Correlation between Antibiotic Use and E.coli Resistances in a Swiss Tertiary Care Hospital
Monitoring antimicrobial use and local antimicrobial susceptibility in hospitals are important tools of antimicrobial stewardship programs. In our institution 2144, E.coli from clinical samples were isolated in 2011. We aimed to correlate resistance of these samples with antibiotic use.
The hospital pharmacy calculated the antibiotic use in grams for each department of the hospital for the years 2008-2011. We converted the use into defined daily doses (DDD) and expressed it as DDD/100 bed days, using the ATC/DDD system promoted by the WHO.
Antimicrobial susceptibility tests (ASTs) were routinely assessed using the disk diffusion method and results interpreted according to CLSI criteria. Resistance rates of E.coli represent the number of resistant isolates divided by the total number of clinical isolates for which AST was performed. All hospital departments with a total of >20 E.coli isolates in urine, blood or other clinical material in the year 2011 were included in the analyses. Per patient only the first isolated E.coli was included. The correlation between antibiotic use from 2008-2011 and E.coli resistance rate in the year 2011 was determined with linear regression analyses.
The total antibiotic consumption for the hospital was 64 DDD/100 hospital days with high variability between the different departments. The hospital-wide resistance rates of E.coli are summarized in Figure 1. Antibiotic use correlated with E.coli resistance for amoxicillin-clavulanic acid [coefficient of determination (r2) 0.409 (p-value: 0.019)], trimethoprim-sulfamethoxazole [r2 0.367 (p-value: 0.028)] and for fluoroquinolones (ciprofloxacin and norfloxacin), [r2 0.566 (p-value: 0.003)]. (Figure 2) There was no correlation between use and resistance for amoxicillin, cefuroxime, ceftriaxone, cefepime, meropenem and gentamicin.
A correlation between inhospital antibiotic use and E.coli resistances was documented exclusively for antibiotics with a high consumption rate which are both available in oral and parenteral form. Two of the antibiotics in question are often prescribed for treatment of urinary tract infections. Our data could serve as basis for antibiotic-specific stewardship interventions.
A. Kronenberg, None
J. Marschall, None