Search

Background: Machine learning (ML) algorithms have been successfully employed for prediction of outcomes in clinical research. In this study, we have explored the application of ML-based algorithms to predict cause of death (CoD) from verbal autopsy records available through the Million Death Study (MDS).Methods: From MDS, 18826 unique childhood deaths at ages 1-59 months during the time period 2004-13 were selected for generating the prediction models of which over 70% of deaths were caused by six infectious diseases (pneumonia, diarrhoeal diseases, malaria, fever of unknown origin, meningitis/encephalitis, and measles). Six popular ML-based algorithms such as support vector machine, gradient boosting modeling, C5.0, artificial neural network, k-nearest neighbor, classification and regression tree were used for building the CoD prediction models.Results: SVM algorithm was the best performer with a prediction accuracy of over 0.8. The highest accuracy was found for diarrhoeal diseases (accuracy = 0.97) and the lowest was for meningitis/encephalitis (accuracy = 0.80). The top signs/symptoms for classification of these CoDs were also extracted for each of the diseases. A combination of signs/symptoms presented by the deceased individual can effectively lead to the CoD diagnosis.Conclusions: Overall, this study affirms that verbal autopsy tools are efficient in CoD diagnosis and that automated classification parameters captured through ML could be added to verbal autopsies to improve classification of causes of death. [ABSTRACT FROM AUTHOR]

Background: Many studies have been conducted on under-five mortality in India and most of them focused on the associations between individual-level factors and under-five mortality risks. On the contrary, only a scarce number of literatures talked about contextual level effect on under-five mortality. Hence, it is very important to have thorough study of under-five mortality at various levels. This can be done by applying multilevel analysis, a method that assesses both fixed and random effects in a single model. The multilevel analysis allows extracting the influence of individual and community characteristics on under-five mortality. Hence, this study would contribute substantially in understanding the under-five mortality from a different perspective., Method: The study used data from the Demographic and Health Survey (DHS) acquired in India, i.e., the fourth round of National Family and Health Survey (2015-16). It is a nationally representative repeated cross-sectional data. Multilevel Parametric Survival Model (MPSM) was employed to assess the influence of contextual correlates on the outcome. The assumption behind this study is that 'individuals' (i.e., level-1) are nested within 'districts' (i.e., level-2), and districts are enclosed within 'states' (i.e., level-3). This suggests that people have varying health conditions, residing in dissimilar communities with different characteristics., Results: Highest under-five mortality i.e., 3.85% are happening among those women whose birth interval is less than two years. In case of parity, around 4% under-five mortality is among women with Third and above order parity. Further, findings from the full model is that ICC values of 1.17 and 0.65% are the correlation of the likelihood of having under-five mortality risk among people residing in the state and district communities, respectively. Besides, the risk of dying was increased alarmingly in the first year of life and slowly to aged 3 years and then it remains steady., Conclusion: This study has revealed that both aspects viz. individual and contextual effect of the community are necessary to address the importance variations in under-five mortality in India. In order to ensure substantial reduction in under-five mortality, findings of the study support some policy initiatives that involves the need to think beyond individual level effects and considering contextual characteristics., (© 2024. The Author(s).)

Background: With 8,82,000 deaths in the under-five period, India observed varied intra-state and inter-regional differences across infant and child mortality in 2018. However, scarce literature is present to capture this unusual concentration of mortality in certain families by examining the association of the mortality risks among the siblings of those families along with various unobserved characteristics of the mother. Looking towards the regional and age differential in mortality, this paper attempts to provide evidence for the differential in mortality clustering among infants (aged 0-11 months), children (12-59 months) and under-five (0-59 months) period among mothers from the Empowered Action Group (EAG) and non-EAG regions of India.Methods: The study used data from the National Family Health Survey (2015-16) which includes all the birth histories of 475,457 women aged 15-49 years. Bivariate and multivariate analyses were used to fulfil the objectives of the study. A two-level random intercept Weibull regression model was used to account for the unexplained mother (family) level heterogeneity.Results: About 3.3% and 5.9% of infant deaths and 0.8% and 1.6% of childhood deaths were observed in non-EAG and EAG regions respectively. Among them, a higher percentage of infant and child death was observed due to the death of a previous sibling. There were 1.67 times [95% CI: 1.55-1.80] and 1.46 times [CI: 1.37-1.56] higher odds of infant and under-five mortality of index child respectively when the previous sibling at the time of conception of the index child was dead in the non-EAG regions. In contrast, the odds of death scarring (death of previous sibling scars the survival of index child) were 1.38 times [CI: 1.32-1.44] and 1.24 times [CI: 1.20-1.29] higher for infant and under-five mortality respectively in the EAG regions.Conclusion: The extent of infant and child mortality clustering and unobserved heterogeneity was higher among mothers in the non-EAG regions in comparison to their EAG region counterparts. With the growing situation of under-five mortality clustering in non-EAG states, region-wise interventions are recommended. Additionally, proper care is needed to ameliorate the inter-family variation in mortality risk among the children of both EAG and non-EAG regions throughout their childhood. [ABSTRACT FROM AUTHOR]

