Bloodstream infections caused by extended-spectrum-β-lactamase-producing Klebsiella pneumoniae:: Risk factors, molecular epidemiology, and clinical outcome

Bloodstream infections caused by extended-spectrum-p-lactamase (ESBL)-producing Klebsiella pneumoniae isolates are a major concern for clinicians, since they markedly increase the rates of treatment failure and death. One hundred forty-seven patients with K pneumoniae bloodstream infections were identified over a 5-year period (January 1999 to December 2003). The production of ESBLs in bloodstream isolates was evaluated by molecular methods. A retrospective case-case-control study was conducted to identify risk factors for the isolation of ESBL, producing K pneumoniae or non-ESBL-producing K pneumoniae isolates in blood cultures. Forty-eight cases infected with ESBL-producing K pneumoniae isolates and 99 cases infected with non-ESBL-producing K. pneumoniae isolates were compared to controls. Risk factors for isolation of ESBL-producing K pneumoniae isolates were exposure to antibiotic therapy (odds ratio [OR], 11.81; 95% confidence interval [CI], 2.72 to 51.08), age (OR, 1.14; 95% CI, 1.08 to 1.21), and length of hospitalization (OR, 1.10; 95% CI, 1.04 to 1.16). Independent determinants for isolation of non-ESBL-producing K pneumoniae were previous urinary tract infection (OR, 8.50; 95% CI, 3.69 to 19.54) and length of hospitalization (OK 1.07; 95% CI, 1.04 to 1.10). When the initial response was assessed at 72 h after antimicrobial therapy, the treatment failure rate for the ESBL-producing K pneumoniae-infected group was almost twice as high as that of the non-ESBL-producing K pneumoniae-infected group (31% versus 17%; OR, 2.19; 95% CI, 0.98 to 4.89). The 21-day mortality rate for all patients was 37% (54 of 147); it was 52% (25 of 48) for patients with ESBL-producing K pneumoniae bloodstream infections and 29% (29 of 99) for patients with non-ESBL-producing K. pneumoniae bloodstream infections (OR, 2.62; 95% CI, 1.28 to 5.35). In summary, this investigation identifies epidemiological characteristics that distinguish ESBL-producing K pneumoniae infections from non-ESBL-producing K pneumoniae ESBL bloodstream infections.