1098

Abstract

model except for the TaperGuard: while leak rate was on average approximately 65x higher than the novel device, fluid this result failed to reach statistical significance (p=0.3). While clear limitations exist due to the approximative nature of a bench-top setup, the novel device provided significantly better results in reducing leak rate in all but one of 8 comparisons, and non-significantly better results in one. While these results are promising, future studies are required for a more accurate and in-vivo comparison in order to evaluate the real benefit of this novel tool.Introduction: Increased prevalence of Extended-Spectrum Beta-Lactamase-producing enterobacteriaceae (ESBL-E) is becoming a clinical concern especially in UTI. The initial antibiotics selection needs to be taken into account for severity of UTI. Methods: This is a prospective cohort study in which patients with UTI were prospectively enrolled in a tertiary university-affiliated hospital in Japan from May 2012 to March 2013. UTI was defined as positive UTI symptoms, urine culture >=105 CPF/ml, and urine white blood cells >=10/HPF or apparent phagocytosis in Gram stain. UTI with ESBL-E were categorized into two groups: urosepsis vs. non-urosepsis. Clinical outcomes in the two groups were compared. Results: 338 consecutive cases of positive urine culture were registered. Chart review revealed 236 cases with UTI. Among 236 cases of UTI, 28.8% (68/236) met SIRS criteria ("urosepsis") and 71.2% (168/236) did not meet SIRS criteria ("non-urosepsis"). 16.2% (11/68) of urosepsis and 13.1% (22/168) of non-urosepsis had ESBL-E (p=0.53). Analysis among urosepsis cases, SAPS-II scores were higher in cases with ESBL-E (n=11 SAPS-II average 53.0) compared to the cases with non-ESBL-E (n=57, SAPS-II average 40.65, p=0.047). Among 22 cases of non-urosepsis with ESBL-E, causative pathogens were all E.coli. 9 cases were managed as inpatients and 13 cases as outpatients. About 50% cases of non-urosepsis with ESBL-E were treated with non-sensitive antibiotics. However all cases but 1 clinically improved successfully. Among 11 cases of urosepsis with ESBL-E, causative pathogens were as follows: 8 cases with E.coli, 2 with P.miralibis and 1 with K.pneumonia. 7 cases improved and 4 died. 5 out of 7 survived cases of urosepsis with ESBL-E were treated with carbapenem. Conclusions: Our study has demonstrated that the prognosis of non-urosepsis with ESBL-E is quite favorable and most of the cases do not necessarily need administration of carbapenems or inpatient management.