10 under certain assumptions: as willingness to pay drops to $1,000 or below per IME prevented, the ranking changes more significantly, as shown in Table 9.4, with vaccines for certain diarrheal diseases rising in the rankings.

A fairly consistent middle-tier of vaccines occurs in the ranking under a variety of assumptions. In addition to those candidates that will contend for higher ranking under certain assumptions, this middletier includes vaccines for Streptococcus group A, M. leprae, V. cholerae, respiratory syncytial virus, parainfluenza viruses, and rabies (Vero cell derived or glycoprotein).

Most of the vaccines that consistently rank low would prevent diseases that are often serious, but mostly restricted to relatively small regions of the developing world. In such areas they may have more benefit than the widespread diseases that rank higher when the developing world is considered as a whole.

Additional sensitivity analyses, discussed below, can be performed to identify elements that may alter decisions.

Scientific opinion differs on some of the judgments incorporated into the proposed method, and uncertainty surrounds some of the data. The system has been applied by using the best estimates and most reliable data the committee could obtain, given its resources. The attempt to be explicit about certain estimates should not be interpreted as indicating that precise, unanimous, or certain comparisons are possible with existing methods or data, when the lack of data makes expert judgment necessary. The implications of these information gaps and differences of opinion about estimates are discussed more fully in Chapter 1. In this light, the committee suggests additional analyses and research to provide further information on the key elements that may alter decisions.

Ideally, to fully assess the effect of alternative IME profiles on the rankings, calculations should be conducted using the whole range of individual sets of IME values. However, because of resource and time constraints, this was not possible in the present study. The perspective adopted to illustrate application of the system was the median set of values from responses of health professionals in developing countries. A median set of values derived from U.S. respondents differed somewhat from the perspective used (see Chapter 4, Tables 4.7 and 4.8).

It is also possible to develop hypothetical age-neutral perspectives as was done for the committee’s first report (Institute of Medicine, 1985). The committee, however, does not endorse either set of median values or the age-neutral perspective for policy formulation. The effect of adopting various IME values is discussed in Chapter 4.

Selecting or constructing a small number of profiles that have distinct differences from the committee median or the age-neutral set would be a practicable way to further examine how various opinions on the undesirability of disease conditions might affect vaccine rankings. For example, IME profiles could be developed that show more or less