ABSTRACT
The demonstration that microalbuminuria is predictive of overt diabetic nephropathy has created a demand for the routine measurement of urinary albumin in diabetic patients. We assessed the sensitivity, specificity, positive and negative predictive values of the conventional dipsticks for urinary protein (Ames Multistix, Bayer Diagnostic, Australia) as the screening test for microalbuminuria in diabetic patients compared with Micral-Test II (Boehringer Mannheim, Germany). Radioimmunoassay for albumin was taken as standard for comparison. With the urinary albumin concentration of 20 mg/L as a discriminating level of microalbuminuria, Micral-Test II had a sensitivity of 98.8 per cent and a specificity of 68.6 per cent whereas Ames Multistix had lower sensitivity but higher specificity. If urinary albumin concentration of 60 mg/L was used instead as a discriminating level of microalbuminuria, none of Ames Multistix by visual reading and only 5 of 32 (15.6%) of those by reflectance photometer had false negative results. By visual reading, the sensitivity of Ames Multistix was increased from 68.1 to 100 per cent with the drop in specificity from 85.7 to 50.2 per cent. On the other hand, the sensitivity was increased from 37.4 to 84.4 per cent but the specificity was maintained if reflectance photometer was used. In conclusion, Ames Multistix was less sensitive than Micral-Test II in detection of urinary albumin concentration above 20 mg/L. At higher urinary albumin concentration above 60 mg/L which indicates a clinically significant microalbuminuria, the sensitivity of Ames Multistix was increased to 100 per cent. Ames Multistix which is much less expensive than Micral-Test II, can be used as the screening test for significant microalbuminuria in clinical practice particularly in cases having financial problems.