Evaluation of dipstick urine studies using a scoring system based on test performance characteristics.

Abstract

A weighted scoring system based on the selection of four measures of the performance characteristics of a laboratory test--sensitivity, specificity, and false positive and false negative rates--was devised to provide a rationale basis for ranking and evaluating dipstick urine screening tests as a criterion for elimination of routine urine microscopic examination. Equal weight was assigned to each performance characteristic, and an arbitrarily chosen scoring scale of 100 was selected for simplicity. When a summation procedure was applied to urine studies available in the medical literature, three categories of author acceptability were found: accept: score greater than 80; conditional accept: score 70-80; reject: score less than 70. Studies incorporating leukocyte esterase (LE) were found to have the highest scores and the highest author acceptability, whereas studies lacking LE had the lowest scores and lowest author acceptability. This easy-to-use, weighted scoring system is proposed as a guide for deciding whether dipstick screening tests are acceptable as a way of eliminating routine microscopic examination.