Abstract
BACKGROUND: Semi-quantitative bacteruria counts (s-QBC) are important in the diagnosis of urinary tract infection (UTI) due to most uropathogens. The prognostic value of s-QBC for diagnosis of UTI due to group B streptococcus (GBS) is unknown. In this study, we assessed the value of s-QBC for differentiating acute GBS UTI from asymptomatic bacteruria (ABU), independent of other potential prognostic indicators. METHODS: Medical record review and urinalysis (UA) values for 1593 patients who had urinary GBS isolated (103 to ≥105 CFU/ml) during a four-year period were analyzed using binary logistic regression to determine the predictive values of s-QBC, age, and gender for infection category (acute UTI, ABU) based on the clinical diagnosis. RESULTS: s-QBC alone had a strong predictive value for infection category but only for ABU. Multivariate logistic regression showed similar predictive power of s-QBC for infection category using age as a co-predictor, which was also independently associated with infection category. Typical s-QBC cut-off values that are commonly used in diagnostic settings had no significant power in predicting infection category. Among other UA measures, proteinuria and hematuria were significantly associated with acute infection. CONCLUSIONS: Together, these data show that s-QBC is not useful in the differential diagnosis of GBS UTI. Among the patients in this study, age was an equally effective prognostic indicator compared to s-QBC for identifying high- and low-risk patients for acute GBS UTI. Collectively, these findings indicate that age-based associations may be equally as useful as s-QBC for predicting infection category in the setting of adult patients with GBS-positive urine cultures.