Sample Registration Systems

Similar to a vital registration system, but covering only a small area. In addition to continuous recording of vital events, twice a year all sample households are surveyed about births and deaths in the past 6 months.

Sample registration is described in greater detail on the following page.

• Requires sophisticated administration and ongoing financial support
• Susceptible to disruption by civil unrest

Demographic Surveillance Sites

Households within a defined geographical area are followed over time and demographic events are identified through regular household visits by trained interviewers.

Surveillance sites are considered in more detail in Session 3 (Sources of Longitudinal Data) [link].

• Costly
• Unrepresentative

Demographic Household Surveys

Surveys carried out in small but representative samples of households. Usually take a full birth history and ask about other health related topics. Can also be tailored to specific topics e.g. HIV knowledge, intimate partner violence etc. Examples include Demographic and Health Surveys (DHS) and UNICEF’s Multiple Indicator Cluster Surveys.

Demographic Household Surveys are described in greater detail in Session 2 (Single-round Surveys) [link].

• Small samples restrict the ability to make precise estimates for population subgroups
• Can be tailored to estimate adult mortality, but often provides an underestimate (Hill et al. 2007)

Verbal Autopsy

Can be used where cause of death data is unavailable or unreliable. Trained interviewers collect information from relatives about the signs, symptoms and events in the period preceding an individual’s death. This information is analysed by trained doctors, and a cause of death assigned on the basis of the report.

• Differing methods between settings mean comparisons are impossible
• Length of the recall period may cause respondents to forget important information
• Training for interviewers increases the cost and time-intensiveness of the method