Study area
This study was conducted among children in two randomly selected districts from province 1 (Jhapa) and province 2 (Bara). These are Terai (lower land) districts bordering to India. Jhapa district is better off than the Bara district in terms of human development index [27].
Study settings and participants
A cross-sectional study was conducted among children 6–59 months in April to June 2018. The sample size for the study was calculated using the formula, n=Z2p(1 – p)/d2 where n is the number of samples required, Z is the value associated with the 95% confidence interval, d is the precision margin of error at 5%, and p is the prevalence of diseases. For this study, we assumed 14.4% prevalence rate of acute malnutrition, 5% margin of error, and 95% confidence level [15, 22]. The sample size of 197 plus an addition 10% non-response rate yielded a total sample size of 217 for each district. The final sample size after data cleaning and excluding incomplete data from the two districts was 398. OTCC in a district were considered as strata. All OTCs, 16 from Jhapa and 19 from Bara, were included in the sampling frame and selected using a systematic sampling method. The required number of children was then divided by the number of OTCs and selected consecutively after a fixed interval.
Inclusion criteria
Eligible mothers of 6 to 59 months children attending OTCs were selected and interviewed.
Exclusion criteria
Children with known chronic illnesses, and congenital abnormality which affects the feeding pattern of the children were excluded from the study.
Questionnaire
The structured questionnaire was developed based on the study objectives. Indicators related to household socio-economic characteristics, education level, toilet facility, and food security [28], nutritional assessment, breastfeeding status, were validated and adopted from the 2016 NDHS questionnaire [15]. All the questions were pretested during the pilot test in a non-sampled OTC.
Outcome variables
Outcome of interest for this study was severe acute malnutrition among children aged between 6 to 59 months [3]. To identify acute malnutrition in children, measurement of MUAC less than 115 mm was used as cut-off criteria based on the 2007 WHO child growth standards [29].
Independent variables
Independent variables were categorized into three levels; household factors, child factors, and maternal factors. A robust literature review was done to regroup the potential factors associated with SAM. Household factors included variables such as place of residence (urban and rural municipality), family type, family size, ethnicity, household income, availability of toilet facility at the household, possession of land size, kitchen garden, and household food insecurity. The family type was classified as nuclear and joint, and the family size was categorized into two categories i) 1 to 5 members, and ii) 5 or more members. Ethnicity variable was grouped into three categories: relatively advantaged (Brahmin/Chhetri), relatively disadvantaged (Janajati/Muslims), and Madhesi and other unidentified. Madhesi is a predominant caste in Nepal who resident in Terai region [30]. The annual income of a household was asked with the respondent to determine the economic status of household. The household income was categorized based on annual family income standards as recommended by the National Bank (Nepal Rastriya Bank) [31]. The possession of land size was grouped into two categories: households having less than 0.5 hectors of land and the households having 0.5 or more hectors of land. Household Food Insecurity Access Scale (HFIAS) measurement tool was used to collect information on food insecurity at household level developed by the Food and Nutrition Technical Assistance Project (FANTA) [28]. Child factors comprised of gender and age of child, number of children in the household variables. Birth order was categorized as first, second, and third, or more. Likewise, the birth interval of between child was classified into two categories; i) less than 2 years and ii) more than 2 years. Maternal level factors included mother’s age, education, occupation (working and employed), breastfeeding practices such as colostrum feeding and exclusive breastfeeding practice. Maternal age was grouped as 15 to 20 years, 21 to 29 years, and 30 and above years. The mother’s education status was categorized as illiterate, literate, primary, secondary, and higher education. Early initiation of breastfeeding indicator included two categories i) within an hour (mother who breastfeed children within an hour of birth) and ii) delayed (mother who breastfeed children after one hour of birth).
Data collection and analysis
Face to face interview was conducted with a mother of eligible children by the trained enumerators using a paper-based structured questionnaire. Legibility and completeness of data were ensured during the data collection period and any inconsistencies were addressed during the fieldwork. Anthropometric tools, SECA digital weighing scale for weight and height board (Stadiometer) for height/length measurement of 6 to 59 months children were used. Shakir tape was used to measure MUAC from the child’s left arm to the nearest 0.1 cm (1 mm) margin. Child’s weight was measured with no or minimum layer of dress and all the measurement was taken during daytime as recommended by WHO 2006 growth standards [32]. Descriptive statistics of 398 children aged 6 to 59 months were presented as frequencies and percentages along with the calculation of Pearson’s chi-square test to determine associations between predictors and outcome variables. Also, we used multivariate logistic regression analysis to report the association between SAM and its determinants. All characteristics associated (p < 0.05) with each outcome in chi‐squared tests were included in the multivariable model. Odds ratios (OR) and 95% confidence intervals (CI) were derived and two-sided p-values less than 0.05 were considered as level of significant. No multicollinearity between independent variables was found. All analysis was performed in Stata software version 15.0 [33].
Ethical considerations and informed consent
The ethical approval for this study was granted by the Nepal Health Research Council, Kathmandu, Nepal. Eligible mothers of under-five years children who were willing to participate in the study were interviewed after obtaining written consent.