Page 75 - GHES-2-2

Global Health Econ Sustain Prevalence and risk factors of childhood diarrhea

2. Methods (i) Maternal factors: Maternal age and the mother’s
education level were included as maternal factors
2.1. Data source and study participants determining childhood diarrhea.
The study utilized secondary data extracted from the (ii) Children-related factors: Out of several child-related
fifth round of the NFHS-5, conducted in India from factors influencing childhood diarrhea, children’s
2019– 2021 (International Institute for Population Science age, birth order, and size were identified as important
[IIPS] & International Classification of Functioning children-related factors affecting diarrheal disease
[ICF], 2019). The NFHS-5 is a nationally representative among children aged under five.
multistage sample survey designed to provide estimates (iii) Household factors: The number of household
of vital indicators at the district, state, and national levels. members, electricity services, caste/tribe, religion,
The primary objectives of NFHS-5 include gathering and wealth quintile were recognized as some of the
essential data on health and family welfare, as well as data significant household factors influencing childhood
on emerging issues such as fertility levels, infant and child diarrheal disease.
mortality, communicable and non-communicable diseases, (iv) Environmental factors: Type of toilet facility, floor
and other health and family welfare indicators across material, wall material, roof material, altitude from sea
various background characteristics. A total of 724,115 level, and regional division were considered responsible
women were interviewed from 636,699 households. environmental factors for childhood diarrhea.
Child-related information was provided by the mothers
on behalf of their children. All living children under the 2.4. Statistical estimation
age of five (N = 232,920) who participated in NFHS-5 Before analyzing the dataset, a thorough check was
were enrolled in this study. After applying inclusion and conducted to identify and address missing values and
exclusion criteria, removing missing values, and filtering irrelevant information. To adjust the clustered sampling
out unnecessary responses such as “don’t know” for each techniques employed in the surveys, a complex survey
variable, the final study participants were 161,368 children module was applied for all analyses, accounting for
residing in rural areas and aged below 5 years. Mothers primary sampling units, sample strata, and sample weight.
were queried about their children’s diarrheal status in Subsequently, descriptive statistics were carried out to
the 2 weeks before the survey. Information on diarrheal understand the distribution of diarrheal disease among
disease in under-five children was obtained through children under 5 years of age across socio-demographic
a women’s questionnaire. The datasets from Indian and environmental characteristics within the study
Demographic Health Surveys (NFHS) are freely available, sample. Moreover, the prevalence of diarrheal disease in
and one can access the dataset from the online repository rural areas among children under five years of age within
of the Demographic and Health Survey Program website 2 weeks preceding the survey was estimated using selected
through the following link: https://www.dhsprogram. explanatory variables. Pearson’s chi-square test was applied
com/data/available-datasets.cfm to investigate the bivariate association between childhood
diarrhea and different explanatory variables. Variables
2.2. Outcome variables exhibiting significance at the 5% level were considered for the
The study utilized responses on childhood (under the age regression analysis. A binary logistic regression model was
of five) diarrheal status in the 2 weeks preceding the survey performed to calculate odds ratios, elucidating the association
as a key outcome of interest. Childhood diarrheal disease, between childhood diarrhea and various sociodemographic,
which is referred to as the passage of three or more loose economic, and living arrangement characteristics. The
or liquid stools per day or a more frequent passage of liquid significance level was set at 5% for the regression model. All
stools than the normal state for an individual, was assessed. statistical analyses were performed using STATA version 15.0
This variable comprised two response categories: “yes” (StataCorp LLC, Texas, United States).
(indicating the presence of diarrhea) and “no” (indicating
the absence of diarrhea), which were coded as “1” and “0,” 2.5. Spatial analysis
respectively. The coding scheme for the outcome variable In this study focusing on the spatial distribution of diarrhea
was consistent with previous literature. disease at the district level in India, data from both NFHS-4
and NFHS-5 have been utilized. First, all the district-
2.3. Explanatory variables wise data were compiled using Microsoft Excel and then
All the possible risk factors associated with childhood exported into CSV-delimited format. Subsequently, the
diarrheal disease were categorized as maternal, child- CSV file was merged with the district shape file in Arc-GIS
related, household, and environmental risk factors: version 10.5 for further spatial analysis.

Volume 2 Issue 2 (2024) 3 https://doi.org/10.36922/ghes.2048