In each of these cases, solid evidence - summarized in the respective case studies - confirms that the impact on health is attributable to the specific health efforts rather than to broad economic and social improvements. In some instances, this confirmation comes through a randomized experimental design, which permits the comparison of the health of people who were included in a particular program with the health of people who have similar baseline characteristics and yet did not participate in the program. Such experimental designs are rare but not unknown: In Mexico, for example, the Progresa program of income transfers was scaled up in a way that was explicitly designed to assess the program’s impact.
In other instances, the confirmation comes from a composite of information about health changes that occurred simultaneously with a program’s implementation. In Sri Lanka, for example, the changes in specific causes of maternal mortality, such as hemorrhage, coincided with targeted improvements in health systems, such as the introduction of transfusion services. In The Gambia, the reduction of a disease that causes meningitis in children was so dramatic and so well-documented following the nationwide introduction of Hib vaccine that no doubt exists about the cause of the epidemiological change. And in other cases, such as the Bangladesh family planning program, statistical analyses that separate out different influences on the health outcomes provide the grounds for claims of success.