Estimating levels and trends of infant and childhood mortality in Kenya by the three variants of the brass method

Abstract

Accurate measurement of the levels of mortality is a fundamental aspect of research in demography; infant and under-five mortality rates need to be estimated both precisely, accurately and should refer to as recent a time as possible. Estimates of infant and child mortality rates are usually derived in two ways: by using the direct method and indirect method. The general objective of this study is to estimate and compare infant and child mortality in Kenya derived from the three variants of the Brass indirect methods; women classified by age group, time since first birth, and duration of marriage. The analysis was based on data from the 1993, 1998, 2003 and 2008/09 Kenya Demographic and Health Surveys. Estimates of infant and under-five mortality in Kenya using the three Brass variants were calculated by the spreadsheet developed by the International Union for the Scientific Study of Population under the Tools for Demographic Estimation (2013). Graphs were generated that show trend-lines in the general direction and levels of infant and child mortality that each variant produces. A straight average of indirect estimates illustrates the general trend in infant and under-five mortality for the period 1993 to 2009 as depicted by the combined indirect estimates and published direct estimates. Paired difference method was used to test whether the differences between each pair of the three Brass variants were statistically significant. Evidence from this study indicates that all variants of the Brass methods produce estimates that are at similar levels and can be relied upon to track accurate trends in infant and under-five mortality. Further, direct and indirect estimates are comparable as the differences were proved to be statistically insignificant. Given that the variants of the method provide similar results with little or no disparity, the choice by users should be influenced by the simplicity and convenience of the technique and availability of required data. The major policy implication drawn from the study is that, despite evidence of a recent reduction in infant and under-five mortality, achieving MDG 4 targets will be a challenge. Therefore, the government of Kenya should accelerate efforts if it is to achieve targets set out in the post MDG’s 2015 agenda and the National Population Policy. In addition, there is need for further research on consistency of these estimates when data is adjusted to reflect biases introduced by HIV/AIDS.