|
" Infant Death Clustering in India: "
Ranjan, Mukesh
Kant Dwivedi, Laxmi
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1111903
|
| Doc. No
|
:
|
TLpq2506381491
|
| Main Entry
|
:
|
Kant Dwivedi, Laxmi
|
|
|
:
|
Ranjan, Mukesh
|
| Title & Author
|
:
|
Infant Death Clustering in India:\ Ranjan, MukeshKant Dwivedi, Laxmi
|
| College
|
:
|
International Institute for Population Sciences University
|
| Date
|
:
|
2019
|
| student score
|
:
|
2019
|
| Degree
|
:
|
Ph.D.
|
| Page No
|
:
|
227
|
| Abstract
|
:
|
Background Heterogeneity in the risk of experiencing child loss between families/mothers in an area is known in demographic literature as death clustering in families. In India, infant mortality has been reduced significantly during last 25 years but the pace of reduction in infant mortality is substantially much faster than the reduction of clustering of deaths among families. Although, reduction was observed among families but the level of clustered infant deaths in such families remained at higher level. Present research is undertaken to understand the phenomenon so that infant mortality could be reduced effectively. Objective Following objectives have been addressed in the thesis. Firstly, the quality of birth history data was appraised. Then the influence of past experience of infant death among families on subsequent infant death was assessed through meta-analysis and systematic literature review by selected countries/regions. Thirdly, the reduction in infant mortality after eliminating death clustering at the family level was investigated. Next, the relative contribution of death clustering along with various socioeconomic and biodemographic factors in explaining the decline of infant deaths has been examined. Lastly, the effects of spatial and non-linear covariates and clustering of deaths on infant mortality in India has been studied. Methods Present study utilized 10 years birth history data from all four rounds of National Family Health Surveys during 1992-2016. In addition, systematic literature review and meta-analysis was carried out on literatures which were published after 1990, belonged to developing countries and included the past experience of infant death as one of the covariates in the model. All rates like NMR, IMR, U5MR, and TFR have been computed using standard DHS procedures. Bivariate and multivariate techniques have been used to fulfill all the objectives. Random effects dynamic probit model has been used to address the problem of endogeneity arises due to correlation between the survival status of previous death and unobserved mother level factor. Geo additive model was used to accommodate spatial and non-linear covariate in the model. Findings In NFHS-4, the displacement of birth still exists in India, though it was observed to be smaller than the previous surveys. Based on 22 studies the pooled estimate of the effect of previous infant deaths in families on the infant death of the index child was a significant factor in Africa and India, whereas in Bangladesh and Latin America it was found to be insignificant. Literatures suggests in family level clustering of deaths, the role of mother’s education, her occupation or income, and maternal competence factor to be important. In NFHS-1, in India, nearly 7 percent of families have experienced two or more infant deaths and they alone contributed over one-half (52%) of the total infant deaths. In NFHS-3, the disproportionate distribution of infant deaths in families shows nearly 46 percent of infant deaths were concentrated in 5 percent families. Among the selected states, Rajasthan was the only state which experienced a 31 percent relative increase in the percentage of families and 2 percent relative increase in clustered infant deaths in such families. In NFHS-4, there were about 2 percent families in India who had experienced two or more infant deaths and contributed nearly two-fifths (37%) of the total infant deaths. Among the selected states, Rajasthan experienced faster reduction in clustered infant deaths in families between NFHS3 and NFHS-4. For mothers, whose age at the first childbirth was more than 30 years, and mothers who received higher education, the death clustering among them has increased between NFHS-3 and NFHS-4.In India, if clustering of infant deaths in families are completely removed from the high mortality states, there would be roughly a decline of nearly 30 percent in IMR. Decomposition analysis of probability of infant deaths between NFHS-3 and NFHS-4 shows that the rate of change of infant deaths for a given population composition was found to be the major factor in declining the probability of infant deaths. The propensity of mother’s education and region of residence has contributed nearly 69 percent and 31 percent respectively, in the total decline of infant deaths between NFHS-3 and NFHS-4. The decline in the propensity of wealth index over the years contributed significantly in the decline of the probability of infant deaths between NFHS-3 and NFHS-4. The Bayesian geo additive discrete time survival model found that after adjusting the non-linear covariates and spatial effect, except for Kerala, the posterior mean of the occurrence of previous infant death in families on experiencing subsequent infant deaths were found to be positive and statistically significant in all selected states of India. In Bihar, districts like Paschim Champaran, Purba Champaran, Gopalganj, Sitamarhi, and Muzzfarpur are at relatively higher risk of infant death. In Uttar Pradesh, the colour of all districts are in the middle range of spatial effect colour showing moderate risk of infant death. In the states of Uttar Pradesh, Jharkhand, Maharashtra, Madhya Pradesh, Odisha, Rajasthan, and West Bengal, there is an “elongated L” shaped curve for mother’s age at childbirth which depicts that in most of the states the risk of infant mortality is higher in the initial years of reproductive period. Conclusions Policymakers should identify and target at the time of ANC visit to those pregnant women who have previously experienced any infant death. Auxiliary nurse midwife (ANM) or Accredited Social Health Activist (ASHA) worker at SHC/PHC level should provide them special care in terms of seeking better maternal and child health (MCH) services right from ANC visit to institutional delivery to post-natal care so that risk of experiencing next infant deaths among such vulnerable women could be eliminated.
|
| Subject
|
:
|
Demography
|
|
|
:
|
Individual family studies
|
|
|
:
|
Sociology
|
|
|
:
|
Southeast Asian studies
|
| |