Strategy to reduce E. coli bacteraemia based on cohort data from a London teaching hospital

Background and aim In 2017, National Health Service Improvement set a 10% reduction target for Escherichia coli bacteraemia by 2018, followed by a 50% reduction in healthcare-associated Gram-negative bacteraemias by 2022. We analysed consecutive cases of E. coli bacteraemia and devised a strategy to achieve these targets. Methods From December 2012 to November 2013, demographic, clinical and microbiological data were prospectively collected on all patients with bacteraemia at the Royal London Hospital in East London, UK. Results There were 594 significant bacteraemic episodes and 207 (34.8%) were E. coli. Twenty-four (11.6%) of the E. coli isolates were extended spectrum beta-lactamase producers, 22 (10.6%) gentamicin resistant and 2 (1.0%) amikacin resistant. The three most common sites of infection were pyelonephritis 105 (56.7%), catheter-associated urinary tract infection 22 (10.6%), and other medical devices and procedures that cause bacteraemia 32 (15.5%). In the pyelonephritis group, trimethoprim resistance in urinary isolates was 16/47 (34.0%) compared with 3/47 (6.4%) for nitrofurantoin. Twelve months postbacteraemia, recurrent bacteraemia rates were 10/105 (9.5%). There were 44 medical device-associated E. coli bacteraemias, and 22 (50%) were urinary catheter associated. There were 10 patients with E. coli bacteraemia caused by procedures, seven genitourinary or biliary tract instrumentation and three postgastrointestinal surgery. Conclusion E. coli bacteraemias related to urosepsis could have been prevented by better empirical treatment and targeted prophylaxis. Urinary catheter quality improvement programmes should contribute to a further reduction. For patients undergoing high-risk urinary or biliary tract procedures or device manipulation, we advocate single-dose amikacin prophylaxis.