Methods and tools to evaluate the quality of vital statistics
Methods to measure death registration completeness
Traditionally, completeness for death registration has not been simple to measure accurately. It has been estimated using several indirect demographic and statistical techniques (for example, the Brass Growth Balance, Generalised Growth Balance, Bennett–Horiuchi or Preston–Coale methods). These methods are best used by national statistical offices, as they require the use of census data and an understanding of the different techniques.
Because of the complexity of these methods, many countries simply determine completeness by dividing the actual recorded deaths by the total estimated number of deaths from a reliable and independent source, such as a national census or an estimate from the United Nations of Global Burden of Disease study. This number can then be compared with the completeness measure produced by the ANACONDA tool.
Some countries include a question in their national census that asks if households experienced a death during the previous 12 months and whether this death was registered. Although this method doesn’t produce an exact, reliable estimate, it can provide a basis for comparison to determine if death statistics from vital registration are plausible.
Another direct approach of measuring completeness is the capture–recapture technique (described above) which directly links registered deaths with another source of mortality data (eg burial records or deaths from a survey).
Some DHSs now ask about whether a death certificate is available for each child reported as having ever died. If a death certificate is not available, an additional question asked whether the death had ever been registered. If this information is available, it can provide a partial picture about the completeness of death registration for children.
A new method to measure death registration completeness
A new empirical method to estimate death registration completeness utilises data that are commonly available at the national and subnational level. This method estimates completeness using the registered crude death rate, the proportion of the population aged 65 years and over, the under-five mortality rate, and the completeness of under-five death registration. Advantages of the method compared to other methods are its better accuracy, improved timeliness, application to subnational areas and greater simplicity of use.
The method was developed based on data from 2,500 country-years of registration data from 110 countries. The empirical method can also be used on any source of mortality data, for example to estimate the percentage of total deaths occur in hospitals. The method is particularly useful for monitoring changes in completeness of death registration, particularly where new interventions have been implemented (such as a new process for death notification and registration).
All countries should carefully measure coverage and completeness levels each year, and have a plan to achieve full coverage and completeness if not already achieved. Although it is not realistic that 100 per cent of vital events will be registered, countries should aim to have at least 90 per cent of all births and deaths registered.
Handbook of Vital Statistics Methods available at: https://unstats.un.org/unsd/demographic/standmeth/handbooks/Series_F7en.pdf
Estimating the completeness of death registration: An empirical method
Many national and subnational governments need to routinely measure the completeness of death registration for monitoring and statistical purposes. Existing methods, such as death distribution and capture-recapture methods, have a number of limitations such as inaccuracy and complexity that prevent widespread application. This paper presents a novel empirical method to estimate completeness of death registration at the national and subnational level.
Authors: Adair T, Lopez AD
Publication date: May 2018
Resource type: CRVS clearinghouse
Related resources: Summary: A new method for estimating the completeness of death registration
Source: PLoS ONE