Economics of Public Health Blog 7

In our last blog we talked about the importance of measuring the benefits which arise from interventions and highlighted one particular method for doing this - the quality adjusted life year (QALY). In that blog we also highlighted some potential limitations of using the QALY in the evaluation of public health interventions, so what other options are out there to measure benefits?
One potential option is Willingness to Pay (WTP). The concept behind WTP is fairly simple - the benefit (or utility) that an individual gains from an intervention can be valued by the maximum amount they would be willing to pay for that intervention. Given that it sounds fairly simple, how do we actually measure an individual’s maximum WTP for an intervention?

To measure WTP, individuals are presented with a description of an intervention and the outcomes that would be associated with that intervention. For example, if we were evaluating a screening service the description would include information like what the screening test would involve, how often it would occur, where it would take place and what potential information would be given following the screening test. After reading this description the respondent would then be asked, hypothetically, how much is the maximum they would be willing to pay for the service described.  This can be quite a difficult task as many of us have never thought about paying for health care so following the description of the intervention there is often some type of additional prompt for respondents to help them think about what their maximum WTP would be. This can take the form of a set of ‘payment cards’ where respondents are given  a set of cards, each with different amounts of money printed on  them and respondents decide if they would pay the amount shown to help them narrow down what their WTP would be. Alternatively they could just be given one money amount and asked if they would pay the amount shown in a straightforward yes or no response, although this requires much larger sample sizes and could lead to bias through respondents’ temptations to answer ‘yes’ when they do not really mean it.
The set up of the intervention description is important in all WTP studies but in the evaluation of public health interventions there are a few additional key features it is important to consider such as; the timeframe associated with the elicitation of public health WTP values, with the benefits of interventions typically occurring and then extending much further into the future and outlining all outcomes including those which may be non-health benefits occurring in non-health sectors. 

One of the advantages of using WTP is that it gives individuals the opportunity to value other potential benefits of the intervention and not just health gain as is the case with QALYs. This may especially be useful in the evaluation of public health interventions where we may not want (or it may not be possible) to reduce the outcomes into a change in health status as expressed in QALY terms. Thinking back to our last blog on QALYs we suggested some public health interventions where health outcome was not even the ultimate goal, such as services to improve access to contraceptives to teenagers or an afternoon club for older people. The more holistic approach of WTP could help us in valuing the benefits of these services.
The other great advantage of measuring the benefits of an intervention in monetary terms (rather than QALYs) is that, in principle, it allows us to compare the health intervention with those that are being proposed in other areas of the public sector. For example the Department for Transport evaluates proposed interventions using WTP to measure the benefits, as does the Department for Environment, Transport and Rural Affairs. This is useful when thinking about resource allocation across the public sector.

So if other areas of the public sector, such as transport or environment, use WTP as the most common measure of benefit in economic evaluation why has WTP not taken off in health?

Fundamentally many people do not feel comfortable assigning monetary values to things which they think are incommensurate with monetary valuation, such as human life or health. However, decision makers are implicitly doing this anyway; whenever they make a decision to limit the extent of availability of an intervention a value is placed on health or life.

Even for those who do feel comfortable in using monetary valuation, another concern with WTP which is frequently cited is its association with ability to pay and the implications of this for equity.  Indeed when asking people to give their WTP responses we encourage them to think about their current income because if they are not trading with an amount which is within their capacity their response has no meaning. So what does this mean for equity? Many public health interventions are targeted at the very poor who have a low ability to pay. If we surveyed only those at whom the intervention is directly targeted, this could result in a benefit to cost ratio which is not a true reflection of the intervention. WTP values should be elicited from a representative sample of the general public, which would include the full income distribution. In the analysis of the data, once WTP values have been estimated for a particular intervention it is important to examine whether people in higher income groups tend to choose one option more frequently than those in lower income groups and whether the WTP values of those in the higher group distorts the overall value given for one or other option. If so, ‘distributional weights’ can be applied in adjusting the raw values. The overall average WTP would then be applied to all regardless of income status – as is the case in work that has been conducted on using WTP to value lives and QALYs saved by safety and health interventions. It is also worth noting that QALYs, through deriving their values from risk and time trade off questions, also suffer from a similar distributional problem as WTP.

Although we may be able to refute some of the more common criticisms of WTP there remain challenges in the design of WTP studies. The long term validity and stability of values is not well understood. There are also concerns as to whether it is appropriate to just ‘add up’ WTP values across the sample or whether we should set out questions which specifically take a ‘societal perspective’ asking respondents to think about the opportunity cost of their valuations. More research is needed into both of these issues; although it has been shown that WTP can be designed to elicit individual values in such contexts.

So where does this leave us now?  In these latest blogs we have described two possible ways in which to measure the benefits of interventions – QALYs and WTP.  Both methods have their advantages and disadvantages and the decision on which one to use should be based on the objectives and context of the intervention being evaluated. If the outcomes of interest are primarily health related then QALYs could be your best bet but if non-health outcomes are important then you may want to consider using WTP. Of course, you may think that neither of these methods suits your evaluation and that it’s time for us economists to go back to the drawing board and design new methods for use in evaluating public health interventions or simply to leave all costs and benefits in disaggregated form as mentioned in an earlier blog.