Variation in admissions from hospital emergency departments in the English NHS

Background Emergency admissions to hospitals via the Emergency Department (ED) are a growing concern internationally due to their impact on healthcare costs. The aim of this study is to document the variation between hospitals in the English National Health Service in respect of their ED admissions before and after taking account of patient casemix and the primary care environment. Methods Fixed-effect regression analysis of 13,888,084 ED attendances from 9,063,518 patients from all NHS hospitals for 2018/19. The primary outcome was the hospital-specific likelihood of a patient being admitted after an ED attendance. Extensive controls were included for characteristics of the patient, their attendance and their GP practice . Results On average, 20% of ED attendances resulted in emergency admissions. Significant variation in admission rates was observed across English hospitals, both before and after adjusting for patient and system characteristics. Increased admission likelihood was found for: older adults, non-White ethnic groups, and patients from more deprived areas. patients arriving by ambulance or during periods when GP practices were closed. patients living closer to their GP. GP practices for which patients are less aware of extended hours and have higher chronic disease prevalence. After allowing for controls, considerable unexplained variation in hospital admission rates persisted, ranging from 12% to 35%, with an uneven distribution across the country. Conclusion There is substantial and persistent variability in emergency admission rates across English NHS hospitals, after controlling for a comprehensive range of patient and system characteristics. This highlights that a uniform approach to managing admissions may be ineffective and that tailored strategies, considering local resources, patient needs, and hospital-specific capabilities, are essential to reduce unnecessary admissions while ensuring equitable access to essential care. Future research should further explore the roles of alternative emergency care services, patient socioeconomic factors, and broader emergency care infrastructure.