| dc.description.abstract | The purpose of this study is to evaluate the preceding birth technique and to see whether the results obtained can be compared to those derived using the Coale Trussell and the direct life table method to give reliable childhood mortality estimates.
The research question that arises is to what extend do estimates of childhood mortality derived using the preceding birth technique compare with those derived using direct and stable population models. The most important source to obtain reasonable estimates of childhood mortality in developing countries has been the use of retrospective questions in censuses and surveys. The best results so far obtained have been from questions about the total number of children ever born and the number surviving to women along their reproductive period. This has been faced with the limitations; namely lack of a current indicator of infant mortality that may be relevant enough for an evaluation of a health programme especially those aimed at reducing the incidence of infant mortality.
The second research question is; in the absence of facility based data, can stable estimates of childhood mortality be obtained from the preceding birth technique using household data. Procedures such as those based on the survival of the preceding child provide more recent estimates of childhood mortality and are less .d~pendent on model patterns of mortality. Because of its simplicity, it can easily be used by other practitioners not trained in demographic techniques as a useful reliable indicator for
monitoring programme activities. ..
The estimation is based on a recent five-year reference period preceding the survey interviews of the 1998 Kenya Demographic and Health Survey. These estimates are worked out using different socio-economic characteristics such as maternal age, type of residence, maternal education and current working status of the mother for comparisons between preceding birth technique (IT) and Coale-Trussell. Other socio economic characteristics such as birth order, sex, ethnicity and religion are added for comparisons between IT and the direct life table method. The estimates are also
obtained for the eight provinces of Kenya. Once the difference between IT and Coale-
-....
Trussell and the direct life table method are obtained, percentages are also calculated for the same for further ease in comparison.
As the study results indicate, general trends expected in childhood mortality are realized in all calculation, with children born to mothers with no education experiencing higher mortality than their counterparts. Those born in Nyanza, birth order seven and above and the male gender also show higher mortalities in their categories. However calculations using the Coale-Trussell method were quite unstable possibly because of the sample sizes but estimate comparisons with the preceding birth technique (PBT) show that the two techniques are fairly comparable. This is attributed to the fact that they are both indirect methods and are based on the stable population theory.
Comparisons between II and the direct life table method showed a tendency of the II estimates to consistently exceed the life table estimates of probability of dying at 24 months of age. There is no comparability in results obtained using PBT and the direct method. This could be attributed to the use of incomplete birth histories in PBT and
complete use of birth history data in the life, table method.
As a recommendation for further research, smce the PBT method was thought
primarily as a method for estimating mortality using clinic records for a small region,
"'_0'l. .'
it would be ideal to evaluate it in multiple clinic settings and compare its computations of IIwith life table estimates of child mortality by 24 months of age
obtained from reliable census, vital registration or survey data for the areas covered
...
by the clinics. Another challenge for researchers would be to examine the underlying
assumption that II approximates probability of dying at 24 months of age; an assumption that has already been challenged by several studies in Latin America and South Asia. To policy makers, the PBT method may prove useful for public health
administrators as an easy to compute indicator of mortality index. | en |
