WHO Mortality Database
Interactive platform visualizing mortality data
Q. What are the sources of the cause-of-death statistics?
A. The cause-of-death statistics are from country civil registration and vital statistics systems. When a death occurs, this event is registered at the local civil registry with information on the cause of death. In some countries, it is the national statistics office, in others it is the ministry of health or registrar-general's office that is responsible for compiling the cause-of-death data for submission to WHO every year.
Q. Are the deaths medically-certified?
A. WHO only publishes here the medically-certified deaths. Diagnoses on causes of death by lay people are not included.
Q. If the data are not complete, have the data been adjusted for under-coverage?
A. The data presented are as submitted by the countries to WHO. No adjustment has been made to account for incompleteness.
Q. What kind of corrections have been made to the data?
A. WHO verifies that the data submitted are coded with the official ICD codes. If non-official codes are used, they would be replaced with the most appropriate official codes.
Q. What are the sources of population data?
A. Some countries submit regularly their mid-year population data. However, for those who do not submit such data, the population estimates from the United Nations Population Division have been used.
Q. Which standard population have been used to calculate the age-standardized death rates?
WHO has used the WHO world standard population. For more info consult https://cdn.who.int/media/docs/default-source/gho-documents/global-health-estimates/gpe_discussion_paper_series_paper31_2001_age_standardization_rates.pdf
Q. How are the number of deaths and population presented?
A. They are all shown as absolute numbers.
Q. How are the death rates presented?
A. They are shown as death rates per 100 000 population.
Q. How are the age-standardized death rates (ASDR) presented?
A. They are all shown as ASDR per 100 000 of the standard population.
Q. Why are some rates not presented although there are number of deaths?
A. Rates are not calculated for countries with a population < 90,000 in 2019 as for small countries any slight variation in numbers would affect to a large extent the rates.
Q. The list of countries does not seem to include all the countries in the world. What are the reasons?
A. Simply because some countries do not report mortality data to WHO. However, for some countries, the data they sent to WHO are not in standard ICD codes or do not have ICD codes, so they cannot be shown here. In many low-resourced countries, the cause-of-death information is difficult to obtain, mainly because the system for recording such information is not functioning or inexistent. In addition, one of the big problems is the lack of medical certifiers to complete the death certificates. When countries submit data, an assessment of the data completeness is done in WHO. The data would only be displayed in the portal if they are estimated to include at least 65% of all deaths occurring in a country.
Q. Why are some latest years of data missing?
A. Countries usually submit data to WHO within 12-18 months after the closure of their records for the calendar year. Data checking, compilation and verification takes considerable time at the country level.
Q. Are data for countries displayed mutually exclusive?
A. All data for countries displayed are mutually exclusive. For e.g. data for Martinique, a French overseas department are not included in the data for France.
Q. What can cause a sudden change in number of deaths from one year to another?
A. There are several reasons. It could be a real drop in mortality due to effective interventions such as enforcement of speed limit contributes to rapid decline in road traffic deaths. Data completeness may have improved over time. Changes from ICD revisions could lead to trend disruptions in some causes as latest ICD revisions have more detailed causes mainly because of increased medical knowledge. Therefore it is not always possible to map exactly some causes across ICD revisions as described in this example.
Deaths due to drug overdoses and alcohol poisoning are commonly mis-assigned to ICD codes for accidental poisoning by drugs and alcohol (e.g. ICD-10 codes X40-X45). Most of these deaths are a result of overdose deaths among individuals with a drug or alcohol use disorder, and therefore should have been assigned the ICD code for the relevant drug use disorders. To address this, ICD-10 codes associated with alcohol and commonly misused drugs were grouped in the drug use disorder category (X41-X42, X44) or alcohol use disorder category (X45). The codes associated with medications that are not commonly misused (X40,X43) were assigned to the poisoning category. It was not possible to carry out the same mapping for misused prescription medications for data coded using prior revisions of ICD due to the lower level of detail in the cause categories. This results in artificial discontinuities in the estimated number of poisoning and drug use deaths in some countries.
Q. When I download the data, sometimes they are not in standard 5 year age-group, what is the reason?
A. Generally the data are available in 5 year age-group, but sometimes countries report data by 10 year age-group. In this case, the data will be shown in the first of the combined age-groups. For e.g. if data are reported for age 10-19 years, then the data are shown in age 10-14 years and the column for age 15-19 years is left blank. Similarly if the last age-group is 75 years and above, the data are shown in ages 75-79 years and the remaining upper age-groups are left blank. This is to preserve a standard layout for all downloaded files.
Q. I would like to analyse causes of deaths other than those listed, how do I proceed?
A. On the same web site where you have found this application, there is a link to download the raw detailed mortality files for each individual country by ICD code, year, age and sex. The data are available for years starting 1950 to date. You should have some experience in handling big databases (over million records) as the files cannot be imported into Excel. You should use a database management or statistical software such as MS Access, Stata, SAS, SPSS or R.