Impact of antimicrobial stewardship programs in public hospitals in Argentina.

Abstract

Background: Antimicrobial stewardship programs (ASP) are useful to fight antimicrobial resistance. We analyze the impact of ASP among public hospitals in Argentina. Methods & Materials: Multicenter study with interrupted-time series design, comparing a pre- and post-intervention periods (PostIP) of 6 months each in 4 public hospitals with no ASP. The prevalence of selected resistant pathogens (SRP) causing nosocomial infections was determined: methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant enterococci (VRE), extended spectrum ß-lactamase- (ESBL) and carbapenemase-producing enterobactariaceae (CPE), and multidrug-resistant Pseudomonas aeruginosa (MDRPA) and Acinetobacter baumannii (MDRAB). Antibiotic use (AU) was measured by defined daily dose (DDD) per 100 occupied bed-days (OBD) and reported in monthly median, in wards and intensive cares units (ICU). At month 7th, an ASP (post-prescription prospective audit modality) was started with restricted antimicrobials defined by each hospital. Results: In the PostIP there was a non-significant decrease in the overall AU (-147 DDD/100 OBD; p = 0.46), with a significant increase in general wards (+75; p = 0.03) and a non-significant decrease in ICU (-171; p = 0.4). Considering each antibiotic overall, the significant changes in AU were: increase of piperacillin/tazobactam (PTZ) (+53; p = 0.01) and vancomycin (VAN) (+15; p = < 0.001), and decrease of 3rd gen cephalosporins (3GC) (-31; p = 0.02) and metronidazole (-11; p = < 0.001). In wards, there was a significant increase in PTZ (+38;p = < 0.001), VAN (+4;p = 0.05), aminoglycosides (+9;p = 0.01), and carbapenemes (+5;p = 0.03), and significant decrease of 3GC (-14;p = 0.02). In ICU, the differences were significant for VAN (+11;p = 0.01) and metronidazol (-10;p = 0.01). In one hospital the overall DDD/100 OBD reached statiscally significant diminution (-49;p = 0.05), in another there was a significant increase (+90;p = 0.01), while the other 2 showed no significant difference (-254; p = 0,17 and +21;p = 0.19). We documented 1134 episodes of infections caused by SRP with no significant difference between the 2 periods. Overall, SRP in decreasing order of frequency were: ESBL, MRSA, MDRAB, MDRPA, CPE and VRE. MDRPA was the only with higher prevalence in the PostIP (p = 0.013). Conclusion: ASP in these hospitals did not significantly affect the AU nor the overall prevalence of SRP; however, partial positive results were obtained in some participating hospitals. Longer follow-up periods are required to better assess the potential benefits of an ASP.