Search the site

The University of Manchester Library

In April 2016 Manchester eScholar was replaced by the University of Manchester’s new Research Information Management System, Pure. In the autumn the University’s research outputs will be available to search and browse via a new Research Portal. Until then the University’s full publication record can be accessed via a temporary portal and the old eScholar content is available to search and browse via this archive.

Incorporating multiple conditions within decision-analytic modelling for use in economic evaluation

Thompson, Alexander

[Thesis]. Manchester, UK: The University of Manchester; 2019.

Access to files

Abstract

Decision-analytic models are the preferred method by the National Institute for Health and Care Excellence for generating evidence on the cost-effectiveness of competing interventions. Increasingly patients are experiencing multiple conditions simultaneously which poses key challenges to this model-based approach. The aim of this thesis was to contribute towards methods for incorporating multiple conditions within decision-analytic models for economic evaluation. Large administrative and survey datasets were used including NHS Patient Reported Outcome Measures, GP Patient Survey and Health Survey for England to explore the relationship between multiple conditions and health-related quality of life (HRQoL). Regression methods were used to quantify the relationship between increasing numbers of conditions and HRQoL. The analysis explored whether co-occurrence of certain conditions, in particular depression, with physical health conditions lead to larger ('synergistic') reductions in HRQoL than would otherwise be expected. Competing methods (including the additive, multiplicative, minimum, linear index and adjusted decrement estimator) for predicting utility values for populations with multiple conditions were evaluated within the GP Patient Survey data. The linear index, which is a parametric estimator, was externally validated in Health Survey for England data. Conceptual challenges associated with modelling the cost-effectiveness of treating patients with multiple conditions were explored in two separate decision-analytic models. A cohort-based state transition model assessed the cost-effectiveness of statin therapy for the primary prevention of cardiovascular disease when patients had an ongoing direct treatment disutility associated with statin therapy. A patient-level model incorporated condition-interactions by assessing the cost-effectiveness of ten antidepressants for patients with depression at risk of cardiovascular disease. A synergistic relationship was observed between depression and the number of reported physical health conditions leading to large falls in HRQoL as patients become increasingly multimorbid. When predicting utility values for patients with multiple conditions, the multiplicative method and linear index performed best with lowest bias and highest precision but both were dependent on key assumptions regarding the inclusion of co-morbidities and the selection of a healthy baseline. Estimates for all methods were biased in the presence of depression, over-estimating predicted utility values compared to true utility values. The linear index was robust in the external validation when the same assumptions were used in the validation cohort as in the derivation cohort. Overall, including multiple conditions within decision-analytic models for economic evaluation was found to be feasible but challenging. The development of methods within this thesis will allow future analysts to estimate the HRQoL for patient populations with multiple conditions where no such data currently exists.

Bibliographic metadata

Type of resource:
Content type:
Administered thesis
Form of thesis:
Alternative format
Type of submission:
Doctoral level ETD - final
Thesis title:
Incorporating multiple conditions within decision-analytic modelling for use in economic evaluation
Degree type:
Doctor of Philosophy
Degree programme:
PhD Medicine 6yr (PHHSR)
Publication date:
2019-11-06T09:29:04
Institution:
The University of Manchester
Location:
Manchester, UK
Total pages:
271
Abstract:
Decision-analytic models are the preferred method by the National Institute for Health and Care Excellence for generating evidence on the cost-effectiveness of competing interventions. Increasingly patients are experiencing multiple conditions simultaneously which poses key challenges to this model-based approach. The aim of this thesis was to contribute towards methods for incorporating multiple conditions within decision-analytic models for economic evaluation. Large administrative and survey datasets were used including NHS Patient Reported Outcome Measures, GP Patient Survey and Health Survey for England to explore the relationship between multiple conditions and health-related quality of life (HRQoL). Regression methods were used to quantify the relationship between increasing numbers of conditions and HRQoL. The analysis explored whether co-occurrence of certain conditions, in particular depression, with physical health conditions lead to larger ('synergistic') reductions in HRQoL than would otherwise be expected. Competing methods (including the additive, multiplicative, minimum, linear index and adjusted decrement estimator) for predicting utility values for populations with multiple conditions were evaluated within the GP Patient Survey data. The linear index, which is a parametric estimator, was externally validated in Health Survey for England data. Conceptual challenges associated with modelling the cost-effectiveness of treating patients with multiple conditions were explored in two separate decision-analytic models. A cohort-based state transition model assessed the cost-effectiveness of statin therapy for the primary prevention of cardiovascular disease when patients had an ongoing direct treatment disutility associated with statin therapy. A patient-level model incorporated condition-interactions by assessing the cost-effectiveness of ten antidepressants for patients with depression at risk of cardiovascular disease. A synergistic relationship was observed between depression and the number of reported physical health conditions leading to large falls in HRQoL as patients become increasingly multimorbid. When predicting utility values for patients with multiple conditions, the multiplicative method and linear index performed best with lowest bias and highest precision but both were dependent on key assumptions regarding the inclusion of co-morbidities and the selection of a healthy baseline. Estimates for all methods were biased in the presence of depression, over-estimating predicted utility values compared to true utility values. The linear index was robust in the external validation when the same assumptions were used in the validation cohort as in the derivation cohort. Overall, including multiple conditions within decision-analytic models for economic evaluation was found to be feasible but challenging. The development of methods within this thesis will allow future analysts to estimate the HRQoL for patient populations with multiple conditions where no such data currently exists.
Keyword(s):
Thesis main supervisor(s):
Thesis co-supervisor(s):
Degree grantor:
Language:
en

Institutional metadata

University researcher(s):
Academic department(s):

Record metadata

Manchester eScholar ID:
uk-ac-man-scw:322321
Created by:
Thompson, Alexander
Created:
6th November, 2019, 09:29:04
Last modified by:
Thompson, Alexander
Last modified:
23rd December, 2019, 12:16:54

Can we help?

The library chat service will be available from 11am-3pm Monday to Friday (excluding Bank Holidays). You can also email your enquiry to us.