Background: Globally, diarrhoea is the second leading cause of death in children under five and a major public health problem. Despite several health care initiatives taken by the government, a large proportion of children still experience diarrhoeal diseases which cause high childhood death in India. This study aims to examine the socio-demographic and environmental factors associated with diarrhoea in children under five in India.Methods: A cross-sectional study was designed using secondary data from the recent round of the National Family Health Survey (NFHS-4), conducted in 2015-16. A total of 247,743 living children below 5 years of age were included in the analysis. Bivariate and multivariate logistic regression models were carried out to assess the factors associated with childhood diarrhoeal disease.Results: In India, about 9% of under-five children experience diarrhoeal disease in the past 2 weeks preceding the survey. Children living in rural areas (Adjusted odds ratio [aOR]: 1.05; 95% CI: 1.01, 1.09), children belonged to scheduled tribe (aOR: 0.83; 95% CI: 0.79, 0.89) and other castes (aOR: 0.92; 95% CI: 0.88, 0.97), Muslim children (aOR: 1.18; 95% CI: 1.13, 1.24), and children resided in the central (aOR: 1.61; 95% CI: 1.52, 1.70) and west (aOR: 1.08; 95% CI: 1.01, 1.15) regions were significantly associated with higher likelihood of diarrhoea in the past 2 weeks. Concerning environmental factors, child stool disposal (aOR: 1.06; 95% CI: 0.98, 1.09), floor materials (aOR: 1.08; 95% CI: 1.03, 1.12) and roof materials (aOR: 1.08; 95% CI: 1.04, 1.13) of the household were found to be significant predictors of childhood diarrhoea occurrence.Conclusions: Diarrhoeal disease is common among children who lived in rural areas, scheduled castes, Muslims, and children from poor families. Regarding environmental factors, stool disposal practices in the household, dirt floor, and thatch roof materials of the household unit are risk factors for diarrhoeal disease. Targeted approach should be initiated to mitigate the problem of the poor health status of children by providing adequate health care. The policy-makers and stakeholders should address adverse environmental conditions by the provision of latrine and improved housing facilities. [ABSTRACT FROM AUTHOR]

Background: Many interventions have attempted to increase vulnerable and remote populations' access to ORS and zinc to reduce child mortality from diarrhoea. However, the impact of these interventions is difficult to measure. From 2010 to 15, Micronutrient Initiative (MI), worked with the public sector in Bihar, India to enable community health workers to treat and report uncomplicated child diarrhoea with ORS and zinc. We describe how we estimated programme's impact on child mortality with Lives Saved Tool (LiST) modelling and data from MI's management information system (MIS). This study demonstrates that using LiST modelling and MIS data are viable options for evaluating programmes to reduce child mortality.Methods: We used MI's programme monitoring data to estimate coverage rates and LiST modelling software to estimate programme impact on child mortality. Four scenarios estimated the effects of different rates of programme scale-up and programme coverage on estimated child mortality by measuring children's lives saved.Results: The programme saved an estimated 806-975 children under-5 who had diarrhoea during five-year project phase. Increasing ORS and zinc coverage rates to 19.8% & 18.3% respectively under public sector coverage with effective treatment would have increased the programme's impact on child mortality and could have achieved the project goal of saving 4200 children's lives during the five-year programme.Conclusions: Programme monitoring data can be used with LiST modelling software to estimate coverage rates and programme impact on child mortality. This modelling approach may cost less and yield estimates sooner than directly measuring programme impact with population-based surveys. However, users must be cautious about relying on modelled estimates of impact and ensure that the programme monitoring data used is complete and precise about the programme aspects that are modelled. Otherwise, LiST may mis-estimate impact on child mortality. Further, LiST software may require modifications to its built-in assumptions to capture programmatic inputs. LiST assumes that mortality rates and cause of death structure change only in response to changes in programme coverage. In Bihar, overall child mortality has decreased and diarrhoea seems to be less lethal than previously, but at present LiST does not adjust its estimates for these sorts of changes. [ABSTRACT FROM AUTHOR]

