Characterising non-urgent users of the emergency department (ED): A retrospective analysis of routine ED data.

The pressures of patient demand on emergency departments (EDs) continue to be reported worldwide, with an associated negative impact on ED crowding and waiting times. It has also been reported that a proportion of attendances to EDs in different international systems could be managed in settings such as primary care. This study used routine ED data to define, measure and profile non-urgent ED attendances that were suitable for management in alternative, non-emergency settings.We undertook a retrospective analysis of three years of Hospital Episode and Statistics Accident Emergency (HES A&E) data for one large region in England, United Kingdom (April 1st 2011 to March 31st 2014). Data was collected on all adult (>16 years) ED attendances from each of the 19 EDs in the region. A validated process based definition of non-urgent attendance was refined for this study and applied to the data. Using summary statistics non-urgent attenders were examined by variables hypothesised to influence them as follows: age at arrival, time of day and day of week and mode of arrival. Odds ratios were calculated to compare non-urgent attenders between groups.There were 3,667,601 first time attendances to EDs, of which 554,564 were defined as non-urgent (15.1%). Non-urgent attendances were significantly more likely to present out of hours than in hours (OR = 1.19, 95% CI: 1.18 to 1.20, P<0.001). The odds of a non-urgent attendance were significantly higher for younger patients (aged 16-44) compared to those aged 45-64 (odds ratio: 1.42, 95% CI: 1.41 to 1.43, P<0.001) and the over 65's (odds ratio: 3.81, 95% CI: 3.78 to 3.85, P<0.001). Younger patients were significantly more likely to attend non-urgently out of hours compared to the 45-64's (OR = 1.24, 95% CI: 1.22 to 1.25, P<0.001) and the 65+'s (OR = 1.38, 95% CI: 1.35 to 1.40, P<0.001). 110,605/554,564 (19.9%) of the non-urgent attendances arrived by ambulance, increasing significantly out of hours versus in hours (OR = 2.12, 95% CI: 2.09 to 2.15, P