Urinary tract infections (UTI) are among the most common bacterial infections and urine samples represent a large proportion of the specimens processed in clinical microbiology laboratories, up to 80% of which, however, yield negative results. Automated microscopy is widely used for urine sediment analysis and has recently been evaluated in a few studies for bacteriological screening of urine samples, achieving high levels of performance. We present a study in which urine samples from both inpatients and outpatients, with either clean-catch or indwelling catheter urine samples, were screened for UTI by urine culture, as the reference method, and the automated urine analyser sediMAX, for the detection of bacteria, leukocytes and yeasts. In total, 3443 urine samples were evaluated. When a single algorithm was adopted for sediMAX to screen the total patient population, 96.4% sensitivity, 75.4% specificity, 57.8% positive predictive value, and 98.4% negative predictive value were found. However, for male outpatients and all patients with indwelling catheter other algorithms were necessary to improve performances. Altogether, with sediMAX false negative rate was 2.4% and false positive rate was 27.6%. In addition, 54% of the investigated samples could have avoided urine culture. After the identification of specific algorithms for different patient subgroups, the automated urine analyser sediMAX can be reliably employed in the screening of UTI.
Read full abstract