Cost-Effectiveness

The simplest type of cost-effectiveness relates deaths prevented to costs. For a measure of effectiveness we can use years of potential life gained  which are calculated as the difference between the expected durations of life with and without the intervention.

         

We see that as compared to the baseline, the total immunization program prevents about 190,000 premature deaths at an additional cost of US$110 million, for a cost-effectiveness ratio of US$579 per premature death prevented. The DPTT program is equally cost-effective  while the BCG program is the least cost-effective  If we added the BCG component to an existing DPTT, we would prevent about 16,000 additional deaths at an extra cost of US$13 million (US$797 per death prevented). Similarly, adding the DPTT program to an existing BCG program would prevent about 173,000 deaths at a cost of US$58 million

         

One can easily calculate YLGs—a useful tool in countries where data are scarce and the primary objective is reducing mortality. However, YLGs ignore benefits stemming from reduced morbidity and, hence, are highly biased against interventions for chronic diseases and other conditions with large morbidity-reducing effects. Although for large classes of diseases, especially common diseases of childhood, the morbidity-reducing effects are relatively small. A broader scope of comparisons among interventions affecting different diseases across the health sector requires a broader measure of effects that takes into account reduced morbidity and mortality.

Weighted Cost-Effectiveness

A measure of benefits that take into account reduced morbidity as well as reduced mortality requires a weighting scheme for the two benefits. The simplest scheme is healthy years of life gained  which weights morbidity and mortality effects equally. HYLGs are the sum of the years of life gained because of reduced mortality and morbidity, adjusted for disability

Cost-Benefit Analysis

Putting a dollar value on the benefits of health projects permits comparing them with projects in other sectors or with otherwise disparate benefits. However, assigning a monetary value to health benefits involves a great increase in complexity. Analysts must be careful not to unwittingly double count effects or include false benefits. Appendix 9B gives some examples of possible benefits from health projects.

         

Conventionally, one categorizes benefits in health as direct or indirect. One derives benefits primarily from morbidity and mortality changes, added quality of services, or gains in efficiency. Direct benefits are those that can be explicitly defined by a monetary value. Examples include avoided treatment costs or gains in efficiency of service delivery. Indirect benefits are those that are nonmonetary and can only be given an implicit monetary value, such as avoided loss of life or ill days, and changes in service quality.

Value of Life

Without question, the most difficult problem in evaluating benefits is placing an indirect value on life gained through reduction in mortality and morbidity. Many techniques have been suggested. The two most prominent are the human-capital approach and the willingness to pay approach. Under the human-capital approach, we view improvements in health status as investments that yield future gains in productivity. Useful as this approach may be to examine the effect of health on economic output, it ignores the consumption value of health. Even after retirement, for example, life has a value. Willingness to pay has become the accepted measure of the value of life. Individual willingness to pay has been estimated by implication from revealed preference studies examining earnings premiums for risky jobs or safety expenditures by consumers. These studies have all been carried out in industrial countries and need to be extended to developing country settings. Informatively, however, these studies consistently produce estimates of the value of life that are greater—usually several times greater—than the discounted present value of per capita income. Thus, in the absence of evidence from revealed preference studies in developing countries, the discounted flow of per capita income provides a highly conservative substitute estimate.