# High frequency emergency department use and heterogeneity of reasons for attendance by children and young people: a retrospective cohort study

## TL;DR

This study examines frequent emergency department visits by children and young people, finding patterns and reasons for repeated attendance over two years.

## Contribution

The study introduces a method to quantify heterogeneity in reasons for emergency department attendance using the Herfindahl index.

## Key findings

- 14.1% of children remained high-frequency emergency department users across two years.
- Children aged 8–12 had more injury-related visits and less varied reasons for attendance.
- Infants under age 1 had more illness-related visits and greater variation in reasons.

## Abstract

To quantify patterns of emergency department (ED) use over two consecutive 12-month periods among children aged 15 and under, and to assess heterogeneity of reasons for attendance in high-frequency users.

Population-based retrospective cohort study of routinely collected ED data.

EDs in the Yorkshire and Humber region, UK, from 31 March 2014 to 1 April 2017.

Children aged 15 and under with ≥1 ED attendance.

Proportion with ≥7 attendances over 2 years and heterogeneity of diagnostic reasons quantified by the Herfindahl index.

The cohort included 71 143 individuals. Although only 13.6% were high-frequency attenders in the first year, over half (55.1%) of these made at least one attendance in the second year. A subset (14.1%) remained high-frequency attenders across both years and were more likely to belong to the most deprived deprivation category. Children aged 8–12 were more likely to attend for injury-related issues and showed lower heterogeneity in reasons for attendance, while infants under age 1 had more illness-related attendances and greater heterogeneity.

A notable proportion of children and young people frequently attend EDs over a 2-year period. This study introduces a method for quantifying heterogeneity in reasons for attendance, which may support future predictive modelling using electronic health records to identify and support high-frequency ED users.