Title: Modelling Issues in Institutional Long-term Care: Placement, Survival and Cost

Author: Haifeng Xie

Year: 2004

Abstract: The population in the UK is ageing rapidly, with nearly a third of the population projected to be over 65 years of age by 2030. Long-term care (LTC) is becoming an even more important issue in this ageing society. In Great Britain, local authorities have a major role in the running and delivering of LTC services at the local level. Therefore, it is important for local authorities to have a sound understanding of the behaviour of the LTC system. Issues of particular interest are: who goes into care and where they go, how long they stay, and how much they cost. Since the answers to these questions will help to ensure the efficient and effective management of LTC at a local level. This thesis addresses these concerns by developing quantitative models focusing on publicly funded residents in LTC. More specifically, we construct prediction models for placements in LTC based on a hierarchical view of the decision making process. A continuous-time Markov model is used to model the movement of publicly funded residents within and between residential care and nursing care, and we derive a procedure for model fitting based on the framework of aggregated Markov processes. We further develop a generic framework for costing current commitments. The costing framework translates a survival model into measurements that are meaningful to budget planners in local authorities. Applications and validations of these models using empirical data provided by the Housing and Social Services Department of the London Borough of Merton (UK) suggest that these models are useful in providing insights into the behaviour of the LTC system. The models developed in this thesis complement other research in this area in providing a full picture of the overall behaviour in the system of LTC.

      Given the important role LTC plays in the efficient running of acute hospitals and the significant costs associated with maintaining elderly people in care homes, the models and findings in this thesis should be of great interest to government departments, insurance companies, health and social services planners, and purchasers and providers of LTC.