Antimicrobial susceptibility of gram-negative pathogens isolated from patients with complicated intra-abdominal infections in South African hospitals (SMART Study 2004-2009) : impact of the new carbapenem breakpoints

Abstract

BACKGROUND: The Study for Monitoring Antimicrobial Resistance Trends (SMART) follows trends in resistance among aerobic and facultative anaerobic gram-negative bacilli (GNB) isolated from complicated intra-abdominal infections (cIAIs) in patients around the world. METHODS: During 2004–2009, three centralized clinical microbiology laboratories serving 59 private hospitals in three large South African cities collected 1,218 GNB from complicated intra-abdominal infections (cIAIs) and tested them for susceptibility to 12 antibiotics according to the 2011 Clinical Laboratory Standards Institute (CLSI) guidelines. RESULTS: Enterobacteriaceae comprised 83.7% of the isolates. Escherichia coli was the species isolated most commonly (46.4%), and 7.6% of these were extended-spectrum b-lactamase (ESBL)-positive. The highest ESBL rate was documented for Klebsiella pneumoniae (41.2%). Overall, ertapenem was the antibiotic most active against susceptible species for which it has breakpoints (94.6%) followed by amikacin (91.9%), piperacillin-tazobactam (89.3%), and imipenem-cilastatin (87.1%), whereas rates of resistance to ceftriaxone, cefotaxime, ciprofloxacin, and levofloxacin were documented to be 29.7%, 28.7%, 22.5%, and 21.1%, respectively. Multi-drug resistance (MDR), defined as resistance to three or more antibiotic classes, was significantly more common in K. pneumoniae (27.9%) than in E. coli (4.9%; p < 0.0001) or Proteus mirabilis (4.1%; p < 0.05). Applying the new CLSI breakpoints for carbapenems, susceptibility to ertapenem was reduced significantly in ESBL-positive E. coli compared with ESBL-negative isolates (91% vs. 98%; p < 0.05), but this did not apply to imipenem-cilastatin (95% vs. 99%; p = 0.0928). A large disparity between imipenem-cilastatin and ertapenem susceptibility in P. mirabilis and Morganella morganii was documented (24% vs. 96% and 15% vs. 92%, respectively), as most isolates of these two species had imipenem-cilastatin minimum inhibitory concentrations in the 2–4 mcg/mL range, which is no longer regarded as susceptible. CONCLUSIONS: This study documented substantial resistance to standard antimicrobial therapy among GNB commonly isolated from cIAIs in South Africa. With the application of the new CLSI carbapenem breakpoints, discrepancies were noted between ertapenem and imipenem-cilastatin with regard to the changes in their individual susceptibilities. Longitudinal surveillance of susceptibility patterns is useful to guide recommendations for empiric antibiotic use in cIAIs.