Search

The purpose of this thesis is to explore the methodological and empirical issues relating to the economic evaluation of minimal access surgery (MAS). Given the likely increase in the utilisation of economic analysis in this area, it is crucial to explore whether the methods of economic evaluation have limitations in the face of the particular characteristics of MAS. The comparison of abdominal hysterectomy (AH) and transcervical endometrial resection (TCRE), for the treatment of menorrhagia, is used as a vehicle to develop methods in relation to MAS. Having reviewed the literature and issues relating to the economic analysis of this group of technologies, the empirical starting point of the thesis is the assessment of the limitations of economic evaluations alongside clinical trials, using a trial comparing AH and TCRE. Three major areas of weakness are identified, and alternative ways of addressing these weaknesses are explored in the remainder of the thesis. The first area of methodological development relates to the measure of benefit used in economic evaluation of MAS. In this clinical context, it is argued that the trade-offs that exist between MAS and conventional surgery, in terms of process characteristics and outcomes, should result in an important role for patients' preferences in the construction of a benefit measure. A cost-utility analysis using the standard quality-adjusted life year (GALY) is undertaken, using trial data augmented with valuation data from a further study. The lack of consistency between individual preferences and standard QALYs suggests a major weakness with this measure of benefit. The strengths and weaknesses of an alternative measure of benefit in cost-utility analysis - the ex ante healthy years equivalent (HYE) - are assessed based on a further valuation study. It is shown that it is feasible to elicit ex ante HYEs from patients and that this measure of benefit exhibits some consistency with other expressions of patients' preferences. However, the HYE is likely to impose a greater measurement burden than the standard QALY. The second area of methodological development in the thesis is the analysis of the generalisability of trial-based economic evaluation, given the limitations that often exist with the external validity of trials. A framework is developed within which trial and observational data can be synthesised. This facilitates the use of sensitivity analysis to explore the robustness of base-case (trial-generated) results to alternative sources of data, which may be more representative of routine practice. The third area of methodological development stems from the importance of patients' preferences in relation to MAS. This element relates to the use of methods to model and to evaluate management strategies which use patients' preferences to determine treatment allocation. It is concluded that preference-based decision making has the potential to be cost effective in relation to TCRE and AH, and MAS applications more generally. The thesis demonstrates the importance of continued development in the detailed methods of economic evaluation.

The concept of opportunity cost is fundamental to economic analysis. Opportunity costs exist because resources are scarce, and so their use in one way means they are not available for use in other ways. Economic evaluation methods that enable opportunity costs to be accounted for in decision-making in healthcare in high-income countries (HICs) are well-developed. However, methods are less well-developed to inform decisions in low- and middle-income countries (LMICs) despite the fact that it is in these countries where healthcare resources are most constrained and the potential health gains per additional resource spent on healthcare are greatest. This thesis provides methods to answer key policy questions for healthcare resource allocation in LMICs by extending existing frameworks and estimating key parameters required to inform them to enable health opportunity costs to be accounted for. It begins by providing empirical estimates of the marginal cost per unit of health produced by the healthcare system for LMICs, which were previously unavailable. How calculations of value founded on such estimates can be used to inform international guidelines about whether specific healthcare interventions should be provided in countries, whether and the extent to which these should be externally supported, and how to prioritise the development of new healthcare interventions for provision in LMICs is demonstrated. Next, the sources for empirical evidence to inform key parameter estimates for a framework of economic evaluation that accounts for the timing of costs and effects in a more appropriate way are identified, and the assumptions implicit in existing guidelines for economic evaluation are exposed and compared. Finally, a framework is developed to inform health benefits package (HBP) design that takes account of health opportunity costs and can analytically consider dimensions pertaining to equity, financing and implementation, while explicitly assessing key trade-offs in package design.

Cost-effectiveness analysis is increasingly being used to support decisions about the allocation of resources in health systems. However, decisions based on cost-effectiveness are usually made under imperfect and incomplete information. This thesis examines the value of understanding and characterizing heterogeneity for decision-making in healthcare. It proposes a methodological framework for a systematic cost-effectiveness subgroup analysis, providing guidance for identification and selection of subgroups. It suggests that the value of heterogeneity should be examined considering two dimensions: the value of making different decisions in different subgroups with current information (here termed static value); and the value of resolving parameter uncertainty conditional to a particular level of heterogeneity (here termed dynamic value). Finally, it provides empirical demonstration of such a framework through its implementation with a real case study. The study of heterogeneity for decision-making led to an examination of the policy agenda for individualization of care. This thesis also presents a conceptual framework to address two aspects of the implementation of individualized decisions. First, the implementation of unrestricted choices for treatment responds to a positive value judgement, which is based on the expected health loss (or gain) associated to it. This chapter presents a novel analytical approach to estimating this magnitude, based on the characterization of the joint distribution of potential outcomes. Second, it is acknowledged that implementing individualized decisions also requires a normative judgement. This refers to a broader question in economic evaluation in health, which is the type of values that society seeks to maximise. Three categories are proposed here (paternalist, altruistic and welfarist) that define alternative normative positions. The positive and normative elements of this framework are illustrated with a real case study. In conclusion, this thesis is a contribution to the development of methods for economic evaluation in healthcare, in particular, for understanding, characterizing and accounting for heterogeneity between members of the population.

This collection of papers aims to addresses why, and how, uncertainty should be incorporated within economic evaluations of health care technologies and how the results can be used to inform decision making. Section one provides a brief overview of the methodological and policy background of using economic evaluation to inform decisions about health care technologies. With the context for utilising economic evaluation to inform reimbursement decisions established, sources of uncertainty are considered at each stage of the evaluation. Section two discusses the identification of relevant evidence to inform estimates of the effectiveness of health care technologies. The key uncertainty addressed is how to assess whether the available evidence can be analysed to provide an unbiased estimate of the treatment effect. As multiple sources of evidence may exist, Section three explains how they can be combined to sum up the available evidence, and in Section four the techniques for characterising uncertainty within that combined analysis are examined. Sections five and six demonstrate why a characterisation of uncertainty is crucial in order to inform decisions about the need for further research and the consequences of making reimbursement decisions under uncertainty. Methods to establish the opportunity cost of uncertainty and sufficiency of the evidence base are reviewed in Section five with a view to informing decisions to acquire further research. In Section six the irreversible aspects of reimbursement decisions, as well as the interdependence of decisions about reimbursement and about further research is illustrated. These factors spell out why an economic evaluation that does not deal with uncertainty will be inadequate for informing reimbursement decisions. Finally, Section seven concludes the thesis and summarises the contribution of the presented papers and areas for further research.