About the data
Definitions
Data sources
Mortality statistics and the indicators derived from them, such as life expectancy and infant mortality, are often cited as measures of a population's welfare or quality of life. They may be used to compare levels of socioeconomic development or to identify populations in need. Cause-specific mortality rates are useful both for placing the current health status of a population in an epidemiological context and for objective evaluation and planning in the health sector. As with all demographic indicators, mortality statistics should be used cautiously, with an awareness of the many difficulties involved in collecting and reporting them.
In developing countries mortality statistics from civil registers are notably defective. Estimates are derived by applying indirect estimation techniques to registration data, or from censuses or surveys, which also are subject to errors and biases. (See the notes to tables 2.1 and 2.2 for further discussion of demographic data.) Mothers may be reluctant to talk about children who have died, and may over- or underestimate the length of a year when answering survey questions about child deaths in the past 12 months (UNRISD 1977). And because many pregnant women die from lack of suitable health care, many maternal deaths go unrecorded, particularly in countries with remote rural populations. This may account for some of the low maternal mortality ratios in the table, especially for African countries. Differences in definitions may also affect the comparability of mortality data over time and across countries.
The available cause-specific mortality data are wholly inadequate; selected indicators are shown here to convey a sense of their potential utility. The main problem lies in determining the cause of death. In many developing countries, particularly in rural areas, trained medical personnel are not available to certify the cause of death. In such cases the cause of death is determined by a layperson, usually a (rural) health worker. The accuracy of such reporting is clearly lower than for cases that have been medically certified. Incomplete reporting introduces other potential biases, as does the use of hospital-based information to impute the health situation of a country as a whole.
Life expectancy and age-specific mortality rates for 1995 are generally estimates based on the most recent census or survey (see Primary data documentation). Maternal mortality ratios are drawn from diverse national sources. Where national administrative systems are weak, estimates are derived from demographic and health surveys using indirect estimation techniques or from other national sample surveys. For a number of countries maternal mortality ratios are derived by WHO and UNICEF (1996) using statistical modeling. Cause-specific mortality rates are standardized using the direct method: age-specific mortality rates are applied to the age distribution of a standard population-in this case the world-and the average is computed. This approach eliminates national differences in cause-specific rates due solely to the age distribution of the population. Cases in which the cause of death was ill defined are distributed among the three groups of causes of death in proportion to the number of deaths in each group.
• Life expectancy at birth indicates the number of years a newborn infant would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life.
• Infant mortality rate is the number of infants who die before reaching one year of age, per 1,000 live births in a given year.
• Under-5 mortality rate is the probability that a newborn baby will die before reaching age 5, if subject to current age-specific mortality rates. As with other demographic data (see notes to tables 2.1 and 2.2), 1995 estimates are often projected on the basis of the most recent census or survey (see Primary data documentation).
• Adult mortality rate is the probability of dying between the ages of 15 and 60, that is, the percentage of 15-year-olds who will die before their sixtieth birthday.
• Maternal mortality ratio is the number of female deaths that occur during pregnancy and childbirth per 100,000 live births.
• Mortality rate by broad cause is standardized for age using the world population as the reference population. l Deaths from communicable diseases include deaths from infectious diseases listed in the WHO's International Classification of Diseases, Ninth Revision (1977), plus influenza and pneumonia, nutritional disorders and anemia, and maternal (including abortion) and perinatal (occurring at about the time of childbirth) causes of death.
• Deaths from noncommunicable diseases include all causes of death other than communicable diseases and injuries and accidents.
• Deaths from injuries and accidents include deaths from all violent causes, whether intentional, unintentional, or unknown.
Mortality estimates are produced by the World Bank's Human Development and International Economics Departments in consultation with World Bank country departments. Important inputs came from the following sources:
Bos and others, World Population Projections 1994-95. Eurostat, Demographic Statistics. United Nations Department of Economic and Social Information and Policy Analysis, World Population Prospects: The 1996 Edition and Population and Vital Statistics Report. Demographic and health surveys from national sources. UNICEF, The State of the World's Children 1997. Maternal mortality ratios are drawn from: WHO, Maternal Mortality: A Global Factbook. WHO and UNICEF, Revised 1990 Estimates on Maternal Mortality: A New Approach. Mortality rates by cause are from WHO, World Health Statistics Annual.