Urinary tract infections caused by multi-drug resistant Proteus mirabilis: Risk factors and clinical outcomes.

BACKGROUND: Proteus mirabilis (PM) as well as other members of the Enterobacteriaceae family are a leading cause of infectious diseases in both the community and acute care settings. The prevalence of multi-drug resistant (MDR) bacterial isolates have increased in the last few years, affecting the prognosis and survival of hospitalized patients. The aim of our study was to determine the risk factors and clinical outcomes of urinary tract infections (UTIs) caused by MDR PM in patients hospitalized in our institution.
METHODS: This was a retrospective matched case-control study. Records of patients with PM-positive urine culture were reviewed, and data were included for analysis.
RESULTS: Univariate analysis revealed that the variables significantly associated with acquisition of MDR PM vs non-MDR PM UTI were younger age ([in years] median 77.5, range 20-94 vs median 78, range 40-94, p = 0.04), other concomitant infectious diseases (57.1 vs 35.7%, p = 0.037),number of prior infectious diseases (mean 0.95 +/- 0.99 vs 0.57 +/- 0.85, p = 0.035), diagnosis of infection at hospital admission (67.9 vs 42.9%, p = 0.008), and prior therapy with antipseudomonal penicillin (17.9 vs 1.8%, p = 0.004),respectively. Mean length of hospitalization was 29.95 days for the MDR group and 30.04 days for the non-MDR group(p = non-significant [NS]). The crude mortality rate following hospital admission was 19/56 (33.9%) vs 14 (25%)in the MDR PM and non-MDR PM groups, respectively(p = 0.300, odds ratio [OR] 1.54, 95% confidence interval[CI] 0.63-3.82). The production of extended-spectrum beta lactamases(ESBL) was found in 100% of MDR-PM vs 31.5%of non-MDR-PM urine isolates (p < 0.001). All variables found to be significantly associated with MDR-PM UTI were included in a logistic regression model. Independent risk factors for MDR-PM UTI were empiric cephalosporin therapy(OR 4.694, 95% CI 1.76-12.516, p = 0.002) and prior antipseudomonal penicillin (piperacillin/tazobactam) therapy during the last year (OR 11.175, 95% CI 1.09-114.2,p = 0.04).
CONCLUSIONS: Prior piperacillin/tazobactam and empiric cephalosporin use were the independent risk factors of MDR-PM strains. All MDR-PM urinary isolates at our institution were ESBL producers. Therefore, carbapenem use remains the only available treatment option for MDR-PM isolates in our institution.