Bayesian random effects modelling with application to childhood anaemia in Malawi

Application of machine learning methods for predicting childhood anaemia: Analysis of Ethiopian Demographic Health Survey of 2016

Childhood anaemia is a public health problem in Ethiopia. Machine learning (ML) is a growing in medicine field to predict diseases. Diagnosis of childhood anaemia is resource intensive. The aim of this study is to apply machine learning (ML) algorithm to predict childhood anaemia using socio-demographic, economic, and maternal and child related variables. The study used data from 2016 Ethiopian demographic health survey (EDHS). We used Python software version 3.11 to apply and test ML algorithms through logistic regression, Random Forest (RF), Decision Tree, and K-Nearest Neighbours (KNN). We evaluated the performance of each of the ML algorithms using discrimination and calibration parameters. The predictive performance of the algorithms was between 60% and 66%. The logistic regression model was the best predictive model of ML with accuracy (66%), sensitivity (82%), specificity (42%), and AUC (69%), followed by RF with accuracy (64%), sensitivity (79%), specificity (42%), and AUC (63%). The logistic regression and the RF models of ML showed poorest family, child age category between 6 and 23 months, uneducated mother, unemployed mother, and stunting as high importance predictors of childhood anaemia. Applying logistic regression and RF models of ML can detect combinations of predictors of childhood anaemia that can be used in primary health care professionals.

Factors associated with anemia among children in South and Southeast Asia: a multilevel analysis

BACKGROUND

South and Southeast Asian countries (SSEA) account for the highest burden of anemia globally, nonetheless, progress towards the decline of anemia has almost been stalled. This study aimed to explore the individual and community- level factors associated with childhood anemia across the six selected SSEA countries.

METHODS

Demographic and Health Surveys of SSEA countries (Bangladesh, Cambodia, India, Maldives, Myanmar, and Nepal) conducted between 2011 and 2016 were analyzed. A total of 167,017 children aged 6-59 months were included in the analysis. Multivariable multilevel logistic regression analysis was used to identify independent predictors of anemia.

RESULTS

The combined prevalence of childhood anemia across six SSEA countries was 57.3% (95% CI: 56.9-57.7%). At the individual level, childhood anemia was significantly higher among (1) mothers with anemia compared to non-anemic mothers (Bangladesh: aOR = 1.66, Cambodia: aOR = 1.56, India: aOR = 1.62, Maldives: aOR = 1.44, Myanmar: aOR = 1.59, and Nepal: aOR = 1.71); (2) children with a history of fever in the last two weeks compared to those without a history of fever (Cambodia: aOR = 1.29, India: aOR = 1.03, Myanmar: aOR = 1.08), and; (3) stunted children compared to those who were not (Bangladesh: aOR = 1.33, Cambodia: aOR = 1.42, India: aOR = 1.29, and Nepal: aOR = 1.27). In terms of community-level factors, children with mothers in communities with a high percentage of community maternal anemia had higher odds of childhood anemia in all countries (Bangladesh: aOR = 1.21, Cambodia: aOR = 1.31, India: aOR = 1.72, Maldives: aOR = 1.35, Myanmar: aOR = 1.33, and Nepal: aOR = 1.72).

CONCLUSION

Children with anemic mothers and stunted growth were found vulnerable to developing childhood anemia. Individual and community-level factors identified in this study can be considered to develop effective anemia control and prevention strategies.

Spatial disparities in impoverishing effects of out-of-pocket health payments in Malawi

BACKGROUND

Out-of-pocket health payments as a means of financing health services are a cause of concern among households in low and middle-income countries. They prevent households from accessing health care services, can disrupt households' living standards by reducing consumption of other basic needs and push households into poverty. Previous studies have reported geographical variations in impoverishing effects of out-of-pocket health payments. Yet, we know relatively little about spatial effects on impoverishing effects of health payments.

OBJECTIVE

This paper assesses the factors associated with impoverishing effects of health payments and quantifies the role of districts spatial effects on impoverishment in Malawi.

