Abstract :
ISSUE: Antimicrobial susceptibility data are often aggregated into “antibiograms,” which provide a summary picture of common organisms and their susceptibility to many antimicrobial agents. Local antibiograms (e.g., by unit, hospital, or facility) provide a starting point for making decisions about empiric antimicrobial treatment. Clinicians are not making use of the antibiogram information, nor taking into account the local patterns of resistance and the further increase of resistant strains resulting from inappropriate antibiotic therapy.
PROJECT: The antibiotic imipenem is used within the acute care areas as a last resort medication for the treatment of serious infections. Current surveillance systems monitoring emerging drug resistance detect susceptibility patterns for each acute care entity. In this study, four hospitals that comprise part of a network were used to compare their resistance patterns to imipenem for the organism Pseudomonas aeruginosa, using their respective antibiograms. The dependent variable is the resistance of the organism; independent variables are the hospitals themselves and the time elapsed. A full model two-way ANOVA analysis was performed, with a subsequent one-way ANOVA performed as the reduced model.
RESULTS: Retrospective review of the antibiogram resistance patterns for three time periods were analyzed from each of the four hospitals. Full model: Two-way ANOVA analysis indicates that Hu Hospital has the largest average resistance. Hi Hospital has a small average, indicating a higher percentage of susceptible isolates of Pseudomonas aeruginosa. Over time, the mean value of resistance did not change very much. Resistance patterns for one of the four hospitals were quite distinctly different (p = 0.012), illustrating the need to monitor local data even when working within a multisite system.
Reduced model: One-way ANOVA results confirmed that the one hospital had substantially higher rates of resistance than the other three hospitals.
LESSONS LEARNED: One of the hospitals had a different antibiogram and resistance pattern than the other three hospitals. Knowledge of local patterns of resistance contained within the antibiogram will allow the infection control nurse to be alert to significant shifts in the populations of organisms within the locale and report those findings to all stakeholders. This information can be shared via Internet and PDA for real-time clinical decisions at the bedside.
