A CEA is the most general, full, economic evaluation and can be distinguished from the other subgroups because the effectiveness of the health program being evaluated is measured in natural units of effect. The most common unit of effect is length of life, such that an analysis would compare the cost per life-year saved between two potential health programs. In medicine, many clinical trials measure survival as the primary outcome and are therefore well suited to be used in a CEA. However, life-years saved may not be the most appropriate outcome measure if the program is designed to improve quality of life (QOL), such as is the case in ophthalmology. It is possible to base a CEA on a natural outcome measure that is assumed to be associated with better health. For instance, the cost per vision-year saved could be calculated in a CEA. Irrespective of the natural outcome unit chosen for the analysis, the purpose of a CEA is to compare the cost per natural health outcome between the health programs under consideration.

In a CMA, one assumes that the effectiveness of the health programs under consideration is equivalent. In a CMA, the cost of two or more health programs is compared and the program with the lower cost is considered the “preferred” option from the health policy maker’s perspective. For example, consider a situation in which an equivalence trial had shown that there was no statistically significant difference between two drugs for the treatment of glaucoma. An equivalence trial is a type of RCT specifically designed to test the hypothesis that a treatment option is as good as another alternative that has previously been proved to be efficacious. Here, the treatment option with the lower cost would be preferred.

In the real world, there are many situations where a number of important health outcomes such as length of life, QOL, and potential complications or consequences with treatment must be considered simultaneously to fully assess the effectiveness of a program. In addition, it may be necessary to directly compare programs that provide drastically different health benefits. CBA and CUA have been designed to account for different health outcomes and may be important in evaluating the true cost-effectiveness of a program and to allow for the comparison of programs or treatment interventions designed to effect health in different ways.

A CBA is also a special form of CEA, except that in this case the effectiveness of a program is measured monetarily. Costs can clearly be measured monetarily, but to measure the effectiveness in this way it is necessary to convert health outcomes into dollars. Consequently, a monetary value must be placed on all health outcomes pertinent to the analysis including length of life, QOL, and other health consequences. If the outcome of interest is simply years of life, then annual earnings per life-year saved can be defined as a monetary measure of effectiveness. However, when other factors such as QOL and potential complications must be considered, effectiveness is generally measured using a willingness-to-pay method. In this technique, a separate study would be conducted and subjects would be asked how much money they would be willing to pay to completely avoid a certain negative health outcome. A CBA should report results in the form of a net benefit in dollars, or the difference between the monetary values of the health benefits derived minus the cost of the health program.⁴

The major advantage with using a CBA is that health programs with widely varying health outcomes can be compared with each other. In addition, by definition a CBA compares the net benefit with the net cost of a health program so that one can determine whether the benefits outweigh the costs of initiating the program. However, assigning a price of a health outcome is a very difficult and controversial task that may only be possible in a limited number of situations. The main disadvantage with this method is that people from different sociodemographic backgrounds may be willing to pay vastly different dollar amounts for the same health outcome. In addition, whether a person lives in a country with a universal health-care system or whether the person has full health insurance will dramatically affect a person’s willingness to pay. These differences can drastically bias a study toward or against a certain demography of the population. Also, these differences make it very difficult to compare CBAs with each other.

A CUA is another type of CEA that allows different health outcomes of a program to be combined into one overall measure of effectiveness, thereby allowing for health programs designed to achieve different health outcomes to be compared. In addition, the difficult task of assigning monetary values to health outcomes is avoided. The effectiveness measure for a CUA is usually a quality-adjusted life year (QALY), where years of life are adjusted using utilities as a weighting factor. Measuring a health outcome in terms of QALYs allows for incorporation of both morbidity and mortality into one measure. Therefore, a CUA can investigate the cost per QOL adjusted year gained from the implementation of a particular health program compared with an alternative.

A utility is a measure of the strength of preference for a particular health outcome and has a theoretical foundation in economics and decision theory. Essentially, a utility is a measure of the value that a person places on a certain outcome or health state. Using utilities, the QOL associated with a particular health state that may have many important aspects can be measured using one method and can be reported with one value. Common methods of utility valuation are the time trade-off (TTO) technique, standard reference gamble (SRG), and rating scale. We will examine the details of utility theory later in the chapter.

