Determinants of child mortality in Kenya
Child mortality has been considered an index that reflects the degree of poverty and deprivation of a population. Millennium Development Goal 4 specifically describes targets which countries should aim for as a show of improvement of the overall health of a population by reducing child mortality. In the recent decades, information on trends aad determinants of child mortality have been the basis of planning, implementation and evaluation of health policies and programs. Proximate determinants that directly influence morbidity and mortality as outlined by Mosley and Chen 1984, are the foundation of the Demographic and Health Surveys which have over the years increased knowledge on the determinants of child survival in many countries. Using the KDHS of 2008/9, this study presents an analysis of determinants of child mortality with the aim of establishing whether child survival strategies which aim at reducing the effect of child mortality determinants should be continued or sustained over years to win the war on child mortality. In some instances, reference will be made to the preceding KDHS of 2003 to bring out the trend element in the reduction of child mortality. For the study methodology frequencies provided a description of the study population while the Logistic regression was used for the bivariate analysis which showed the association between study variables and dependent variables. Of significance was age at first birth of the mother, birth order, birth interval, mother's and father's level of education, mothers working status-confounded by whether formal or informal and households access to sanitation and water. Multivariate analysis using the three models of demographic factors, socio economic and cultural factors and environmental factors was used to measure the net effect of demographic variables on the risk of death before age five. The study findings showed that at the multivariate level, significant determinants of child mortality according to KDHS of2008/9 are the duration of the succeeding birth interval; where a child was less likely to die when the interval between births is 24 months and above. The region of residence was also a significant variable; with children from all regions in the country being less likely to die when measured against Nairobi. The data from the study also showed that socio-culturally a child whose parents are Muslims or do not subscribe to any religion was less likely to die and finally, a child residing in a household with no access to piped water has reduced likelihood of mortality. The study therefore concluded that factors affecting child mortality can over time change in intensity and should therefore be reassessed regularly. Key of these is region of residence, religion and access to piped water. However there are still other factors that retain their significance as determinants of child mortality at the same level e.g. age at first birth of the mother, birth order of the reference child and duration of the succeeding birth interval are consistent with reviewed literature on child mortality determinants where for example a child of higher birth order is more likely to die and one whose birth interval is 24 months or more is less likely to die. Key recommendations deduced from this study and also based on the literature reviewed for the purpose of the study show that there is indeed room for further analysis of determinants of child mortality found to be significant. This include region of residence, religion and access to piped water. At the same time, the study recommends that because of the role succeeding birth interval of 24 months and above have played in reducing mortality, the incentives that have reinforced these intervals should be encouraged if child mortality is to be reduced over time.