Table 10-2 Cost Considerations According to the Study Perspective
Study perspective | Cost examples | |
Societal | Public payer | Direct costs for services and good covered by the public payer. For example, in healthcare, these costs would include costs for hospitalization, primary care, rehabilitation, community care, long-term care, medical and other professional services, drugs, medical devices, diagnosis and screening services, facilities. |
Private insurer | Services and devices covered outside of the public insurance system | |
Direct beneficiaries, families and caregivers, replacement costs for employers | Copayments. Includes for example in the healthcare system copayments for services and drugs, cost of travel, premiums paid to private insurers, time spent for travel and treatments. | |
Productivity loss | Unpaid work of caregivers and lost productivity due to reduced capacity, long-term disability, early mortality | |
Source : Adapted from Agence Canadienne des Médicaments et des Technologies de la Santé. (2006). Guidelines for the Economic Evaluation of Health Technologies: Canada. 3rd Edition.
Time Horizon
In the reference case, the time horizon should be long enough to capture all potential differences in costs and outcomes associated with the interventions being compared (Fox-Rushby & Cairns, 2005; Gold et al., 1996). The same time horizon must be applied to costs and outcomes for analytical consistency. (Bryan et al., 2017, p. 31)
Future costs and benefits should be discounted. Different discount rates are used to test the robustness of the results. Debates exist regarding the appropriate discount rate to apply to costs and effects, particularly concerning its value and whether the same should be used for both costs and effects. Additionally, there are ethical discussions about whether long-term effects should be discounted, as discounting assigns a smaller value to effects occurring in the distant future. In fact, some argue that effects should not be discounted at all, citing principles of intergenerational justice.
Uncertainty
Sensitivity analysis should be conducted when uncertainty exists to make sure the results are robust.
Decision Criteria
An intervention is dominating when the costs are lower, and the results are more effective. Similarly, an intervention is dominated if the costs are higher and the results are less effective than the comparators’ (Brousselle et al., 2011c; Bryan et al., 2017). A dilemma arises when an intervention is more costly than the alternative yet produces better effects. In this case, the discussion can switch from efficiency to affordability (Brousselle et al., 2011c). Health economists will also recommend using the Incremental Cost-Effectiveness Ratio (ICER) which is “the difference in expected costs between two interventions divided by the difference in expected outcomes” (Bryan et al., 2017, p. 55):
(Costs of A – Costs of B) / (Effects of A - Effects of B).
In Cost-Benefit analysis, the incremental benefits will be compared to differential costs, with the difference between the two being the net social benefit (Brousselle et al., 2011c; Drummond et al., 1998): Social benefits = (Benefits of A - Benefits of B) - (Costs of A - Costs of B).
Social Return on Investment (or Return on Investment) studies calculate a ratio of monetary benefits over costs for each intervention. The higher the ratio, the higher the return.
An increasing number of attempts can be observed to express the economic, social, and environmental impacts of interventions in monetary terms within ROI and Cost–Benefit analyses. For example, studies are attempting to calculate the net social benefits of reducing greenhouse gas emissions or, conversely, the net social costs of increasing greenhouse gas emissions (National Center for Environmental Economics, 2023). Where a market exists, environmental impacts will be estimated using market prices. For example:
The loss or gain to the commercial fishing industry attributed to a change in water quality could be valued as the difference between the landed quantity before and after the change in water quality multiplied by the value of the species harvested. (Bohmholdt, 2014, p. 55)
When market values are not available, similar methods to the ones used in cost-benefit analysis are used (Bohmholdt, 2014). However, these valuation methods do not account for the intrinsic value of life or for concepts such as biodiversity, equity, happiness, etc.
Limitations of Existing Approaches
Synthetic indices, which are the end-product of ROI, Cost-Benefit and Cost-Utility analyses, have an appealing aesthetic quality and present a seductive format for decision-makers. However, they hide many methodological decisions in a domain where no perfect methods exist. For example, in Cost-Utility analysis, the estimation of utility varies according to the method used and whether they are measured directly or estimated using a classification system (Brousselle et al., 2011c; Brousselle & Lessard, 2011). Utility scores also vary according to the participants (Brousselle et al., 2011c; Brousselle & Lessard, 2011). Another example is the use of discounting and discount rates that are the subject of many ethical and methodological debates.
