Impact of an antimicrobial stewardship program at a university hospital of respiratory diseases

Abstract

Background: Antimicrobial Stewardship Programs (ASP's) have the potential to reduce antimicrobial exposure and decrease microbial resistance, and to decrease costs. The objective of this study was to characterize the overall impact of an ASP on antimicrobial utilization, antimicrobial resistance and mortality rates. Methods & Materials: Quasi-experimental study, before and after ASP implementation, from January to May 2016 and from January to May 2017 respectively, in an 8-bed adult intensive care unit at a university hospital of respiratory diseases. ASP included: elaboration of local guides (LG's); empiric therapy according to LG's; taking cultures before starting antibiotic therapy (AT); de-escalation of therapy as soon as culture results were available; discontinuation of AT if evidence of infection was absent; short courses of AT. Outcomes for analyses included: retrospective data for antimicrobial utilization measured as defined daily doses (DDD); bacterial susceptibilities of the first five most frequent organisms isolated in each period; and mortality rates for all causes and for infectious causes according to ICD-10 classification. Results: The overall reduction of all antibiotic utilization was 19%, including a decrease by 81% in meropenem use, 11% for imipenem, 39% for ciprofloxacin, 93% for ceftriaxone, 23% for ceftazidime and 33% for vancomycin. The most frequently isolated organism in the ASP period were Pseudomonas aeruginosa, with a reduction by 50% of Acinetobacter baumanii, the most frequent isolation in the pre-ASP period. Pseudomonas aeruginosa showed resistance to ceftazidime and cefepime in 55% of the cases in the pre-ASP period, whereas only 5% were resistant to the same antibiotics in the ASP period. The susceptibility pattern of Klebsiella pneumoniae in the pre-ASP period was 50% KPC and 83% ESBL, in the ASP period there was no carbapenem resistance and 75% ESBL. There was a 71% reduction in overall mortality rates and a 35% reduction in mortality due to infectious causes in the ASP period. Conclusion: There were limitations in this study due to missing individual patient data. There is a trend towards a decrease in the utilization of antibiotics and a shift of the epidemiological and susceptibility patterns of multidrug resistance organisms. The implementation of ASP was not accompanied by an increase in mortality.