Evaluation of a dipstick test for microalbuminuria in three different clinical settings, including the correlation with urinary albumin excretion rate.

UNLABELLED: A dipstick test for microalbuminuria was compared with urinary albumin excretion and urinary albumin concentration. Elevated urinary albumin excretion was defined as > 30 mg/24 h, elevated urinary albumin concentration in early morning urine as = > 20 mg/l and sticks = > 20 mg/l were considered positive. 1,071 samples with urinary albumin concentration 20-200 mg/l from 258 diabetic subjects were evaluated in three settings: I. 3 trained nurses testing samples from day-clinic diabetic patients. Sticks v.s. urinary albumin excretion sensitivity 86%, specificity 97%, predictive value of negative test 97%, correlation coefficient 0.79. Values for same setting but v.s. urinary albumin concentration were almost identical. II. 1 laboratory technician testing not-hospitalised diabetic patients: Sticks v.s. urinary albumin concentration: sensitivity 91%, specificity 85%, predictive value of negative test 95%, correlation coefficient 0.82. III. 58 general practitioners testing not-hospitalized diabetic patients: Sticks v.s. urinary albumin concentration: sensitivity 66%, specificity 92%, predictive value of negative test 83%, correlation coefficient 0.70.
CONCLUSIONS: In the hands of trained nurses and laboratory technician the Micral-Test showed good correlation with urinary albumin excretion and urinary albumin concentration for day-clinic and not-hospitalized patients and can be recommended as a screening tool. General practitioners obtained a lower sensitivity probably due to lack of experience and incorrect handling of the sticks leading to systematical errors. Training in the use of the stick must be emphasized, since under such circumstances the results are satisfactory.