In this post, we explain key concepts of cost-effectiveness analysis in healthcare and how they are linked: opportunity cost, marginal productivity (of a healthcare system), net benefit, and cost-effectiveness threshold.
In another post, to be published tomorrow, we will propose how to use cost-effectiveness thresholds and the existing estimates of marginal productivity in a cost-effectiveness analysis of policy options in healthcare. We also suggest some stock sentences to use when explaining results to audiences from an economics, medical, or lay background.
Cost-effectiveness analysis aims to inform decisions by providing information about the benefits and costs of various courses of action (such as funding a drug, package of care, surgery, or diagnostic test, among others) – all termed here ‘options’.
Knowing which option is cost-effective is easy if one intervention is both better (i.e. results in more benefits) and cheaper (i.e. results in lower costs) than the alternatives. It is less clear cut if an intervention is better but more expensive, or worse but cheaper.
The costs are actually the benefits that would be lost had we used the resources for the most highly valued available option – in economics, we call these lost benefits the opportunity costs. Of course, the most highly valued option depends on the perspective of who is valuing it. For simplicity, let’s assume for now that we are taking the perspective of the healthcare system.
An option is cost-effective if it gets us the most benefits net of the losses due to the opportunity cost. This principle can be easily adapted for options that are better but more expensive as well as with options that are worse but cheaper.
Let’s start with an option that is better but more expensive than the next best alternative. This option is cost-effective if the benefits gained outweigh the benefits lost due to the (opportunity) cost. The benefits lost are the benefits that we will have if we spend the option’s costs on something else.
What about an option that is worse but cheaper than what we currently have? We apply the same principle: this option is cost-effective if the benefits gained from the savings outweigh the benefits lost (given that it is worse). The benefits gained are the benefits that we will have if we spend the option’s savings on something else.
To know the opportunity cost, we need to know the rate at which a service (e.g. a publicly funded healthcare service) generates benefits (e.g. health) given the available resources (e.g. budget) – this is the productivity.
Most options cost a small amount relative to the size of the service’s budget; that is, the resources commanded by the decision between options are usually small relative to the overall pool of available resources. Therefore, we are interested in the productivity of the sector at the margin – that is the marginal productivity.
The marginal productivity may be different from the average productivity. In the case of healthcare, it is often thought that there are diminishing returns to scale. This means that, as inputs increase, outputs increase too but at a slower rate. In other words, the last dollar of health expenditure will yield less benefit than the first. Hence, in this case, marginal productivity is lower than average productivity.
In the context of cost-effectiveness analysis and opportunity cost, the marginal productivity is expressed in terms of the amount of monetary resources needed to produce a unit of benefit. For example, for the healthcare system, the unit of benefit could be a measure of health such as the quality-adjusted life year (QALY). The opportunity cost is then the option’s costs divided by the marginal productivity.
The net benefit is simply the difference between the option’s gains and its losses, given the opportunity cost. In an evaluation comparing various options, the cost-effective option is the one with the greatest net benefit.
The traditional approach to identifying the cost-effective option relies on the cost-effectiveness threshold. The cost-effectiveness threshold is the maximum additional cost that is deemed acceptable for an additional unit of health benefit.
Under this approach, to determine whether an option is cost-effective, we compare the option’s additional cost per unit of additional benefit (the incremental cost-effectiveness ratio or ICER) to the cost-effectiveness threshold. A better and more expensive option is cost-effective if the ICER is lower than the cost-effectiveness threshold. Conversely, a worse and cheaper option is cost-effective if the ICER is greater than the cost-effectiveness threshold.
Determining cost-effectiveness by comparing an ICER to the cost-effectiveness threshold will lead to the same conclusion as comparing the net benefit of the different options using the cost-effectiveness threshold to calculate the opportunity cost.
How have cost-effectiveness thresholds been defined?
Broadly, three types of approaches have been taken to set the value of the cost-effectiveness threshold to inform decisions in healthcare: ‘supply-side’ values; ‘demand-side’ values; and implied values or norms.
So-called ‘supply-side’ values correspond to the marginal productivity of the health care system for specific measures of health. At the time of writing, estimates of the marginal productivity of the health care system are available for the UK (see here and here), the Netherlands (here and here), Australia, Sweden, Spain, South Africa and China based on within-country data, and for most other countries based on estimates from cross-country data (see here, here and here).
So-called ‘demand-side’ values represent what the value of health and healthcare expenditure ought to be. These often take the form of either a) aspirational values of what value of ought to be placed on health, or b) the value that people are thought to place on health (also known as the consumption value of health).
Implied values or norms
Some countries and global organisations recommend using thresholds based on implied values (e.g. based on past decisions) or norms. For example, the National Institute for Health and Care Excellence (NICE) in the UK recommends using a threshold range of £20,000 to £30,000 per QALY to judge the cost-effectiveness of new options. The Institute for Clinical and Economic Review (ICER) in the USA recommends $100,000 or $150,000 per QALY are used [PDF]. In the context of low- and middle-income countries, the range of 1 to 3x gross domestic product per capita has often been used.
Some cost-effectiveness analyses consider not only the effects within the healthcare system but also the effects in other sectors, such as effects related to the social care sector, to the educational sector, and outcomes other than health (e.g. non-health dimensions of wellbeing) as well as costs borne by patients, family, and carers.
What about effects outside the healthcare system?
There is a strong theoretical case that costs falling on different sectors should not be added up if their marginal productivities, hence their opportunity costs, are likely to differ (see a discussion here). Therefore, the same cost in different sectors may result in different impact in terms of benefits. Also, the cost-effectiveness threshold set for one sector, say the healthcare system, is not necessarily applicable to other sectors.
Using sector-specific marginal productivities is potentially controversial as it could be interpreted as supporting policies to fund more productive sectors at the expense of less productive ones.
What does this all mean for your cost-effectiveness analysis?
Stay tuned for the second post, where we explain our ideas for using cost-effectiveness thresholds and marginal productivity in a cost-effectiveness analysis.