Introduction Violent injury places a large burden on the NHS. We had implemented information sharing in our ED in 2007 and aimed to see which patient groups were most affected by information sharing, as this would provide clues as to how this complex intervention works.

Methods Retrospective time series study of all the assault victims presenting for ED care between 2005 and 2014 at a single ED in England.

Results 10 328 patients presented during the study period. There was a 37% decrease in the number of patients presenting after assault, consistent with national trends. The proportions of people arriving by ambulance, and the proportion of men did not change during the study period. There were no important changes in the age of our assault patients in this study. Greater, disproportionate, decreases in rates of violence were seen in patients who presented at the weekend up (incidence rate ratio (IRR)=0.57, 95% CI 0.50 to 0.64) versus weekdays (IRR=0.72; 95% CI 0.62 to 0.83) There were also disproportionately greater decreases over the study period in patients who were discharged with no hospital follow-up (IRR=0.51, 95% CI 0.45 to 0.56) versus those leading to either an inpatient admission (IRR=1.05, 95% CI 0.84 to 1.31) or outpatient follow-up (IRR=1.23, 95% CI 0.93 to 1.64).

Conclusions The epidemiology of violent injury at our institution has changed over the last 10 years and is most marked in a reduction of visits at the weekend, and in those who leave without follow up.