Use this URL to cite or link to this record in EThOS:
Title: Statistical models for small area public health intelligence on chronic morbidity
Author: Dutey-Magni, Peter
ISNI:       0000 0004 6496 2495
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Local indicators of chronic morbidity are needed to conduct needs assessments, plan health care services, allocate funds and monitor health inequalities. Model-based estimation is increasingly perceived as a possible avenue to enhance future local population health statistics methodology. In the UK, model-based small area estimation attracts particular interest both as a possible alternative to traditional population census enumeration and as a way to expand the range of indicators currently available. These methods however remain complex and still neglected in official statistics production. The present thesis brings applied contributions to this field by examining the potential of model-based estimation in England and Wales. First, a systematic literature review identifies the latest statistical developments and key methodological weak points. This informs the designs of three empirical academic papers designed around 2011 census health outputs. The first study builds two models predicting the crude prevalence of long-term limiting illness and self-rated health, and examines their reliability compared with 2011 census estimates. Secondly, an observational study analyses the spatial structure of morbidity for twenty age by ethnic groups in 2011 census long term limiting illness data. This assesses the potential to borrow strength across space and demographic groups, and to improve prediction efficiency. The final study proposes a survey design approach determining sample size requirements to achieve a desired level of statistical reliability. It is tested in a simulation study on 2011 census long-term limiting illness data. Together, these contributions provide applied testing work on well-established European population health indicators which inform the reliability of model-based estimation methods in a UK context.
Supervisor: Tzavidis, Nikolaos Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available