Surveillance of bloodstream infections in intensive care units in England, May 2016–April 2017: epidemiology and ecology

Abstract

Bloodstream infections (BSIs) in patients in intensive care units (ICUs) are associated with increased morbidity, mortality and economic costs. Many BSIs are associated with central venous catheters (CVCs). The Infection in Critical Care Quality Improvement Programme (ICCQIP) was established to initiate surveillance of BSIs in English ICUs. A web-based data capture system was launched on 1st May 2016 to collect all positive blood cultures (PBCs), patient-days and CVC-days. National Health Service (NHS) trusts in England were invited to participate in the surveillance programme. Data were linked to the antimicrobial resistance dataset maintained by Public Health England and to mortality data. Between 1st May 2016 and 30th April 2017, 84 ICUs (72 adult ICUs, seven paediatric ICUs and five neonatal ICUs) based in 57 of 147 NHS trusts provided data. In total, 1474 PBCs were reported, with coagulase-negative staphylococci, Escherichia coli, Staphylococcus aureus and Enterococcus faecium being the most commonly reported organisms. The rates of BSI and ICU-associated CVC-BSI were 5.7, 1.5 and 1.3 per 1000 bed-days and 2.3, 1.0 and 1.5 per 1000 ICU-CVC-days in adult, paediatric and neonatal ICUs, respectively. There was wide variation in BSI and CVC-BSI rates within ICU types, particularly in adult ICUs (0-44.0 per 1000 bed-days and 0-18.3 per 1000 ICU-CVC-days). While the overall rates of ICU-associated CVC-BSIs were lower than 2.5 per 1000 ICU-CVC-days across all age ranges, large differences were observed between ICUs, highlighting the importance of a national standardized surveillance system to identify opportunities for improvement. Data linkage provided clinically important information on resistance patterns and patient outcomes at no extra cost to participating trusts.