also illustrates important limitations of the approach, given the data currently available or likely to be available in the near future.
The cost model discussed below was developed for one set of reproductive health interventions, known collectively as the Mother-Baby Package. The costs considered here are only public-sector costs—for example, there is no cost estimate for patient travel time—and even these estimates are subject to considerable uncertainty and would vary among settings. Nevertheless, we believe that it is useful to show how costs may be affected by elements of program design and scale.
As noted throughout the report, interventions vary considerably in the degree to which their effectiveness in developing countries is known. All the interventions included in the package have at least passed a minimal test of plausibility, that is, they have been judged by experts convened by the participating agencies to reduce a substantial amount of maternal/neonatal deaths or disability. In other words, the interventions embedded in the package make good sense, to judge from the clinical record, but that record has not yet furnished the numerical estimates needed for precise ranking.1
The data available do support an examination of the public-sector costs associated with the package's interventions and activities and enable us to determine the likely sensitivity of the cost estimates to changes in key assumptions. We believe that such calculations are informative and suggest promising directions for future research.
Throughout, we present hypothetical cost estimates that are built upon a series of assumptions. Apart from the published international prices of various drugs and supplies, there are few numbers that are based on empirical data. This exercise is not unlike a population projection in spirit. It is, in the end, no more than an extrapolation of assumptions. Its value lies in highlighting those assumptions that appear to be most critical and that therefore merit closest scrutiny.
The elements of the intervention package, described by the World Health Organization (1994), do not span the full range of reproductive health discussed in Chapters 2-5. Rather, they are focused on the events
Some global effectiveness estimates for maternal mortality and neonatal mortality (percentage of deaths averted) are presented in the World Health Organization (1994:Table 1, Table 5). The major compilation of effectiveness studies (Chalmers, Enkin and Keirse, 1989), contains an assessment of a number of the interventions considered in the Mother-Baby Package.