Value of urine microscopy in predicting histological changes in the kidney: double blind comparison.

Fresh, first morning specimens of urine from 22 consecutive patients were examined by quantitative microscopy on the morning of renal biopsy; the renal biopsy samples were evaluated "blindly." Five patients showed no abnormality in the biopsy samples but eight had minimal, one mild, six moderate, and two severe histological changes. Comparison of the results of quantitative microscopy of urine with the presence or absence of histological evidence of disease showed that sensitivity was 88%, specificity 83%, accuracy 86%, positive predictive value 93%, and negative predictive value 71%. When combined with microscopy of a second urine specimen these values were 100%, 50%, 87%, 85%, and 100% respectively. There was a significant relation between number of casts and severity of the histological changes (p less than 0.01). Comparison of renal functional abnormalities with histological findings gave values of 64%, 100%, 73%, 100%, and 50% respectively. It is concluded that quantitative microscopy of the first morning specimen of urine is a sensitive test with high predictive value for the presence or absence of renal disease. If no casts are detected in two early morning specimens the likelihood of finding anything more than minimal changes in a biopsy sample is virtually zero.