• The committee was explicitly asked to consider therapeutic vaccines directed against chronic conditions such as autoimmune diseases and cancers.

Before embarking on a discussion of the methods used and the results obtained, it is necessary to present several caveats, as well as an historical perspective of vaccine development since publication of the 1985 report. This report, particularly the appendixes, is not the definitive cost-effectiveness analysis of any one specific vaccine-based prevention or treatment strategy. For several of the conditions under study in this report, detailed and elaborate economic analyses have been conducted and published by other researchers. For other conditions, there are none. Published studies, however, rely on different assumptions and apply different techniques. Thus, the committee could not depend on the results in the published literature for comparable estimates of the economic burdens of the conditions to be studied.

In order to compare candidate vaccines with each other, as this committee was charged with doing, the analyses needed to be conducted in comparable ways. Resource limitations meant that the committee focused on big picture descriptions of disease burden and potential gains from vaccine use. Because fine details regarding epidemiology, clinical presentation in individuals and within a population, and costs of care are not always readily available for the 30 candidate vaccines considered, the committee necessarily made many assumptions for the data used in the model. The committee assessed the published literature and the opinions of experts in the field, and then the members of the committee used their collective scientific and clinical judgment to estimate values when there was no published consensus.

Third, the committee knows that there will be readers of this report who will object to the individual numbers used in (or omitted from) the analyses. The committee urges all readers to consider the report and its appendixes as a whole and believes the following questions can be answered in the affirmative:

• If individual values for a given parameter of a given disease seem contrary to expectation, are they relatively correct compared to the values for the other conditions?

• Do the data used as a whole represent an unbiased qualitative synthesis of the published values?

• Are all conditions being treated similarly?

The committee offers the model as a tool for decisionmakers as they assess the needs for vaccine development and the opportunities for a vaccine development strategy. An untested model would be highly suspect. Therefore, the committee used data in the model that represent the consensus of the scientific community.