A true CEA must measure both the cost and the effectiveness of a health program against the next best alternative. As mentioned earlier, the cost-effectiveness of a health program will usually be expressed in terms of an incremental cost-effectiveness ratio (ICER). Ideally, this will be defined as the difference in cost between the health program under question and the next best alternative (the numerator, or cost) divided by the difference in effectiveness between the health program under question and the next best alternative (the denominator, or effectiveness). It is important to differentiate between a marginal ICER and an average ICER. In the former, the cost and effectiveness both represent differences in costs and effectiveness between the treatment in question and the next best alternative, whereas in the latter the costs and effectiveness are measured independently of any alternative strategy.¹ Through the use of an ICER, it is possible to discern the true opportunity cost of a program, or the health outcomes that could be achieved by implementing the program of interest as opposed to the next best available option. By examining health policy in this way, it is possible to compare the cost-effectiveness of various health interventions in a consistent manner.

The first fundamental question that must be answered in an economic evaluation is the perspective that the decision maker is taking when conducting the analysis. For instance, the decision of whether photodynamic
therapy for patients with age-related macular degeneration is cost-effective could be drastically different depending on whether the decision is being made from the perspective of a for-profit third-party insurer or society at large. The insurer’s viewpoint may simply take into account the incremental cost of treatment and a health outcome in terms of vision-years saved or QALYs gained. However, society’s viewpoint may have to consider the cost of blindness that could include the utilization of many social and disability services provided by a government. The conclusion of a CEA could easily be different depending on the perspective taken. Unless a CEA is inherently being undertaken from a specific viewpoint (e.g., from the perspective of a third-party insurer or hospital), it has been recommended that the most general societal perspective be used.⁴ However, if the evaluation is undertaken from a societal viewpoint, it may be relatively easy and informative to provide other viewpoints in a CEA.

In designing a CEA, a clear problem that can be realistically answered through the analysis must be identified. The objective, method, and target population of the program alternatives must also be clear. A description of the effectiveness of the health intervention should be included as it is not logical to investigate the cost-effectiveness of something that has not been proved to be effective. To visualize health outcomes being modeled in the analysis, it may be useful to draw a flow diagram using a hypothetical cohort of people who begin the program and follow that cohort through every possible event outcome. Consultation with medical and economic experts is usually required.

Survival is a basic and very useful outcome measure in a CEA and can be the sole outcome or can be incorporated with other data. If survival does not fully explain the health outcome that is conferred by a program, then another method to measure effectiveness must be used. It is often easy and useful to base a CEA on an intermediate outcome measure that is assumed to be associated with better health. For instance, the cost per vision-year saved could be calculated for an ophthalmic intervention. If an intermediate health outcome is being used, a strong link with QOL or survival must be established.

Real-world situations commonly arise where a number of important health outcomes such as survival, QOL, and potential complications or consequences of treatment must be considered simultaneously to fully determine the effectiveness of a program. It is also desirable to be able to compare programs that provide drastically different health benefits. When the effectiveness of a health outcome is measured in dollars, the economic evaluation is known as a CBA. The most critical aspect in a CBA is valuing health using money; often the value of a health outcome, health state, or health scenario is measured by the willingness to pay or by annual earnings on the basis of expected length of life. As mentioned earlier in the chapter, there are many inherent biases involved in doing this.

The most comprehensive measure of health outcome combines both length and QOL into a QALY. A QALY can be conceptualized further by examining Figure 3.1, where the y-axis represents health-related quality of life (HRQL), the x-axis represents duration of life, and the curve represents various health states that a hypothetical person could potentially go through within a certain period. The area under the curve represents the QALYs associated with that particular set of health states over the specified time frame.

Measuring Health-Related Quality of Life. It has been generally accepted that QOL should be measured with a broad-based definition of health, accounting for physical/mobility function, emotional/psychological function, sensory function, cognitive function, pain, dexterity, and self-care.⁵ However, there remain many alternatives in measuring HRQL. HRQL measurement tools can be classified as generic, which attempt to measure overall HRQL, or specific, which focus on certain aspects of health such as disease, population, and function.⁶