ABSTRACT

OBJECTİVE:

To assess whether or not automated urinanalysis is helpful for detection of asymptomatic bacteriuria and urinary tract infections in pregnant patients. Accuracy is evaluated by using urine culture as a gold standard.

MATERİALS and METHODS:

Midstream first void urine samples obtained from 102 pregnant women were tested using automated urinanalysis to detect nitrite, total bacteria and leukocyte counts. Urine cultures were performed using blood agar and eosin methylen blue agar, and more than 100,000 colony forming units per ml indicated urinary tract infection or asymptomatic bacteriuria. Each result was compared with urine culture. Validity of the automated urinanalysis was investigated, using sensitivity, specificity, positive and negative predictive values for each parameter and in combination.

RESULTS:

Nitrite positivity was found to be the most accurate (96%) variable for detection of UTI and AB, and showed excellent correlation with the urine culture (r=0.8, p<0.001). The combined use of three parameters reduced the sensitivity (33%) but increased specificity (100%), accuracy (92%) rates and positive and negative predictive values (100% and 92%, respectively). Total visible cost was €211.14 for automated urine analysis and €448.38 for urine culture for the 102 patients. However, when the cost of under/over treatment according to results of automated urine analysis is taken into account, its cost exceeds that of urine culture.

DİSCUSSİON:

Based on its high specificity and positive predictive value, patients with positive nitrite in automated urine analysis can be treated empirically until results of urine culture obtained. However, negative results for nitrite do not exclude asymptomatic bacteriuria and urinary tract infections, and urine culture is still needed for definite diagnosis. Although, it provides useful information, a routine use of automated urinanalysis for screening of bacteriuria proved neither valid nor cost-effective in pregnant patients.