Trimodal age distribution of frequent attendance at the emergency department: a descriptive analysis of national, English, secondary care data using a retrospective cohort

Objectives: Frequent use of emergency departments (EDs) places a considerable burden on healthcare systems. Although frequent attenders are known to have complex physical, mental health and social needs, national-level evidence on their characteristics and patterns of attendance remains limited. This study aimed to provide a comprehensive, population-level description of frequent ED attendance in England, with a focus on age-based subgroups.

Design: Retrospective cohort study.

Setting: EDs in England via the Hospital Episode Statistics and the Emergency Care Dataset data linked with primary care prescribing and mortality data, between March 2016 and March 2021.

Participants: The dataset received from National Health Service Digital contained approximately 150 million ED attendances by 30 million adult (>18 years) patients over the time period April 2016 to March 2021. A random sample of 5 million people was used for this analysis.

Outcome: measures The primary outcome was the number of attendances in each financial year by frequent attenders compared with the remaining patients, split by age bands. Patients were classified as frequent attenders if they had ≥5 or ≥10 ED attendances within a rolling 12-month period. Secondary outcomes included demographic, diagnostic and prescribing characteristics, as well as the number of different ED sites visited.

Results: A Gaussian mixture model was used to identify age-based subgroups. Descriptive statistics were used to summarise key features; 95% CIs were reported where applicable. Among 3.91 million unique adult ED attenders, there were 8.7 million attendances. Of these, 222 160 individuals (5.7%) had ≥5 attendances in a year, accounting for 12.6% of total attendances. A trimodal age distribution was identified, with three distinct peaks corresponding to ages 18–34, 35–64 and 65+. Frequent attenders were more likely to live in deprived areas and have a history of psychotropic or analgesic prescribing. Mental health diagnoses and polypharmacy were particularly common in the younger and middle-aged groups. Multisite attendance was uncommon, with over 80% of frequent attenders using only one ED site annually.

Conclusions: This national analysis reveals a trimodal age pattern among frequent ED attenders, with differing clinical and socio-demographic profiles across age groups. These findings highlight the need for age-tailored approaches to managing high-intensity ED use and inform targeted service development.