ing from the clustering of over 70 provincial mortality profiles. Four different age patterns of mortality for males and four for females are found in Russia, associated with different cause-of-death patterns. Variations within the country, however, are dominated by the difference between rural and urban areas: in rural areas, injuries and cardiovascular disease have a more pronounced impact on mortality in early and middle adult ages, while in urban areas, cardiovascular disease and neoplasm have a strong impact on older-age mortality.
In what ways are the Russian age-sex mortality profiles unique? Vassin and Costello find that among females, the predominant urban and rural age patterns of mortality in Russia are fairly similar to the Coale and Demeny ( 1966; Coale et al., 19831983) West and North regional model life tables, respectively. The male age patterns of mortality, in contrast, are not similar to the Coale and Demeny tables, but they are not unique to Russia, either. Rather, they have been seen before in earlier time periods in Hungary, Finland, and France, and are most similar to contemporary mortality patterns among African-American males in the United States.
The rural-urban differential in mortality is also examined in the chapter by Kingkade and Arriaga. In the NIS outside of Russia, rural populations generally have higher death rates than their urban counterparts; the exception is males in the Central Asian states, where there is reason for suspicion about the quality of mortality statistics, especially in rural areas. In the southern NIS countries, loss of life due to infectious diseases is primarily a rural phenomenon, while loss of life due to degenerative diseases and injuries is greater among the urban than rural populations of these states.
Information on mortality by age, sex, and causes of death has been used for decades to set priorities in the health sector. In many countries, the control of major risk factors and diseases has been achieved following the implementation of targeted control programs. Such programs are commonly designed through epidemiological analysis of the causes of death and examination of the cost-effectiveness of alternative control measures. To analyze the main causes of death, an aggregate indicator of premature mortality is used. The indicator used in two chapters in this volume is potential years of life lost. This indicator generally reflects the number of years of life lost to premature mortality, assuming that the deceased would otherwise have lived out a full life span to some expected age at death. The concept is useful for comparing the burden of disease, or health losses, as it reflects the loss of life taking into account the age of those who died, as well as disease-specific incidence and case fatality. In contrast, mortality rates weight deaths at all ages equally. The use of this measure is also useful for estimating the cost-effectiveness of health interventions, since it can be used to compare benefits across age and sex groups.