METHODS

The paper uses a cross sectional integrated household survey data collected from April 2016 to April 2017 among 12447 households in Malawi. Impoverishing effect of out-of-pocket health payments was calculated as the difference between poverty head count ratio before and after subtracting health payments from total household consumption expenditures. We assessed the factors associated with impoverishment and quantified the role of spatial effects using a spatial multilevel model.

RESULTS

About 1.6% and 1.2% of the Malawian population were pushed below the national and international poverty line of US$1.90 respectively due health payments. We found significant spatial variations in impoverishment across districts with higher spatial residual effects clustering in central region districts. Higher socio-economic status (AOR=0.34, 95% CI=0.22-0.52) decreased the risk of impoverishment whereas hospitalizations (AOR=3.63, 95% CI 2.54-5.15), chronic illness (AOR=1.56, 95% CI=1.10-1.22), residency in rural area (AOR=2.03, 95% CI=1.07-4.26) increased the risk of impoverishment.

CONCLUSIONS

Our study suggests the need to plan financial protection programs according to district specific needs and target the poor, residents of rural areas and those with chronic illnesses. Policy makers need to pay attention to the importance of spatial and neighborhood effects when designing financial protection programs and policies.

"Childhood Anemia in India: an application of a Bayesian geo-additive model"

Background

The geographical differences that cause anaemia can be partially explained by the variability in environmental factors, particularly nutrition and infections. The studies failed to explain the non-linear effect of the continuous covariates on childhood anaemia. The present paper aims to investigate the risk factors of childhood anaemia in India with focus on geographical spatial effect.

Methods

Geo-additive logistic regression models were fitted to the data to understand fixed as well as spatial effects of childhood anaemia. Logistic regression was fitted for the categorical variable with outcomes (anaemia (Hb < 11) and no anaemia (Hb ≥ 11)). Continuous covariates were modelled by the penalized spline and spatial effects were smoothed by the two-dimensional spline.

Results

At 95% posterior credible interval, the influence of unobserved factors on childhood anaemia is very strong in the Northern and Central part of India. However, most of the states in North Eastern part of India showed negative spatial effects. A U-shape non-linear relationship was observed between childhood anaemia and mother’s age. This indicates that mothers of young and old ages are more likely to have anaemic children; in particular mothers aged 15 years to about 25 years. Then the risk of childhood anaemia starts declining after the age of 25 years and it continues till the age of around 37 years, thereafter again starts increasing. Further, the non-linear effects of duration of breastfeeding on childhood anaemia show that the risk of childhood anaemia decreases till 29 months thereafter increases.

Conclusion

Strong evidence of residual spatial effect to childhood anaemia in India is observed. Government child health programme should gear up in treating childhood anaemia by focusing on known measurable factors such as mother’s education, mother’s anaemia status, family wealth status, child health (fever), stunting, underweight, and wasting which have been found to be significant in this study. Attention should also be given to effects of unknown or unmeasured factors to childhood anaemia at the community level. Special attention to unmeasurable factors should be focused in the states of central and northern India which have shown significant positive spatial effects.

A Scoping Review of Spatial Analysis Approaches Using Health Survey Data in Sub-Saharan Africa

Spatial analysis has become an increasingly used analytic approach to describe and analyze spatial characteristics of disease burden, but the depth and coverage of its usage for health surveys data in Sub-Saharan Africa are not well known. The objective of this scoping review was to conduct an evaluation of studies using spatial statistics approaches for national health survey data in the SSA region. An organized literature search for studies related to spatial statistics and national health surveys was conducted through PMC, PubMed/Medline, Scopus, NLM Catalog, and Science Direct electronic databases. Of the 4,193 unique articles identified, 153 were included in the final review. Spatial smoothing and prediction methods were predominant (n = 108), followed by spatial description aggregation (n = 25), and spatial autocorrelation and clustering (n = 19). Bayesian statistics methods and lattice data modelling were predominant (n = 108). Most studies focused on malaria and fever (n = 47) followed by health services coverage (n = 38). Only fifteen studies employed nonstandard spatial analyses (e.g., spatial model assessment, joint spatial modelling, accounting for survey design). We recommend that for future spatial analysis using health survey data in the SSA region, there must be an improve recognition and awareness of the potential dangers of a naïve application of spatial statistical methods. We also recommend a wide range of applications using big health data and the future of data science for health systems to monitor and evaluate impacts that are not well understood at local levels.

