The socioeconomic and community influences on ‘potentially avoidable’ emergency admissions to hospital for the older people of London

Abstract

This PhD thesis explores the relationship between „potentially avoidable‟ emergency admissions to hospital and socio-economic and community care influences for the older population of London. The thesis first explores how socioeconomic conditions are associated with „potentially avoidable‟ hospital admissions ratios (SARs) for the older population at a local level. I bring in a new index of social fragmentation for older people, specifically designed to reflect potential increased need for older people. I show how a high amount unpaid care is important in explaining variations in admissions, but, unsurprisingly, it is strongly correlated with morbidity. I therefore create an index of high demand, incorporating these variables. Using multiple regressions I show how deprivation and a high demand for care predominate in explaining variation in 'potentially avoidable' emergency admissions for the older population of London. It then continues to show multiple admissions, or frequent admissions, where patients have more than one admission to hospital in any year from last admissions is also closely associated to the effects of deprivation, but not to a high demand for care, suggesting service factors from Primary Care Trusts (PCTs) and Local Authorities (LAs) may be in operation. I then explore the interactions between PCT and LA social services care funding and service provision at the scale of PCTs and local ward level socio-economic conditions, in relation to SARs using multilevel analysis. This demonstrates how PCT funding and social services spending on older people reduce the effects of deprivation at PCT level, however significant positive effects of deprivation remain both within and between wards. It shows how there are cross level effects of iii deprivation and PCT funding in operation with frequent 'potentially avoidable' emergency admissions for the older population of London. The final results chapter focuses on the question of „potentially avoidable‟ admission at a finer scale (individuals) within a district of London. Multilevel modelling is applied to explore the relationships of individual risks of avoidable admission with the patient‟s GP practice characteristics and quality of care provided by the patients‟ GP surgery, whether the patient receives homecare and the socio-economic condition of the area the patient resides in. The thesis concludes with a discussion of the overall findings concerning factors that seem to be driving potentially avoidable admissions for older people and the implications for policy.