Estimating infant mortality trends from child survivorship data

Area of Science:

• Demography
• Public Health
• Statistical Modeling

Background:

• Estimating mortality trends is crucial for public health.
• Brass's procedure is a standard method using census/survey data.
• Previous applications often focused on younger women's reproductive data.

Purpose of the Study:

• To generalize Brass's procedure for improved mortality trend estimation.
• To assess the reliability of data from older age groups for infant mortality estimation.
• To compare estimates from different age groups with vital registration data.

Main Methods:

• Generalization of Brass's mortality estimation procedure.
• Application to census/survey data from Costa Rica and peninsular Malaysia.
• Comparison of derived infant mortality rates with vital registration figures.

Main Results:

• Estimates from women aged 30-50 showed comparable accuracy to those from women aged 20-30.
• Excluding data from women over 30 or 35 may lead to biased mortality estimates.
• Estimates using data from women under 20 were found to be significantly biased.

Conclusions:

• The generalized procedure effectively estimates mortality trends.
• Data from women aged 30-50 is valuable for infant mortality rate estimation.
• Rethinking age-group exclusions in mortality estimation is recommended, especially for younger mothers.