Evaluation of tests for microalbuminuria screening in patients with diabetes

Background. The first step in the diagnosis of diabetic nephropathy is to measure albumin in a spot urine sample. The aim of this study was to assess the accuracy of urinary albumin concentration (UAC), urinary albumin-to-creatinine ratio (UACR), and the Micral-Test II in a random urine specimen (RUS) for microalbuminuria screening in diabetes mellitus. Methods. Two hundred and seventy-eight patients collected 24 h timed urine specimens followed by RUS. Albumin (immunoturbidimetry) and creatinine were measured in protein-negative (Combur-Test) urine samples. Samples were classified as normoalbuminuric [24 h urinary albumin excretion rate (UAER) < 20 mu g/min; n = 189] and microalbuminuric (UAER =20-199 mu g/min; n = 89). Micral-Test II readings were performed in 130 RUS. Receiver operating characteristics (ROC) curves were constructed using UAER as the reference standard. Results. The areas under the ROC curves were similar for UAC (0.934 +/- 0.032) and UACR (0.920 +/- 0.035; P = 0.626), but the Micral-Test II had lower accuracy to diagnose microalbuminuria (area = 0.846 +/- 0.047) than UAC (P = 0.014). The first cutoff point with 100% sensitivity for UAC was 14.4 mg/l (specificity = 77.2%), and 15.7 mg/g for UACR (specificity = 73.0%). Concerning the Micral-Test II, sensitivity and specificity for the 20 mg/l cutoff point were 90.0 and 46.0%, respectively. The agreement between UAER and the Micral-Test II for microalbuminuria diagnosis was 55.8% (kappa = 0.22; P < 0.001). The cost of diagnosing microalbuminuria was U$1.74 (UAC), U$2.00 (UACR) and U$4.09 (Micral-Test II) per patient. Conclusions. Measurement of UAC in a RUS was the best choice for the diagnosis screening of microalbuminuria in diabetic patients, considering cost and accuracy.