Abstract:
Infant and under-five mortality are the sensitive indices of development and often
reflect a country's quality of life. But there is little understanding of whether
Bangladesh will be capable of achieving all of the Millennium Development Goals
(MDGs) especially the MDG 4 relating to infant (0-11 months) and under-five (0-59
months) mortality. In this view, an attempt has been made to estimate childhood
mortality and to identify the covariates of infant and under-five mortality measuring
the contribution of each of the covariates to overall variation, with a view to reducing
infant and under-five morbidity and mortality in Bangladesh so that the country may
be able to reach the MDG 4 target by the year 2015. For this, data are taken from the
Bangladesh Demographic and Health Survey (BDHS), 2004, for the ten-year period
preceding the survey (with a total sample size 7427, of whom 5004 from rural and
2423 from urban areas), having an eligible woman (ever-married and aged 10-49)
with at least one or more child(ren) and the study is executed separately for the
overall country level, rural level and urban level of Bangladesh. For analyzing the
data, indirect techniques (Brass, Sullivan and Trussell methods), abridged life table
technique and well-known statistical tools such as Pearson chi-square (X2
) test, logistic regression model and discriminant analysis have been performed.
From the results, it is revealed that the probabilities of dying (nqx) calculated, using
Brass, Sullivan and Trussell methods are increasing with the increase of age of
mothers. Again in case of data classified by duration of marriage of mothers, it is
observed that the nqx is increasing with the increase of marital duration. But for the
urban male, these probabilities are decreasing from the marital duration of 0-4 years
to 10-14 years and then again increasing up to the last marital duration group. The
results of z2 test indicate that among all the selected variables e.g. mothers' education,
age of mother at birth, is child twin?, toilet facility, T.T. injections, breastfeeding,
antenatal visits, after birth checked health, parity and preceding birth interval are
statistically and significantly associated with the infant and under-five mortality. The
results of logistic regression analysis using these ten significant variables (the two
variables such as child's sex and sources of drinking water are excluded because of
their insignificant association showed through x2 test) revealed that is child twin?,
breastfeeding, antenatal visits and after birth-checked health have highly significant
(p<O.O 1) influence on infant and under-five mortality. The impact of toilet facility on
infant and under-five mortality, except under-five mortality of urban level, is not
statistically significant. The results of abridged life table indicate that the survival
function Clx) decreases smoothly throughout the period both the rural and urban areas
of Bangladesh and these changes are little after the neonatal period. The nqx function
is decreasing with the increase of age and it is higher in rural area than that of urban
area. In case of the expectation of life, it is higher for literate and working mothers
than that of illiterate and not working mothers respectively. The results also indicate
that life expectancy is the highest in age group of 1-2 months for all the studied areas.
In discriminant analysis, the stepwise procedure has been picked up and only the
significant variables are ranked according to the rank of Wilk's Lambda values and
the results show that breastfeeding is the most important variable. Also the canonical
discriminant function coefficients (unstandardized and standardized) for the predictor
variables have also been calculated and the results are almost similar to Wilk's
Lambda. The related results of discriminant function also indicate that the
discriminant function is statistically significant and discriminates well.
However, improvements in the health system for increased coverage of antenatal and
postnatal care, for avoiding pregnancy related complications especially during twin
births and for improving the breastfeeding practices (both inclusive and exclusive) are
essential which may the effective strategies to reach families and communities with
targeted messages and information.
Description:
This thesis is submitted to the Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi Bangladesh for the Degree of Master of Philosophy (MPhil)