Background: Even though the overall infant mortality rate and child mortality rate have considerably declined in India, the marginalised groups-Scheduled Caste, and Scheduled Tribe, continue to have higher mortality rates. This study looks at the changes in IMR and CMR amongst disadvantaged and advanced social groups at the national level and in three states of India., Data and Methods: Data from five rounds of National Family Health Survey spanning nearly three decades have been used to measure IMR and CMR according to the social groups for India and for selected states-Bihar, West Bengal, and Tamil Nadu. The relative hazard curves were drawn to understand which social group's children face a higher risk of dying in the first year of birth and between first year and age four in those three states. Further, a log-rank test was applied to examine whether the differences between the survival curves or distributions of the three social groups are statistically significant. Finally, the binary logit regression model was applied to investigate the effect of ethnicity, and other socio-economic and demographic covariates on the risk of infant and child deaths (1-4 years) in the country and in selected states., Results: Hazard curve shows the probability of death within one year of birth was highest among the children belonging to the ST families, followed by SCs in India. And, the CMR was found to be higher amongst the STs compared to all other social groups as well at the national level. While Bihar had a remarkably high infant and child mortality rates, Tamil Nadu possessed the lowest child death rates irrespective of class, caste, and religion. The regression model revealed that most of the caste/tribe gaps in infant and child deaths might be attributed to the place of residence, mother's level of education, economic status, and the number of children in a family. Notably, the multivariate analysis showed that ethnicity was an independent risk factor, when controlled for socioeconomic status., Conclusion: The study detects the persistence of significant caste/tribe differentials in infant and child mortality in India. Poverty, education and health care access issues could be the possible reasons for the premature deaths of the children from deprived castes and tribes. There is a need to critically analyse the current health programmes aimed at reducing IMR and CMR to make them attuned to the needs of the marginalised communities., (© 2023. The Author(s).)

Background: India has the world's highest total number of under-five deaths of any nation. While progress towards Millennium Development Goal 4 has been documented at the state level, little information is available for greater disaggregation of child health markers within states. In 2000, new states were created within the country as a partial response to political pressures. State-level information on child health trends in the new states of Chhattisgarh and Jharkhand is scarce. To fill this gap, this article examines under-five and neonatal mortality across various equity markers within these two new states, pre-and post-split. Methods: Both direct and indirect estimation using pooled data from five available sources were undertaken. Inter-population disparities were evaluated by mortality data stratification of rural-urban location, ethnicity, wealth and districts. Results: Both states experienced an overall reduction in under-five and neonatal mortality, however, this has stagnated post-2001 and various disparities persist. In cases where disparities have declined, such as between urban-rural populations and low- and high-income groups, this has been driven by modest declines within the disadvantaged groups (i.e. low-income rural households) and stagnation or worsening of outcomes within the advantaged groups. Indeed, rising trends in mortality are most prevalent in urban middle-income households. Conclusions: The results suggest that rural health improvements may have come at the expense of urban areas, where poor performance may be attributed to factors such as lack of access to quality private health facilities. In addition, the disparities may in part be associated with geographical access, traditional practices and district-level health resource allocation. [ABSTRACT FROM AUTHOR]

