Adult mortality in Zambia: An ecological model
Date
2018-06
Authors
Chisumpa, Vesper Hichilombwe
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Background: Adult mortality has largely remained a major health concern and an under-researched area thus it does not feature prominently on the national development agenda of Zambia, even in development plans, health policy, national health strategic plan, consequently there are no specific programmes to address adult mortality in the country. Equally, the past Millennium Development Goals (MDGs) did not have a specific focus on adult mortality; the Sustainable Development Goals (SDGs) as well have a broader health agenda but no specifics on adult mortality. Adult mortality just like child mortality is also linked to the socioeconomic development of a country. In Zambia, adult mortality has remained among the highest in southern Africa as noted by the National Population Policy of Zambia. The probability of dying between ages 15 and 60 years for the period 2010-2015 was estimated at 306.5 deaths per 1,000 persons alive at age 15 and above. Yet the adult population in age group 15 to 59 years constitutes about 50 per cent of Zambia's total population of over 13 million. Adult mortality has been an issue of public health concern for several decades now because of its impact at individual, household, community and national level in development terms as the consequences are immense to be neglected. Studies have shown that the community context plays a mediating role by influencing the type of life style which eventually determines individual outcomes like adult mortality. Previous studies have also argued that adult mortality is associated with factors such as education, income, marital status, religion, age, sex, neighbourhoods, smoking, and alcohol among others. However, these studies did not adequately address some of the contextual factors such as place of death, community health care utilisation, and community illness treatment received as they relate to adult mortality. This study addresses these issues with special attention to their effects on adult mortality at community level by applying an ecological model to reveal the previously unexamined associations between contextual factors and adult mortality variations. The study set out to achieve four specific objectives: (i) establish the level of adult mortality situation in Zambia; (ii) examine the causes of adult mortality in Zambia; (iii) explore the age- and cause-specific mortality contributions, and differentials in adult mortality; and (iv) determine the extent to which individual-, household-, and community-level factors influence adult mortality variations in Zambia.
Methodology: The study utilised two datasets, the 2010 census (10 per cent sample) and 2010-2012 Sample Vital Registration with Verbal Autopsy survey (SAVVY). The 2010 census reported 16,445 total deaths of which 6,693 occurred in age group 15-59; whereas the 2010-2012 SAVVY recorded 2,759 total deaths of which 1,078 were adult deaths in age group 15-59. The 2010 census data were utilised in deriving adult mortality rates in objective one as they were more
appropriate with the methods applied. The 2010-2012 SAVVY dataset was utilised in addressing
all the study objectives. Objective one was achieved by employing direct (life table and
siblinghood), and indirect (Hill's Generalized Growth Balance (GGB), Bennett and Horiuch's
Synthetic Extinct Generations (SEG)) demographic methods of adult mortality estimation to
establish the level of adult mortality at national and sub-national levels in Zambia, that is, the
probability of dying between ages 15 and 60 years ( 45 15 q ). Objective two was achieved by
computing proportions of causes of death, age-sex and cause-specific mortality rates to examine
the causes of death among adults in the age group 15-59 years. Objective three was achieved by
constructing cause-deleted life tables to determine the impact of cause of death elimination on
adult mortality. In addition, decomposition analysis was performed to determine the age- and
cause-specific adult mortality rates’ contributions to widening the life expectancy gap between
males and females. Multivariate multilevel survival analysis was employed to determine the extent
to which individual-, household-, and community-level factors influence adult mortality
variations in Zambia, to achieve objective four. Multivariate multilevel survival analysis was
employed because it is the appropriate method for the nature of time-to-event data, that is, the
risk of dying between ages 15 and 60 years. The unit of analysis was deceased adults in age group
15-59 years. The dependent variable is adult mortality operationally defined as the risk of dying
between ages 15 and 60 years and was measured as "1" mortality of deceased persons in age group 15 to 59, and "0" deaths of persons in other ages. Independent variables at individual-,
household-, and community-levels were selected based on the literature reviewed and the study
conceptual framework. Analysis was performed at univariate, bivariate and multivariate levels.
The study hypothesized that place of death, community health service utilisation, and community
illness treatment received were associated with the risk of adult mortality. Data analysis was
performed using Stata 14 and Microsoft Excel.
Description
A Thesis submitted to the Faculty of Humanities, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the award of the degree of Doctor of Philosophy in Demography and Population Studies
June 2018
Keywords
Citation
Chisumpa, Vesper Hichilombwe (2018) Adult mortality in Zambia: An ecological model, University of the Witwatersrand, Johannesburg, <http://hdl.handle.net/10539/26759>