This study investigated the molecular epidemiology of Acinetobacter baumannii in the University of Debrecen in relation to antibiotic consumption. Overall and ward-specific antibiotic consumption was measured by the number of defined daily doses (DDD) per 100 bed-days between 2002 and 2012. Consumption was analysed against the number of A. baumannii positive patients per 100 bed-days, number of isolates per positive sample, and proportion of carbapenem resistant A. baumannii, using time-series analysis. Altogether 160 A. baumannii isolates from different wards were collected and analysed. Carbapenemase genes bla OXA-23-like , bla OXA-24-like , bla OXA-48-like , bla OXA-51-like , bla OXA-58-like and integrons were sought by PCR. Relatedness of isolates was assessed by PFGE. Prevalence and carbapenem resistance of A. baumannii were statistically associated with carbapenem consumption. Prevalence data followed carbapenem usage with three quarterly lags (r50.51-0.53, P,0.001), and meropenem and ertapenem, but not imipenem usage, affected prevalence. Colistin usage, in turn, lagged behind prevalence with one lag (r50.68-0.70, P,0.001). Six clusters were identified; the neurology ward with the lowest carbapenem consumption was associated with the carbapenem-susceptible cluster, as well as with the carbapenem-susceptible isolates in the cluster with variable susceptibility. Wards with high carbapenem usage almost exclusively harboured isolates from carbapenem-resistant clusters. All clusters were dominated by isolates of one or two wards, but most wards were represented in multiple clusters. Increases in prevalence and carbapenem resistance of A. baumannii were associated with usage of meropenem and ertapenem but not of imipenem, which led to the spread of multiple clones in the University.

Because asymptomatic carriage of extended-spectrum beta-lactamase (ESBL) producers is a risk factor for infection, data on colonization dynamics are important when planning infection control. This study investigated fecal colonization with ESBL producers among inpatients, outpatients and medical students and compares the characteristics of ESBL producers among these groups. Carriage rates were investigated in 5581 fecal samples; 4343 from inpatients (330, 1397, 619 and 1864 from adult ICUs [intensive care units], adult non-ICUs, pediatric ICUs and pediatric non-ICUs, respectively), 814 from outpatients and 424 from screening of medical students. ESBL producers were characterized by co-resistance, integrons carried, and aminoglycoside resistance and ESBL genes. Dynamic regression models were built to identify relationships between combinations of time series of monthly antibiotic consumption, prevalence of carriers and infected subjects. Inpatients, ICU patients and adults showed higher prevalence than outpatients, non-ICU patients or children (7.4%, 9.3% and 12.0% vs. 3.1%, 6.1% and 4.1%, respectively). Klebsiella pneumoniae was more frequent in ICU patients; dominance of CTX-M-15 producers was more marked in adult than in pediatric inpatients. ESBL carriage was shown to be a consequence of infection in adults in the time-series analysis; antibiotic consumption had little effect. The epidemiology of colonization with ESBL producers differed between pediatric ICU, adult ICU and adult non-ICU patients. In adults, carriage of ESBL producers seems to be the consequence of infection, especially in ICU patients; the main source of colonization is nosocomial acquisition. In contrast, children are less likely to acquire colonizer strains in hospitals; importation of ESBL producers by colonized children seems to be significant.