Furthermore, the rationale and methods used for giving a monetary value to life, as well as to social and environmental impacts, remain highly contested (Brousselle et al., 2016). For instance, giving a monetary value to all interventions’ results rests on the principle that everything can be valued, including intangible elements, and nothing has an intrinsic value that is incommensurable. Many economists attribute a monetary value to life, for example, but how would you value your own life if you were asked to give it a price?
Economic evaluations are aimed at making decision-making more rational. However, to facilitate those processes, we rely on elegant synthetic indicators; these can hide many ethically and methodologically controversial choices which, in the end, make the criterion used for decision-making problematic.
A Proposal for Useful Economic Evaluations for Planetary Health
This proposal is based on several observations. First, applying planetary health lenses involves considering dimensions that are not normally integrated into economic evaluations, such as intervention’s impacts on equity, prosperity, biodiversity, pollution, and land and water systems. Recent efforts have been made to move in this direction. However, these analyses present the same ethical and methodological problematic aspects as when giving an economic value to human life in economic evaluations. Second, as evaluators we have a responsibility to be transparent. The less complex the methodology, the more transparent the information tends to be.
Over the years, some researchers and health economists have advocated for Cost-Consequence analysis (CCA) as the preferred approach for economic evaluation. CCA is often considered a partial economic evaluation approach (Bryan et al., 2017; Rabarison et al., 2015) because it presents results in a disaggregated format; however, it’s this feature that makes it a good alternative. In CCA,
the predicted effects of the intervention (i.e., clinical outcomes including adverse events, quality of life impact, utility impact) are usually listed in tabular form without aggregating across different dimensions (Coast, 2004; Mauskopf et al., 1998; Williams et al., 2008). Other types of outcomes, such as humanistic outcomes (e.g. increased power over one’s life), can be included (Coast, 2004; Simoens & Laekeman, 2005). The same process is applied to resources (i.e., direct medical costs, hospital costs, direct non-medical costs, and indirect costs, such as loss of earnings), with attribution of values and quantities being shown separately (Mauskopf et al., 1998; Williams et al., 2008). CCA can include data from a variety of sources, including, but not limited to, clinical trials and observational and administrative databases (Mauskopf et al., 1998). Some authors stress that this type of analysis allows decision-makers to use their local data on costs and consequences (Eddama & Coast, 2008; Mauskopf et al., 1998). CCA can also present the impact of the intervention on different population groups (Mauskopf et al., 1998). Furthermore, it allows the consideration of different perspectives of analysis and the identification of all relevant dimensions for data interpretation (Mauskopf et al., 1998). The non-aggregated format offers a transparent reading, allowing decision-makers to see clearly what types of information are included and omitted, and where information is quantitative or qualitative (Coast, 2004). (as cited inBrousselle & Lessard, 2011, p. 836)
The costs and effects in CCA are usually presented in a tabular format, though can also be presented as a logic model. Brousselle and colleagues have used this methodology for conducting economic evaluations of public health interventions (Benmahrnia et al., 2017; Tchouaket et al., 2013) (see Figure 10.1). The logic model in Figure 10.1 describes the intervention by presenting all its resources, activities, and effects. The grey boxes identify the elements that were included in calculating the intervention’s benefits (Tchouaket et al., 2013).
Figure 10.1 Logic Model of Quebec’s Water Fluoridation Program
Source: Tchouaket, E., Brousselle, A., Fansi, A., Dionne, P.-A., Bertrand, E., & Fortin, C. (2013). The economic value of Quebec's water fluoridation program. Journal of Public Health, 21(6).
A logic model can be used in economic evaluation for a representation of the full range of results, from direct effects to indirect effects and externalities, as well as to include the logic sequence of results from outputs to outcomes to impacts. Results that can be readily and objectively converted into monetary values may be identified, valued, and compared to costs, enabling an assessment of whether the intervention offers a return on investment. However, a logic model presentation makes considering the whole range of impacts possible without making controversial methodological and ethical choices. It offers transparent information for decision-makers. Most importantly, the use of logic models in economic evaluation allows the inclusion of all relevant planetary health dimensions, alongside those specific dimensions decision-makers care about.
Conclusion
Economic evaluation is almost a field on its own. It is a field that is extremely normative and has many guidelines that are built on previous guidelines. The field has continued to evolve in the same direction for years, despite significant research studies showing that decisions based on economic evaluation recommendations have not necessarily been the more desirable nor the most socially acceptable. Many methodological problems are unresolved. Similarly, persistent ethical issues do not seem to influence the practice. However, it is possible to practice economic evaluation in a manner that allows for genuinely information-based and transparent decision-making while considering the full range of the intervention’s impacts.