Checking the accuracy of vital events records
Examples of strategies to improve the quality of vital events records
Strategy 1: Strengthen capacity in registration and local health offices to check the microdata before they are transferred further.
At the moment of registration, some simple checks need to be implemented to verify that the information entered on the record is correct and complete. With proper SOPs, it is easy to train staff to check the raw data for content errors and omissions. SOPs should guide staff to check each data item and verify the veracity of the supporting documentation – that is, the notification agent, the medical death certificate, the identity papers (if required) and any other documentation requested for the registration to take place.
Besides checking that the registration forms are complete, staff may also perform some simple plausibility checks, such as:
- The age of a mother whose age is listed as younger than 15 or older than 49
- The sex of a decedent whose cause of death does not seem plausible – for example, men do not die from maternal causes, women do not die from prostate cancer, and adults generally do not die from causes associated with the neonatal period.
Plausibility checks should happen at the point of registration to ensure that, when the data are compiled, the microdata do not contain a content error.
Additionally, staff should receive training in data cleaning and basic quality assessment to check for potential errors in the data. Such checks on the compiled data should be a standard part of their duties, and should be performed routinely on all data to ensure that compilation errors are eliminated and the data are consistent before transfer and further use.
Strategy 2: Information technology can be used to check for errors in data entry and their transmission
Customised software tools have been developed to assist data entry and compilation. These tools ensure data items are not skipped and key characteristics are not left blank. Sometimes these tools use dropdown menus to assist with data entry, and they can also have built-in plausibility checks to help ensure the data entered are as accurate as possible. For example, if the age of a mother is outside the range of 15–49 years, the computer may display an error message asking the user to confirm the mother’s age before continuing with the rest of the birth registration.
Additionally, using information technology (IT) can improve the timeliness of data. The long delays often observed between the collection of the data and the production of periodic reports is often due to slow transmission of data from the periphery to the centre when this is done manually or without computers. A comprehensive assessment of the national CRVS system that includes a detailed map of the existing business processes is likely to reveal where it is most urgent and effective to begin introducing computerised data and records transfer.