Exploring the association between the degree of pyuria and urinary tract infections

We aim to define the optimal white cell count threshold that correlates with a clinically significant urinary tract infection, as there is insufficient data exploring this in the adult population. We conducted a retrospective cohort study at the Royal Melbourne Hospital analyzing urine samples collected over 6 months in 2022. Urinary tract infection (UTI) was defined as the presence of symptoms (dysuria, urgency, frequency, flank pain, or loin to groin pain) and isolation of a uropathogen with colony counts greater than 10(7) CFU/L. The relationship between urinary white cell count, growth of uropathogen, and likelihood of UTI was estimated using locally weighted scatterplot smoothing. Of the 6,328 samples included, at a urinary white cell count of less than 10 per microliter, 38% grew a microorganism, while 7% of the total grew a uropathogen. For our sub-analysis part C, at the same WBC count, 2% of samples fulfilled our criteria for UTI. The optimal WBC range for identifying a UTI was 30-50 WBC/mu L, demonstrating the most pragmatic balance of sensitivity (92.3%-94.9%, 95% CI 85.9-98.1) and specificity (41.6-47.2, 95% CI 38.6%-50.3%) for a UTI. A lower-than-expected number of UTIs were confirmed in our study, likely due to inappropriate indications for culture collection and the types of urine specimens collected. However, the optimal white cell cutoff for identifying a UTI was higher than the defined pyuria cutoff of 10 WBC/mu L. Utilizing a higher urinary WBC cutoff could improve urine culture processing protocols.