Publication

Back to overview

Using health and demographic surveillance system (HDSS) data to analyze geographical distribution of socio-economic status; an experience from KEMRI/CDC HDSS

Type of publication Peer-reviewed
Publikationsform Original article (peer-reviewed)
Author Amek Nyaguara, Vounatsou Penelope, Obonyo Benson, Hamel Mary, Odhiambo Frank, Slutsker Laurence, Laserson Kayla,
Project Development of spatial statistical methodology for the analysis of health demographic surveillance system (DSS) data
Show all

Original article (peer-reviewed)

Journal Acta Tropica
Volume (Issue) 144
Page(s) 24 - 30
Title of proceedings Acta Tropica
DOI 10.1016/j.actatropica.2015.01.006

Abstract

Continuous monitoring in health and demographic surveillance sites (HDSS) allows for collection of longitudinal demographic data, health related, and socio-economic indicators of the site population. We sought to use household survey data collected between 2002 and 2006 in the Kenya Medical Research Institute in collaboration with Centers for Disease Control and prevention (KEMRI/CDC) HDSS site in Asembo and Gem Western Kenya to estimate socio-economic status (SES) and assess changes of SES over time and space. Data on household assets and characteristics, mainly source of drinking water, cooking fuel, and occupation of household head was annually collected from 44,313 unique households during the study period. An SES index was calculated as a weighted average of assets using weights generated via Principal Component Analysis (PCA), Polychoric PCA, and Multiple Correspondence Analysis (MCA) methods applied to the pooled data. The index from the best method was used to rank households into SES quintiles and assess their transition over time across SES categories. Kriging was employed to produce SES maps at the start and the end of the study period. First component of PCA, Polychoric PCA, and MCA accounted for 13.7%, 31.8%, and 47.3%, respectively of the total variance of all variables. The gap between the poorest and the least poor increased from 1% at the start to 6% at the end of the study period. Spatial analysis revealed that the increase in least poor households was centered in the lower part of study area (Asembo) over time. No significant changes were observed in Gem. The HDSS sites can provide a platform to assess spatial-temporal changes in the SES status of the population. Evidence on how SES varied over time and space within the same geographical area may provide a useful tool to design interventions in health and other areas that have a close bearing to the SES of the population.
-