Use this URL to cite or link to this record in EThOS:
Title: Trajectory and sequence analysis of administrative data
Author: Rao, Ahsan Mumtaz
ISNI:       0000 0004 9356 6109
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Health policy makers struggle to curb readmission rates despite various health policies. The National Health Service is spending more than half of its healthcare budget to manage emergency inpatient admissions. There is a subgroup of patients who use disproportionately more health resources and have higher rates of unplanned hospitalisation. They are defined as high-impact users. The characteristics and long-term healthcare use of high-impact users are poorly understood. Previous methods and models to identify and predict them have not performed well. More evidence is required to assess the causes of emergency admissions among them and to investigate whether they undergo a repeated cycle of similar events that trigger their long-term use of healthcare. Trajectory modelling and sequence analysis has been used in social and psychological sciences to understand changes in behaviour in the population. It has shown to have an advantage to categorise pupils based on common developmental pathways, and identify the chronological order of events in the life of the pupils. The hypothesis of the research is that trajectory modelling and sequence analysis can be applied to epidemiological administrative data to assess use of hospital care among high-impact users. Trajectory and sequence analysis was successfully applied to various patient cohorts with different medical conditions, using both hospital and primary care administrative datasets. Within each population cohort, discrete groups of patients with independent trends of hospital care use were identified. High-impact users accounted for a significant proportion of the patient population. They had persistently high readmission rates. Significant predictors associated with high-impact users were identified. High-impact users had worse outcomes than the rest of the patient population. They had distinct common sequences of causes of readmissions compared with other groups. Cardiopulmonary conditions were the main contributors to the sequences of readmissions among high-impact users in all patient populations.
Supervisor: Aylin, Paul ; Darzi, Ara ; Bottle, Alex Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral