Trends in causative organisms reported for Surgical Site Infections in English NHS Hospitals, 1999-2008

Background: The Surgical Site Infection Surveillance Service (SSISS) was established in 1997, England involving the surveillance of surgical site infections (SSI) in 14 surgical categories. Data on causative pathogens are collected.

Objective: To investigate the distribution of mico-organisms reported in SSI specimens and to evaluate changes in the epidemiology of these organisms.  

Methods: Data on operations collected between January 1999-December 2008 were included. Excluded: low volume surgeries and SSIs identified post-discharge. A maximum of three organisms can be reported for each case hence the analysis was based on counts of organisms. A GLM model with a log link function for the poisson distribution was used to analyse trends. The model was re-parameterised with different base years to evaluate the changes in the number of organisms between selected years.

Results: From 1999-2008, 246 hospitals contributed data on 457910 operations, of which 9155 SSIs were identified prospectively. Of these SSIs, 79% had valid organism data (7211/9155), providing a total of 9,673 organisms for analysis. The distribution of organisms varied by category. In hip prosthesis, Methicillin-resistant Staphylococcus (MRSA), Methicillin-sensitive Staphylococcus (MSSA) and coagulase-negative Staphylococcus (CNS) were the most common, accounting for 54% of all reported hip SSIs. In coronary artery bypass graft surgery (CABG), Enterobacteriaceae organisms accounted for 32% of all CABG SSIs, followed by MRSA and CNS: combined, they accounted for 62% of all CABG SSIs. In large bowel surgery, Enterobacteriaceae organisms predominated, followed by MRSA and Enterococci spp: combined, they accounted for 60% of all large bowel SSIs. Although the frequency of a specific pathogen differs by category, MRSA was the most common pathogen per year for all combined categories until 2008, decreasing from 27% in 2005 to 14% in 2008 with a significant decrease from 2007 to 2008 (RR 2008 vs base year 2007=0.65, 95% CIs 0.52-0.82; p<0.001). Meanwhile, the proportion of SSIs with MSSA remained unchanged. For all combined categories, Enterobacteriaceae organisms accounted for approximately 20% of all SSIs in each year and were second most common group. The proportion increased significantly from 21% in 2007 to 27% in 2008, making Enterobacteriaceae organisms the most common among SSIs in 2008 (RR 2008 vs base year 2007=1.29, 95% CI 1.06–1.58; p=0.010).

Conclusions: The distribution of SSI organisms differ at category-specific level.  The epidemiology of MSSA proportions indicates that the trends do not reflect the downward trends observed for MRSA. The proportion of SSIs with Enterobacteriaceae organisms had been steady over time although there was a significant increase from 2007 to 2008. The evidence from this study indicates that the focus on MRSA infection control has not broadened to the control of infection from other organisms.