[Clinical predictors of asymptomatic bacteriuria during pregnancy]

To estimate the prevalence of asymptomatic bacteriuria among pregnant women attended at our university prenatal care clinic and to identify probable clinical predictors.

Across-sectional study was carried out from August 2008 to October 2009 at the Bahiana School of Medicine involving 260 pregnant women without symptoms of urinary tract infection. The following exclusion criteria were considered: presence of clinical signs such as fever, dysuria, vesical tenesmus, lumbar pain, history of active genital bleeding or loss of amniotic fluid, use of antimicrobial agents in the 30 days prior to sample collection, and refusal to participate in the project. The presence of single pathogen bacterial colonization ≥10(5) CFU/mL in the urine sample obtained from the middle jet was considered to be a dependent variable. The predictive factors evaluated were as follows: age, race, marital status, schooling, gestational age, hypertension, anemia, vaginal infection, sickle cell trait and previous history of urinary tract infection, urinary symptoms related to the lower urinary tract (frequency, urgency and nocturia) and data obtained from the urine summary (leukocyturia, increased bacterial flora, hematuria, proteinuria, and presence of nitrite). Statistical analysis was performed with the Statistical Package for the Social Sciences (SPSS) software version 13.0 and the level of significance was set at p<0.05. Prevalences were expressed as percentage, and the confidence interval considered was 95% (95%CI).

The prevalence of asymptomatic bacteriuria was 12.3% (95%CI=8.3-16.3). E. coli was the most frequent etiologic agent (59.4%). Logistic regression indicated that urgency to void (OR=5.99; 95%CI=2.20-16.31; p<0.001); leukocyturia (OR=2.85; 95%CI=1.04-7.83; p=0.042) and increased bacterial flora (OR=10.62; 95%CI=3.95-28.56; p<0.001) were independent predictors of asymptomatic bacteriuria.

The prevalence of asymptomatic bacteriuria in the studied population was high. The prediction score created for the final logistic regression model has an accuracy of 91.9% for bacteriuria.