Investigating the spatial variation and risk factors of childhood anaemia in four sub-Saharan African countries

Background

The causes of childhood anaemia are multifactorial, interrelated and complex. Such causes vary from country to country, and within a country. Thus, strategies for anaemia control should be tailored to local conditions and take into account the specific etiology and prevalence of anaemia in a given setting and sub-population. In addition, policies and programmes for anaemia control that do not account for the spatial heterogeneity of anaemia in children may result in certain sub-populations being excluded, limiting the effectiveness of the programmes. This study investigated the demographic and socio-economic determinants as well as the spatial variation of anaemia in children aged 6 to 59 months in Kenya, Malawi, Tanzania and Uganda.

Methods

The study made use of data collected from nationally representative Malaria Indicator Surveys (MIS) and Demographic and Health Surveys (DHS) conducted in all four countries between 2015 and 2017. During these surveys, all children under the age of five years old in the sampled households were tested for malaria and anaemia. A child’s anaemia status was based on the World Health Organization’s cut-off points where a child was considered anaemic if their altitude adjusted haemoglobin (Hb) level was less than 11 g/dL. The explanatory variables considered comprised of individual, household and cluster level factors, including the child’s malaria status. A multivariable hierarchical Bayesian geoadditive model was used which included a spatial effect for district of child’s residence.

Results

Prevalence of childhood anaemia ranged from 36.4% to 61.9% across the four countries. Children with a positive malaria result had a significantly higher odds of anaemia [AOR = 4.401; 95% CrI: (3.979, 4.871)]. After adjusting for a child’s malaria status and other demographic, socio-economic and environmental factors, the study revealed distinct spatial variation in childhood anaemia within and between Malawi, Uganda and Tanzania. The spatial variation appeared predominantly due to unmeasured district-specific factors that do not transcend boundaries.

Conclusions

Anaemia control measures in Malawi, Tanzania and Uganda need to account for internal spatial heterogeneity evident in these countries. Efforts in assessing the local district-specific causes of childhood anaemia within each country should be focused on.

District Effect Appraisal in East Sub-Saharan Africa: Combating Childhood Anaemia

Background

Anaemia in children is a significant health problem that receives little attention. This study aimed at determining the factors significantly associated with anaemia in children aged 6 to 59 months in Kenya, Malawi, Tanzania, and Uganda while accounting for the spatial heterogeneity within and between the districts of the four countries. In addition, the performance of the districts with regard to their impact on anaemia was assessed and ranked.

Methods

A generalised additive mixed model with a spatial effect based on the geographical coordinates of the clusters was used. A district-level random effect was included to further account for the heterogeneity as well as to rank the performance of the districts based on the best linear unbiased prediction (BLUP).

Results

The results depicted significant spatial heterogeneity between and within the districts of the countries. After accounting for such spatial heterogeneity, child-level characteristics (gender, malaria test result, and mother's highest education level), household-level characteristics (household size, household's wealth index Z-score, the type of toilet facility available, and the type of place of residence), and the country of residence were found to be significantly associated with the child's anaemia status. There was a significant interaction between the type of place of residence and the country of residence. Based on the BLUP for the district-level random effect, the top 3 best- and worst-performing districts within each country were identified.

Conclusion

The ranking of the performance of the districts allows for the worst-performing districts to be targeted for further research in order to improve their anaemia control strategies, as well as for the best-performing districts to be identified to further determine why they are performing better and then to use these districts as role models in efforts to overcome childhood anaemia.

Multilevel Analysis of the Effects of Individual- and Community-Level Factors on Childhood Anemia, Severe Anemia, and Hemoglobin Concentration in Malawi

BACKGROUND

The purpose of this article was to examine individual- and community-level factors associated with childhood anemia, severe anemia, and hemoglobin (Hb) concentration in Malawi.

METHODS

Using data from the 2010 Malawi demographic and health survey (MDHS), the multilevel regression models were constructed to analyze 2597 children aged 6-59 months living in 849 communities.

