Enterobateriaceae replacing MRSA as the leading causative organism of surgical site infections in England: analysis of trends 2000-2009

Background: Gram-positive cocci have historically accounted for the majority of surgical site infections (SSIs) in England. In light of the decreases in MRSA bacteraemia in England since 2006, analyses of SSI data were undertaken to identify any parallel changes in the epidemiology of SSIs.

Objective: To investigate changes in the epidemiology of pathogens causing SSIs between 2000-2009, focusing on Enterobacteriaceae and staphylococci.

Methods: National SSI surveillance data collected by NHS hospitals in England between 2000-2009 for 13 categories of surgery were extracted for analysis. SSIs detected during hospital stay and at readmission were included. Analyses were based on all causative organisms reported (maximum of three per SSI). A GLM model for a binomial distribution with a log link function to estimate risk ratios (RR) was used. The model was reparameterised with different base years to evaluate changes in proportions between defined years.

Results:  Between 2000-2009, 244 hospitals contributed data on 542,767 operations from which 11,358 developed SSIs. Of these SSIs, 77% (8,735/11,358) had valid organism data with 27% containing multiple causative organisms. Records with valid SSIs provided 11,647 organisms for analysis due to reporting of multiple causative organisms.  Enterobacteriaceae accounted for 21% of all SSIs (2,474/11,647) comprising: E. coli (16%); Proteus spp (8%); E. cloacae (5%); Klebsiella spp (5%), other named organisms (12%) and organisms not reported to genus level (54%). Enterobacteriaceae were the most common cause of SSIs in large bowel surgery (34%; 692/2,073); coronary artery bypass graft (33%; 370/1,131); small bowel surgery (31%; 85/279) and spinal surgery (29%; 8/27). For all categories combined, MRSA was the predominant reported cause of SSI until 2007 when overtaken by Enterobacteriaceae. The proportion of SSIs due to Enterobacteriaceae increased significantly from 20% in 2007 to 25% in 2008 (RR=1.26; p=0.004). The increase from 25% in 2008 to 29% was also significant (RR=1.15; p<0.001). These effects remained significant after adjusting for surgical category in each of these periods (p=0.020; p=0.049 respectively). Relative decreases in MRSA as a cause of SSI occurred from 2006 with significant decreases between 2007-2008 (RR=0.65; p<0.001) and between 2008-2009 (RR=0.62; p<0.001). MSSA showed small but non-significant increases from 2006 onwards. No significant patterns emerged for CNS which was stable at approximately 10%. 

Conclusions: The relative importance of Enterobacteriaceae as a cause of SSIs increased progressively since 2007 whilst significant decreases in MRSA infection were observed. These effects remained after adjusting for surgical category. The changing aetiology of SSI may have implications for reassessing choice of antimicrobial prophylaxis for certain surgical procedures.