Further research is needed to reduce the impact of selection bias by indication through creating more comparable patient groups and applying risk adjustment.

We determined the incidence, risk factors and antimicrobial susceptibility associated with Escherichia coli bacteraemia in England over a 24 month period. Case data were obtained from the national mandatory surveillance database, with susceptibility data linked from LabBase2, a voluntary national microbiology database. Between April 2012 and March 2014, 66,512 E. coli bacteraemia cases were reported. Disease incidence increased by 6% from 60.4 per 100,000 population in 2012–13 to 63.5 per 100,000 population in 2013–14 (p < 0.0001). Rates of E. coli bacteraemia varied with patient age and sex, with 70.5% (46,883/66,512) of cases seen in patients aged ≥ 65 years and 52.4% (33,969/64,846) of cases in females. The most common underlying cause of bacteraemia was infection of the genital/urinary tract (41.1%; 27,328/66,512), of which 98.4% (26,891/27,328) were urinary tract infections (UTIs). The majority of cases (76.1%; 50,617/66,512) had positive blood cultures before or within two days of admission and were classified as community onset cases, however 15.7% (10,468/66,512) occurred in patients who had been hospitalised for over a week. Non-susceptibility to ciprofloxacin, third-generation cephalosporins, piperacillin–tazobactam, gentamicin and carbapenems were 18.4% (8,439/45,829), 10.4% (4,256/40,734), 10.2% (4,694/46,186), 9.7% (4,770/49,114) and 0.2% (91/42,986), respectively. Antibiotic non-susceptibility was higher in hospital-onset cases than for those presenting from the community (e.g. ciprofloxacin non-susceptibility was 22.1% (2,234/10,105) for hospital-onset vs 17.4% (5,920/34,069) for community-onset cases). Interventions to reduce the incidence of E. coli bacteraemia will have to target the community setting and UTIs if substantial reductions are to be realised.

Community prescribing of antibiotics for RTIs significantly decreased following the introduction of the Quality Premium, with the greatest reduction seen in younger patients.

Previous research has shown that Escherichia coli infection rates peak in the summer; however, to date there has been no investigation as to whether this is seen in both hospital and community-onset cases, and how this differs across regions. We investigated and quantified E. coli bloodstream infection (BSI) seasonality. A generalized additive Poisson model was fitted to mandatory E. coli BSI surveillance data reported in England. There was no impact of seasonality in hospital-onset cases; however, for the community-onset cases, there was statistically significant seasonal variation over time nationally. When examined regionally, seasonality was significant in the North of England only. This variation resulted in an absolute increase of 0.06 (95% CI 0.02-0.1) cases above the mean (3.25) in each hospital trust for each week of the peak summer season, and a decrease of (-) 0.07 (95% CI -0.1 to -0.03) in the autumn. We estimate that fewer than one hospital bed-day per week per hospital is lost because of seasonal increases during the summer. Our findings highlight the need to understand the distinct community and hospital dynamics of E. coli BSI, and to explore the regional differences driving the variation in incidence, in order to design and implement effective control measures.