Antimicrobial usage and resistance trend relationships from the MYSTIC Programme in North America (1999-2001)

Background: The MYSTIC Programme is a global, longitudinal antimicrobial surveillance network of hospitals that frequently utilize carbapenems. One aspect of the programme is the ability to capture antimicrobial consumption data from participating institutions. The current report evaluates these relationships for Enterobacteriaceae and Pseudomonas aeruginosa over the initial 3 year period of the programme in the USA. Methods: Between 10 and 15 medical centres participated during 1999-2001, each submitting up to 200 isolates/year (7003 strains overall). Evaluations of the relationship between drug usage and antimicrobial resistance in P. aeruginosa and Enterobacteriaceae for the carbapenems (imipenem and meropenem), cefepime, ceftazidime, ciprofloxacin, gentamicin and piperacillin-tazobactam were determined. Data were analysed based on: (1) aggregate usage results; (2) medical centre-specific usage compared with resistance rates; and (3) medical centre-specific usage results compared with yearly changes in resistance rates (DeltaR). The parameter of drug usage was the defined daily dose (DDD)/100 patient days calculated from total grams administered, using WHO definitions. Results: Resistance (1999-2001) among Enterobacteriaceae did not change significantly for beta-lactams, but tended to increase slightly for gentamicin (+1.1%) and ciprofloxacin (+3.1%). P. aeruginosa resistance rates (1999-2001) for gentamicin (+9.0%) and ciprofloxacin (+10.2%) increased, in contrast to a significantly decreased resistance rate for meropenem (-7.7%). Formulary-use changes were noted: increased meropenem and ciprofloxacin use and decreased consumption for imipenem, aminoglycosides, ceftazidime and cefepime. Aggregate ciprofloxacin DDD/100 days rates were directly related (+3.3 DDD) to Enterobacteriaceae and P. aeruginosa resistance changes, whereas among P. aeruginosa, usage and resistance were inversely correlated for gentamicin (-3.8 DDD; +9.0% resistant). Medical centre-specific antimicrobial usage calculations did not demonstrate a correlation to rates of resistance (r = -0.38 to 0.61) or yearly changes in resistance rates (r = -0.56 to 0.43). Conclusions: The availability of aggregate USA medical centre antimicrobial usage data enabled us to identify several important trends in the incidence of resistance among P. aeruginosa and Enterobacteriaceae: (1) increased use of ciprofloxacin associated with a higher resistance among Enterobacteriaceae; and (2) a correlation between ciprofloxacin categories of resistance and levels of resistance to other antimicrobial classes in P. aeruginosa. Medical centre-specific antimicrobial usage and resistance did not demonstrate direct statistical relationships, and require a continued search for other monitoring methods that can better identify antimicrobial/environmental factors that lead to resistance.