Use of age-specific hospital catchment populations to investigate geographical variation in inpatient admissions for CYP in England

Use of age-specific hospital catchment populations to investigate geographical variation in inpatient admissions for children and young people in England: retrospective, crosssectional study.
BMJ Open 2018;8:e022339. doi:10.1136/bmjopen-2018-022339

• A technique to understand hospital catchment populations for children and young people by modelling the historical flow of emergency admissions from each local area. This allows the population each hospital serves to be identified and their demographic characteristics to be estimated. These catchment populations for children support work to understand if differences are related to how different hospitals manage a condition or because they serve a unique population.
• See also "There’s no longer a catch: a new way to measure how children and young people are treated”, Nuffield Trust comment, 25 April 2019

Abstract

Objectives To develop a method for calculating agespecific hospital catchment populations (HCPs) for children and young people (CYP) in England. To show how these methods allow geographical variation in hospital activity to be investigated and addressed more effectively.
Design Retrospective, secondary analysis of existing national datasets.
Setting Inpatient care of CYP (0-18 years) in England.
Participants Hospital Episode Statistics (HES) data were accessed for all inpatient admissions (elective and emergency) for CYP from birth to 18 years, 364 days, for 2011/2012–2014/2015. In 2014/2015, 857 112 admissions were analysed, from an eligible population of approximately 11.9 million CYP.
Outcome measures For each hospital Trust, the catchment population of CYP was calculated; Trustlevel admission rates per thousand per year were then calculated for admissions due to (1) any diagnostic code, (2) primary diagnosis of epilepsy and (3) epilepsy listed as primary diagnosis or comorbidity.
Results Estimated 2014/2015 HCPs for CYP ranged from 268 558 for Barts Health NHS Trust to around 30 000 for the smallest acute general paediatric services and below 10 000 for many Trusts providing specialist services. As expected, the composition of HCPs was fairly consistent for age breakdown but levels of deprivation varied widely. After standardising for population characteristics,
admission rates with a primary diagnosis of epilepsy ranged from 14.3 to 157.7 per 100 000 per year (11.0-fold variation) for Trusts providing acute general paediatric services. All-cause admission rates showed less variation, ranging from 4033 to 11 681 per 100 000 per year (2.9- fold variation).
Conclusions Use of age-specific catchment populations allows variation in hospital activity to be linked to specific teams and care pathways. This provides an evidence base for initiatives to tackle unwarranted variation in healthcare activity and health outcomes.