## Mortality indicators

(go to Outline)

The point estimate for the crude mortality rate and age-specific mortality rates when measured using surveys are usually presented with confidence intervals. Moreover, the rates themselves should always contain the population constant and time period. You cannot cite a mortality rate as just a number (for example, 2.3); this is as meaningless as giving the weight of something without the unit of weight (for example, it weighs 19). In addition, the presentation should include the time to which the estimate is applicable.

__Badghis Province, Afghanistan__

The crude mortality between 5 March 2001 and 23 March 2002 was estimated to have been 0.7 deaths/10,000/day (95% confidence interval: 0.49, 0.96).

Prospective death surveillance allows you to follow trends in mortality over time. The best way to display such data are in a graph of mortality incidence rates.

__Burundian refugee camps in western Tanzania, 1997__

(Source: International Rescue Committee)

This graph shows age-specific mortality rates for malaria in the first 7 months of 1997. It is clear that malaria mortality is seasonal, at least in 1997. Children less than 5 years of age have a higher mortality rate from malaria during the malaria season.

Mortality can also be presented as the number of excess deaths caused by the emergency (see section Number of excess deaths for more information). As with any important estimate from a survey, this point estimate should also be presented with confidence intervals.

__Iraq 2006__

The number of excess deaths between the U.S. invasion in April 2002 and July 2006 were estimated to be 654,965 (95% confidence interval: 392,979-942,636).

Cause of death is usually presented as the proportion of all deaths ascribed to each cause. The total of all the proportions should equal 100%. The presentation can be in the form of a table or pie chart. See section Measurement of causes of death in surveys for a discussion of important caveats regarding the measurement of causes of death in surveys.

__Eastern Democratic Republic of the Congo__

The distribution of causes of death as shown: