Objectives To describe key demographic and clinical characteristics of frequent attenders to two paediatric accident and emergency (A&E) departments.
Design Retrospective case note review of all children attending the paediatric A&E departments from 31st March 2009 to 1st April 2010.
Setting Two large District General Hospitals in London, UK with different models of A&E care; Hospital A paediatric led, Hospital B A&E led with paediatric input for under 1 year olds and all potential admissions.
Subjects 44,500 children aged 0 to 16 years
Outcome measure Frequent attendance defined by four or more attendances per child per year.
Exposure measures Demographic (age, sex, ethnicity, socioeconomic position (SEP) and distance lived from hospital) and clinical characteristics (presenting complaint and outcome) as assessed on first A&E attendance. Income deprivation affecting children index (IDACI) was derived from the postcode as a measure of SEP.
Methods Multivariable logistic regression was used to assess potential independent associations in each hospital.
Results 27% of children attending the A&E accounted for 50% of all attendance episodes. Three percent from hospital A (HA) (n=478) and four percent from hospital B (HB) (n=625) presented 4 or more times in the year. All results are presented fully adjusted for age, sex, ethnicity, IDACI, distance lived from hospital, presenting complaint and outcome at first attendance (table 1). Frequent attendance, was associated with greater odds for younger children (HA and HB), living close to the hospital (HA only) being male, living in an area of higher deprivation and admission on first presentation (HB only) Lower odds were found for frequent attendance for those presenting with injury (HA and B), reporting ethnicity as black, other or unstated (compared to white) (HB only) Certain first presenting complaints were associated with odds of frequent attendance, albeit numbers were small and confidence intervals wide.
Conclusion Key demographic and clinical characteristics were associated with more frequent attendance to paediatric A&E, but these differed for the two populations. Analyses of these types of data facilitates evidence informed service configuration