Long-Term Impact of an Educational Antimicrobial Stewardship Program on Hospital-Acquired Candidemia and Multidrug-Resistant Bloodstream Infections: A Quasi-Experimental Study of Interrupted Time-Series Analysis

Abstract

Background. The global crisis of bacterial resistance urges the scientific community to implement intervention programs in healthcare facilities to promote an appropriate use of antibiotics. However, the clinical benefits or the impact on resistance of these interventions has not been definitively proved.Methods. We designed a quasi-experimental intervention study with an interrupted time-series analysis. A multidisciplinary team conducted a multifaceted educational intervention in our tertiary-care hospital over a 5-year period. The main activity of the program consisted of peer-to-peer educational interviews between counselors and prescribers from all departments to reinforce the principles of the proper use of antibiotics. We assessed antibiotic consumption, incidence density of Candida and multidrug-resistant (MDR) bacteria bloodstream infections (BSIs) and their crude death rate per 1000 occupied bed days (OBDs).Results. A quick and intense reduction in antibiotic consumption occurred 6 months after the implementation of the intervention (change in level, -216.8 defined daily doses per 1000 OBDs; 95% confidence interval, -347.5 to -86.1), and was sustained during subsequent years (average reduction, -19,9%). In addition, the increasing trend observed in the preintervention period for the incidence density of candidemia and MDR BSI (+0.018 cases per 1000 OBDs per quarter; 95% confidence interval, -.003 to.039) reverted toward a decreasing trend of -0.130 per quarter (change in slope, -0.029; -.051 to -.008), and so did the mortality rate (change in slope, -0.015;-.021 to -.008).Conclusions. This education-based antimicrobial stewardship program was effective in decreasing the incidence and mortality rate of hospital-acquired candidemia and MDR BSI through sustained reduction in antibiotic use.