Uncertainty about facts and heterogeneity in values

This chapter covers methods for describing how lack of knowledge impacts on the conduct and findings of distributional cost-effectiveness analysis (DCEA). It also sets out methods for describing how different value judgments can alter the findings. It explains why and how to distinguish uncertainty about facts from heterogeneity in values, and the role of each in informing decisions. It shows how the standard tools of uncertainty analysis in economic evaluation—including deterministic and probabilistic sensitivity analysis, and value of information analysis—can be applied to DCEA to provide information about uncertainty in the estimated health distributions and summary measures of equity impact. The chapter also shows how to use deterministic sensitivity analyses to investigate the implications of alternative value judgments and inequality metrics for DCEA findings and recommendations.

Health by disease categories

This chapter shows how to quantify and compare the baseline health between and within disease categories using three main health concepts—current health, future health, and lifetime health—and shows how each concept can be measured using three main metrics—absolute health achievement, absolute health shortfall, and relative health shortfall. Different health concepts can have quite different equity implications about which disease group is worse-off in terms of health or has greater ‘severity of illness’. Metrics based on current or future health might be useful for a decision maker who believes that past health is water under the bridge. Metrics based on lifetime health might be useful for a decision maker who believes that priority should be given to people who suffer a greater burden of illness from the perspective of the whole life course from birth until death.

Designing a distributional cost-effectiveness analysis

This chapter is a guide to designing a distributional cost-effective analysis (DCEA) for research commissioners and producers. It describes the main components of DCEA and how they can be combined in different ways to inform decision-making in various contexts. The appropriate design of a DCEA will depend on the objectives of the relevant decision-making authorities; the political, economic, and institutional environment, and the prevailing social norms in the population being served; as well as the resources available for the analysis. We start by reviewing key questions about the decision-making context and illustrate how the answers to those questions can shape the study design, using our training example of nicotine replacement therapy in England. We then present the equity-efficiency impact plane, a visual tool for illustrating the relationship between the impact of a decision on efficiency (e.g. net health benefit) and equity (e.g. reduction in unfair health inequality). We then go through the main components of DCEA, including simple, low-cost forms of equity analysis that may be useful when detailed bespoke modelling of distributional consequences is not a practical option.

Costs and health effects

This chapter shows how to model the distribution of costs and health effects of a health programme by equity-relevant characteristics. We introduce the ‘staircase of inequality’, a framework for thinking about the main steps in the causal pathway leading to differences in costs and effects by equity-relevant characteristics. We then illustrate the application of this framework with the example of public funding options for nicotine replacement therapy (NRT) in England. In this case, the main steps along the staircase of inequality are: need (who smokes), receipt (which smokers use smoking cessation services and how delivery costs vary), short-term effect (which service users succeed in quitting), and long-term effect (which short-term quitters become long-term quitters and how many healthy years they gain). At each step, we model differences in costs and effects by socioeconomic status and geography. We then combine this information to simulate the distribution of lifetime costs and health effects. The framework can also be applied to other equity-relevant variables, such as ethnicity or gender, other concepts of health, and other time horizons. We discuss the key assumptions and data requirements for implementing this approach, highlighting potential data sources and providing guidance for setting these parameters when evidence is scarce.

Health by social variables

This chapter explains how to estimate baseline distributions of lifetime health by social group that can be used in equity-informative economic evaluation. It shows how social inequalities in health at baseline can have an influence on the relative inequality impact funding new health programmes and how the relevant distribution of health is defined by social value judgments about what constitutes unfair health inequality. A step-by-step guide to estimating distributions of health-adjusted life expectancy is provided, covering common analytical methods and measures of health-related quality of life. It also discusses appropriate sources of data and includes several practical examples demonstrating how baseline distributions have been produced in different settings.

Level-dependent equity weights

This chapter explains how you can use level-dependent social welfare functions (SWFs) to evaluate health distributions in a manner that is founded on explicit, challengeable, and consistent ethical principles. A level-dependent social welfare function (SWF) weights health gains for one person or group relative to another as a function of their absolute health levels. By contrast, a rank-dependent SWF weights health gains for one person or group relative to another as a function of their health ranks in the wider population. A level-dependent SWF is also called a prioritarian SWF. The basic prioritarian principle is a specific value judgment giving priority to the worse-off based on their absolute level of health. The aim of this chapter is to describe level-dependent SWFs and how they can be used to conduct equity-efficiency trade-off analysis. A SWF can be consistent with many different sets of weights, which vary according to the chosen value of an equity parameter. We show how level-dependent equity parameters can be used to perform a systematic sensitivity analysis of how different degrees of concern for the worse-off have different policy implications.

Principles of health equity

This chapter shows how informal normative concerns about health equity raised by decision makers can be translated into formal health equity objectives that are amenable to quantitative analysis using distributional cost-effectiveness analysis (DCEA). It also clarifies the kinds of ethical concerns that can and cannot be addressed by DCEA. Four ways of thinking about ethics are distinguished: value maximizing, moral rights, fair shares, and fair processes. When considering whether and how DCEA can be used to address a specific equity concern raised by decision makers, it is important to clarify which of these four ways of thinking underpins the concern. DCEA methods for evaluating distributions are based on the value maximizing way of thinking, but the information provided by DCEA about the distributional consequences of decisions can also help to address equity concerns relating to moral rights, fair shares, and fair processes.

Health opportunity costs

The net benefit of a health programme is defined as the benefits it generates minus its opportunity costs. Accounting for opportunity costs—the value of resources in their most highly valued alternative use—is an essential part of economic evaluation. This chapter discusses the role of health opportunity costs in equity-informative economic evaluation and describes some approaches to quantifying both total health opportunity costs and their distribution. Alternative methods are proposed where data availability is limited, emphasizing careful consideration of the funding context along with scenario analyses to examine the implications of alternative levels and patterns of health opportunity cost.

Describing equity impacts and trade-offs

This chapter introduces the basic measurement concepts needed to understand and critically appraise distributional cost-effective analysis (DCEA) findings about equity impacts and trade-offs. It starts by introducing the key questions that research users need to ask about DCEA findings, drawing on basic concepts from the literature on measuring inequality and social welfare. It then introduces two diagrams for summarizing trade-offs between equity and efficiency impacts. Close attention is paid to the value judgments underpinning different equity impact metrics, with a view to helping research users engage in well-informed conversations about those value judgments.

Rank-dependent equity weights

This chapter explains how to evaluate the distributional consequences of health programmes by comparing inequality penalized measures of population health. The approach taken is founded on use of rank-dependent social welfare functions (SWFs) that evaluate population health outcomes in terms of equity-weighted average health, using weights that depend only on rank in the distribution of health. The underlying ethical principles and resource allocation implications are compared with those of the level-dependent welfare function approach introduced in the next chapter. Conditions under which quantitative results derived from a rank-dependent SWF are consistent with the preference ordering of programmes established by dominance analysis are identified. The approach is easily extended to allow for aversion to health inequality in relation to an equity-relevant variable, such as socioeconomic status.