Background: Relating Information on causes of deaths to implementation of health interventions provides vital information for program planning and evaluation. This paper from Ballabgarh Health and Demographic Surveillance System (HDSS) site in north India looks at temporal trends and gender differentials in the causes of death among under-five children. Methods: Data on causes of death for 1972-74, 1982-84, 1992-94, 2002-04 were taken from existing HDSS publications and database. Physicians' assigned causes of death were based on narratives by lay health worker till 1994 and later by verbal autopsy. Cause Specific Mortality Fractions (CSMF) and Cause Specific Mortality Rates (CSMR) per 1000 live births were calculated for neonatal (<1 month) and childhood (1-59 months) period. Gender difference was estimated by calculating ratio of CSMR between girls and boys. Available information on coverage of childhood interventions in the HDSS was retrieved and compiled. Results: The CSMF of prematurity and sepsis was 32% and 17.6% during neonatal period in 2002-04. The share of infections in all childhood deaths decreased from 55.2% in 1972-74 to 43.6% in 2002-04. All major causes of mortality (malnutrition, diarrhea and acute lower respiratory infection) except injuries showed a steep decline among children and seem to have plateued in last decade. Most of disease specific public health interventions were launched in mid eighties. . Girls reported significantly higher mortality rates for prematurity (RR 1.52; 95% CI 1.01-2.29); diarrhea (2.29; 1.59 - 3.29), and malnutrition (3.37; 2.05 - 5.53). Conclusions: The findings of the study point out to the need to move away from disease-specific to a comprehensive approach and to address gender inequity in child survival through socio-behavioural approaches. [ABSTRACT FROM AUTHOR]

Background: Most households in low and middle income countries, including in India, use solid fuels (coal/coke/lignite, firewood, dung, and crop residue) for cooking and heating. Such fuels increase child mortality, chiefly from acute respiratory infection. There are, however, few direct estimates of the impact of solid fuel on child mortality in India.Methods: We compared household solid fuel use in 1998 between 6790 child deaths, from all causes, in the previous year and 609 601 living children living in 1.1 million nationally-representative homes in India. Analyses were stratified by child's gender, age (neonatal, post-neonatal, 1-4 years) and colder versus warmer states. We also examined the association of solid fuel to non-fatal pneumonias.Results: Solid fuel use was very common (87% in households with child deaths and 77% in households with living children). After adjustment for demographic factors and living conditions, solid-fuel use significantly increase child deaths at ages 1-4 (prevalence ratio (PR) boys: 1.30, 95%CI 1.08-1.56; girls: 1.33, 95%CI 1.12-1.58). More girls than boys died from exposure to solid fuels. Solid fuel use was also associated with non-fatal pneumonia (boys: PR 1.54 95%CI 1.01-2.35; girls: PR 1.94 95%CI 1.13-3.33).Conclusions: Child mortality risks, from all causes, due to solid fuel exposure were lower than previously, but as exposure was common solid, fuel caused 6% of all deaths at ages 0-4, 20% of deaths at ages 1-4 or 128 000 child deaths in India in 2004. Solid fuel use has declined only modestly in the last decade. Aside from reducing exposure, complementary strategies such as immunization and treatment could also reduce child mortality from acute respiratory infections. [ABSTRACT FROM AUTHOR]

Background: India has achieved impressive gains in child survival over the last two decades; however, it was not successful in attaining MDG 2015 goals. The study's objective is to inquire how the survival status of the preceding child affects the survival of the next born child., Methods: This is a retrospective analysis of data from the National Family Health Survey, 2015-16. Analysis was restricted to women with second or higher-order births because women with first-order births do not have a preceding child. Proportional hazards regression, also called the Cox regression model, has been used to carry out the analysis. Kaplan-Meier (K-M) survival curves were also generated, with a focus on preceding birth intervals., Results: Results found that female children were more likely to experience infant mortality than their male counterparts. Children born after birth intervals of 36+ months were least likely to experience infant mortality. Mother's education and household wealth are two strong predictors of child survival, while the place of residence and caste did not show any effect in the Cox proportional model. Infant and child deaths are highly clustered among those mothers whose earlier child is dead., Conclusion: Maternal childbearing age is still low in India, and it poses a high risk of infant and child death. Education is a way out, and there is a need to focus on girl's education. The government shall also focus on raising awareness of the importance of spacing between two successive births. There is also a need to create a better health infrastructure catering to the needs of rich and poor people alike., (© 2021. The Author(s).)