Quantification of pyuria by two methods correlation and interobserver agreement.

The aim of this study was to correlate the results of two methods of quantifying pyuria and to compare the diagnostic reliability of the two methods in terms of interobserver agreement. Three independent observers counted leucocytes in 56 urine samples by using high-power-field (HPF) microscopy on centrifuged urine sediment and by using a Bürker haemocytometer on uncentrifuged urine. The commonly accepted diagnostic level of ten leucocytes/mm3 in the haemocytometer was found to be equivalent to two leucocytes/HPF. This agrees with other reports but contrasts with the usual notion of a diagnostic level of five to ten leucocytes/HPF, and implies that urinary tract infection in the presence of symptoms should not be ruled out when less than five leucocytes/HPF are found. Kappa statistics were used to evaluate the interobserver agreement. The observers agreed on the diagnosis in 96% (K = 0.92) of the urines when the haemocytometer was used, and in 93-96% (K = 0.81-K = 0.91) when leucocytes were counted per HPF. The differences in Kappa values were not statistically significant. Our material indicates that the method of quantifying pyuria in centrifuged urine sediment by HPF microscopy may be used as the routine method in general practice.