A three-point time series study of antibiotic usage on an intensive care unit, following an antibiotic stewardship programme, after an outbreak of multi-resistant Acinetobacter baumannii.

Antibiotic use in intensive care units (ICUs) can promote antimicrobial resistance. Outbreaks of multi-resistant bacteria significantly affect patient outcomes and delivery of care. Antibiotic stewardship programmes (ASPs), combining root-cause analyses and multi-faceted prevention strategies, are necessary, often at significant cost and time. Which elements of such strategies have the largest impact on antibiotic usage following an outbreak is unclear. The aim of this study was to investigate how antibiotic usage in a university hospital ICU changed with a non-protocolised ASP following a disruptive outbreak of multi-resistant Acinetobacter baumannii (MRAB). This was a three time-period observational cohort study. The primary endpoint was the change in overall antibiotic usage (daily defined dose, DDD, antibiotic-days, antibiotic-courses) for consecutive ICU patients staying >48 h, over three 6-month study time periods pre-MRAB (2008, n = 84) and post-MRAB (2010, n = 88; 2012, n = 122). Secondary endpoints were changes in antibiotic usage and patient demographics, in predefined admission categories (Medical Emergency, ME; Surgical Elective, SEL; and Surgical Emergency, SE). The mean age (54.6 ± 17.7, 58.1 ± 17.9, 62.8 ± 19.1 years*) and severity of illness (APACHE 14.8 ± 8.0, 16.7 ± 6.8, 18.3 ± 6.1*) increased, particularly medical admissions. There was a sustained reduction in DDD antibiotic usage [1895.1 (2008), 1224.2 (2010), 1236.6 (2012) per 1000 patient-days] but no overall change in antibiotic-days or antibiotic-courses. Antibiotic usage (antibiotic-days) fell significantly in surgical emergency admissions [20.2 ± 32.1, 4.6 ± 7.4*, 5.9 ± 7.3]. There was a sustained drop in beta-lactam, quinolone, glycopeptide and macrolide usage. Following an MRAB outbreak, and subsequent operational changes including enhanced ASPs (non-protocolised), there was a sustained overall fall in antibiotic usage in spite of an increase in disease severity over 5 years.