Abstract
Urine microscopy for detecting pus cells is a common investigation ordered in hospitalized general medical patients as part of routine care. A few previous studies have shown that sterile pyuria is not uncommon in this population. We studied the prevalence of pyuria among patients hospitalized with non-urinary tract infection (UTI) diagnosis in the medical wards. We excluded patients with asymptomatic bacteriuria (ASB). Pyuria was quantified in uncentrifuged urine using the chamber counting method, and ≥ 10 pus cells per mm (3) was considered significant. We also compared this method with the commonly used but less accurate method of counting pus cells/high power field using centrifuged urine (routine method). We studied 196 patients; 113 (57.7%) were males. Most (175[89.3%]) patients were hospitalized for an infection. We found that 18.4% of the study group had sterile pyuria, and it was strongly associated with the presence of concomitant microscopic hematuria (unadjusted odds ratio, 3.74 [1.65 to 8.50]; P=0.002). We found no association of pyuria with female gender, diabetes, acute kidney injury, or current antibiotic use. By routine method, 56 (28.6 %) patients had significant pyuria. In comparison to the chamber counting method, the routine method was 69.4(63-75.8) % sensitive and 80.6(75.1-86.2) % specific. The positive and negative predictive values were 44.6 (37.7- 51.6) % and 92.1 (88.4 - 95.9) %. We concluded that sterile pyuria and microscopic hematuria could be present in a proportion of hospitalized general medical patients without UTI or ASB. Clinical judgment is essential in interpreting the significance of abnormal urinalysis reports.