To examine how socioeconomic deprivation influences referral pathways to emergency departments (EDs) and to assess how these pathways affect subsequent hospital outcomes.

Retrospective observational study.

Emergency department of a large teaching hospital in the East of England, providing secondary and tertiary care.

482 787 ED attendances by patients aged 16 years and over, recorded between January 2019 and December 2023. Patients were assigned Index of Multiple Deprivation (IMD) deciles based on residential postcode.

Referral source (general practitioner (GP), National Health Service (NHS) 111, ambulance, self-referral, other), total ED time, 4-hour breach, hospital admission and unplanned return within 72 hours.

Substantial socioeconomic inequalities were observed in referral pathways. Patients from the most deprived areas were significantly less likely to be referred by a GP (4.7%) than those from the least deprived areas (14.7%) and more likely to arrive via ambulance (32% vs 24%). These differences persisted after adjusting for demographic, clinical and contextual variables. Ambulance referrals showed the longest ED stays, ranging from 347 to 351 min across IMD deciles (overall 95% CI 343 to 363) and the highest probability of 4-hour breaches (51%; 95% CI 50% to 53%). Self-referrals had the greatest rates of unplanned returns within 7 days (up to 7.1%; 95% CI 5.5% to 8.7%). In contrast, NHS 111 and GP referrals were associated with shorter stays, lower breach rates and fewer reattendances. Minimal variation in outcomes was observed across deprivation levels once referral source was accounted for.

Inequalities in how patients access emergency care, particularly reduced GP and NHS 111 referrals among more deprived groups, appear to underpin disparities in ED outcomes. Referral source captures important clinical and system-level factors that influence patient experience and resource use. Interventions to improve equitable access to structured referral pathways, particularly in more deprived areas, may enhance both the efficiency and fairness of emergency care delivery. Further research using national data is needed to assess broader policy implications and economic costs associated with differential access.