Abstract BackgroundNotifications of non-O157 Shiga Toxin Producing Escherichia Coli (STEC) infections have increased in England since local laboratories implemented PCR testing of gastrointestinal pathogens. This study describes the epidemiology of non-O157 STEC cases between 2012 and 2016 and identifies risk factors for Haemolytic Uraemic Syndrome (HUS) from clinical symptoms, virulence factors (shiga toxin (stx)1, stx2 and eae,) and stx subtypes.MethodsWe analysed laboratory typing results for isolates of non-O157 STEC including presence of stx1, stx2 and eae between 2012 and 2016. We reviewed local records for symptoms, case risk group and onward transmission. We undertook univariate and multivariable analyses to identify exposures associated with HUS amongst non-O157 STEC cases.Results5527 specimens were tested for non-O157 STEC. 1190 were positive for one of stx1, stx2, eae, faecal culture or serology for a non-O157 STEC serogroup. Of 797 culture or serology confirmed isolates, 71 had stx subtypes 2a, 2c and/or 2d.HUS was seen in 3.4% (41/1190) of cases with any laboratory evidence of non-O157 E. coli exposure and associated with stx2 and eae positivity (Odds ratio (OR) 12.6; 95% Confidence Interval (95%CI) 5.7-29.1.Amongst culture or serology confirmed cases, 4.4% (35/797) had HUS, associated with vomiting (OR 8.9; 95%CI 2.8-28.1), bloody diarrhoea (OR 5.5; 95%CI 1.7-18.1), symptomatic epidemiological link (OR 4.4; 95%CI 4.4-48.9), age under 6 years (OR 4.2;95%CI 1.3-13.2) . All HUS cases were stx subtypes 2a, 2c, and/or 2d.ConclusionsThis work has improved our understanding of the contribution of non-O157 serogroups to the STEC severe disease burden. We recommend inclusion of virulence factors, age and presence of vomiting or bloody diarrhoea in algorithms to identify non-O157 STEC cases at risk of severe disease. Funding No funding received