Decomposition analysis of the compositional and contextual factors associated with poor-non-poor inequality in diarrhoea among under-five children in low- and middle-income countries.

OBJECTIVES

The aim of the study was to assess the magnitude of wealth inequalities in the development of diarrhoea among under-five children in low- and middle-income countries (LMICs) and to identify and quantify contextual and compositional factors' contribution to the inequalities.

DESIGN

This is a cross-sectional study.

METHODS

We used cross-sectional data from 57 Demographic and Health Surveys conducted between 2010 and 2018 in LMICs. Descriptive statistics were used to understand the gap in having diarrhoea between the children from poor and non-poor households and across the selected covariates using Fairlie decomposition techniques with multivariable binary logistic regressions at P = 0.05.

RESULTS

Of the 57 countries, we found a statistically significant pro-poor odds ratio in only 29 countries, 7 countries showed pro-non-poor inequality and others showed no statistically significant inequality. Among the countries with statistically significant pro-poor inequality, the risk difference was largest in Cameroon (94.61/1000), whereas the largest pro-non-poor risk difference in diarrhoea was widest in Timor-Leste (-41.80/1000). Important factors responsible for pro-poor inequality varied across countries. The largest contributors to the pro-poor inequalities in having diarrhoea are maternal education, access to media, neighbourhood socio-economic status, place of residence, birth order and maternal age.

CONCLUSION

Diarrhoea remains a major challenge in most LMICs, with a wide range of pro-poor inequalities. These disparities were explained by both compositional and contextual factors, which varied widely across the countries. Thus, multifaceted geographically specific economic alleviation intervention may prove to be a potent approach for addressing the poor and non-poor differentials in the risk of diarrhoea with policies tailored to country-specific risk factors. There is a need for further investigation of factors that drive pro-non-poor inequalities found in 9 of the LMICs.