Correlation of urinalysis and dipstick results with catheter-associated urinary tract infections in surgical ICU patients

Objective: To determine the utility of urinalysis and dipstick results in predicting urinary tract infections in catheterized ICU patients. Design and setting: Urine samples were collected for 4 months from patients admitted a ten-bed surgical ICU of an urban public teaching hospital designated by the state of New York as a level I trauma center. The correlation was analyzed between urinalysis and dipstick results from urine samples and subsequent quantitative culture results. Patients: All patients with indwelling urinary catheters admitted to the ICU were considered eligible; 106 patients were enrolled, and 300 individual urine samples were analyzed. Measurements and results: There were 44 catheter-associated urinary tract infections. Nitrite presence was the best indicator of infection (91.8% specificity) but was not a reliable clinical test due to a sensitivity of 29.5% and positive and negative likelihood ratios of 3.52 and 0.56, respectively. None of the other parameters (leukocyte esterase, white blood cell count, urobilinogen, presence of yeast or bacteria) were independently correlated with the culture results either individually or in combination. Conclusions: Based on our data we cannot recommend the use of urinalysis or dipstick in screening for potential catheter-associated urinary tract infections.