Health-related quality of life; chronic obstructive pulmonary disease and smoking

Abstract

This thesis explores and investigates the challenges around measurement of Quality of Life (QoL) / Health State Utility Value (HSUV) in Chronic Obstructive Pulmonary Disease (COPD), as a chronic disease and its major risk factor, smoking. This thesis is based upon four separate studies, which present original research of 1) systematic literature review on HSUV in COPD, 2) application of the HSUVs in COPD disease progression models, 3) economic evaluation study alongside a clinical trial aimed to improve HSUV in COPD and 4) econometric analysis of the effect of smoking habit transition on the HSUVs.

The first study investigates the mean HSUVs in COPD patients in general and specifically in each stage of the disease by using systematic literature review and meta-analysis of studies which reported patients-level utility values elicited by EQ-5D. In order to explore the degree of heterogeneity around the utility values, effects of a variety of clinical and study characteristics have been examined through subgroup analyses. This study represents one of the first meta-analysis and subgroup analysis of HSUV in COPD. It demonstrates considerable inconsistency in utility measures among COPD-related published literature. This study highlights that in case of high level of heterogeneity, appropriate sensitivity analyses are recommended for more accurate health economic appraisals.

The second study concerns the compatibility of available COPD progression models with good practices guideline for decision analytic modelling. This study conducts a systematic review of the HSUVs assigned to the different stages of COPD used in modelling studies and compares these with summary measures from meta-analyses of available utility studies. This study demonstrates that on average, COPD decision models used higher values than estimated mean HSUVs from the meta-analysis of the patient-level data. The study suggests that improvement in the consistency of modelling studies may be achieved if published recommendations on good modelling practice, especially the data identification, are followed closely as suggested.

The third study is an economic evaluation of the telephone-based cognitive behavioural (TB-CBT) therapy for depression/anxiety comorbidities in COPD patients. Alongside a clinical trial, a cost-utility analysis is performed to measure cost and quality-adjusted life years gained based on the Assessment of Quality of Life (AQoL-4D) measure as a preference-based HSUV scale. This study shows that TB-CBT can be considered as a cost saving approach. This study, by using the concept of loss aversion from prospect theory which is based on individual preference, provides a distinctive interpretation of the incremental cost-effectiveness ratio (ICER) in the south-west quadrant of ICER plane.

The fourth study elucidates the effect of the transition from “Smoker” to “Ex-smoker” on QoL (measured by SF-36) in the general Australian population. Panel data from thirteen waves of a nationally representative longitudinal survey of Household Income and Labour Dynamics of Australia is used and piecewise two-way fixed effect linear regression models are adapted. Of the eight SF-36 dimensions, only physical health factors showed pervasively and significant improvements after the smoking transition, irrespective of age and sex and other related time-invariant covariates. This study is one of the first studies analysing the relationship between smoking and QoL measures in general population, taking the advantages of panel data which provides unique opportunity to account for individual heterogeneity and focuses on within-person changes in QoL as smoking status change while controlling for unobserved time-invariant individual characteristics (fixed effects) on observed covariates.