin excretion (UAE > 20µg/min). There was a significant correlation between UAE and ACR (r = 0.902, p < 0.05). The ACR accuracy was 90%, with sensitivity of 88% and specificity of 90%. CONCLUSION: This study shows the diagnostic value of the albumin/creatinine ratio in detecting microalbuminuria.
Albumin/creatinine ratio; Microalbuminuria; Random urine sample