RESULTS

The results showed that both childhood anemia and severe anemia were negatively associated with child's age, no fever in the previous 2 weeks and height-for-age, and positively associated with residing in poor household. Childhood anemia was negatively associated with community female education. Child's age, no fever in the previous 2 weeks and maternal Hb levels were positively associated with child Hb concentration, while residing in poorest households was negatively associated with children's Hb concentration.

CONCLUSION

Comprehensive public health strategies aimed at reducing childhood anemia need to focus more on the significant characteristics addressed in this study.

Maternal anemia is a potential risk factor for anemia in children aged 6-59 months in Southern Africa: a multilevel analysis

BACKGROUND

The effect of maternal anemia on childhood hemoglobin status has received little attention. Thus, we examined the potential association between maternal anemia and childhood anemia (aged 6-59 months) from selected Southern Africa countries.

METHODS

A cross-sectional study using nationally representative samples of children aged 6-59 months from the 2010 Malawi, 2011 Mozambique, 2013 Namibia, and 2010-11 Zimbabwe demographic and health surveys (DHS) was conducted. Generalized linear mixed models (GLMMs) were constructed to test the associations between maternal anemia and childhood anemia, controlling for individual and community sociodemographic covariates.

RESULTS

The GLMMs showed that anemic mothers had increased odds of having an anemic child in all four countries; adjusted odds ratio (aOR = 1.69 and 95% confidence interval [CI]:1.37-2.13) in Malawi, (aOR = 1.71; 95% CI: 1.37-2.13) in Mozambique, (aOR = 1.55; 95% CI: 1.08-2.22) in Namibia, and (aOR = 1.52; 95% CI: 1.25-1.84) in Zimbabwe. Furthermore, the odds of having an anemic child was higher in communities with a low percentage of anemic mothers (aOR = 1.52; 95% CI: 1.19-1.94) in Mozambique.

CONCLUSIONS

Despite the long-standing efforts to combat childhood anemia, the burden of this condition is still rampant and remains a significant problem in Southern Africa. Thus, public health strategies aimed at reducing childhood anemia should focus more on addressing infections, and micronutrient deficiencies both at individual and community levels in Southern Africa.

Spatial pattern and determinants of anaemia in Ethiopia

Anaemia is a condition in which the haemoglobin concentration falls below an established cut-off value due to a decrease in the number and size of red blood cells. The current study aimed (i) to assess the spatial pattern and (ii) to identify determinants related to anaemia using the third Ethiopian demographic health survey. To achieve these objectives, this study took into account the sampling weight and the clustered nature of the data. As a result, multilevel modeling has been used in the statistical analysis. The analysis included complete cases from 15,909 females, and 13,903 males. Among all subjects who agreed for hemeoglobin test, 5.22% males, and 16.60% females were anemic. In both binary and ordinal outcome modeling approaches, educational level, age, wealth index, BMI and HIV status were found to be significant predictors of anaemia prevalence. Furthermore, this study applied spatial methods to generate maps at regional level which could be useful for policy makers where great efforts should be concentrated to reduce the prevalence of anaemia. As revealed by Moran's I test, significant spatial autocorrelation were noted across clusters. The risk of anaemia was found to vary across different regions, and higher prevalences were observed in Somali and Affar regions.

Structured Additive Quantile Regression for Assessing the Determinants of Childhood Anemia in Rwanda

Childhood anemia is among the most significant health problems faced by public health departments in developing countries. This study aims at assessing the determinants and possible spatial effects associated with childhood anemia in Rwanda. The 2014/2015 Rwanda Demographic and Health Survey (RDHS) data was used. The analysis was done using the structured spatial additive quantile regression model. The findings of this study revealed that the child’s age; the duration of breastfeeding; gender of the child; the nutritional status of the child (whether underweight and/or wasting); whether the child had a fever; had a cough in the two weeks prior to the survey or not; whether the child received vitamin A supplementation in the six weeks before the survey or not; the household wealth index; literacy of the mother; mother’s anemia status; mother’s age at the birth are all significant factors associated with childhood anemia in Rwanda. Furthermore, significant structured spatial location effects on childhood anemia was found.