Menu
Menu
Menu
Menu
Diarrhea is a leading cause of morbidity and mortality in the world but mostly in Sub-Saharan Africa. These could be prevented if universal coverage of current available interventions were implemented. The study aimed to identify factors associated with the choice of health care source among caretakers seeking treatment for under-5 children with diarrhea illness. Using women’s questionnaire we extracted a subset of data of children aged 0 to 59 months from the 2017 Demographic & Health Survey. Questions regarding history of childhood diarrhea for the past 24 hours or last 2 weeks prior to the survey were key in data extraction. Caregivers were asked to report the place where they sought treatment. In this study, 4 types of health facilities were defined: public, private, pharmacies, and other unspecified sources. A multinomial logistic regression model was used to identify sources of health facility used and corresponding factors associated with the choice. Factors associated with choice of health care source included education (educated women were less likely to self-medicate their children [relative risk ration (RRR) = 0.46; 95% confidence interval (CI) = 0.22-0.94]), income (better income earning families were more likely to seek care from private facility such as pharmacy [RRR = 1.87; 95% CI = 1.14-3.09]), and rural living (those in rural areas were more likely to seek treatment from other unspecified sources [RRR = 7.33, 95% CI = 1.40-38.36]). Public health facilities (79.9%) were the main source of health care service; however, reducing under-5 mortality due to diarrhea illness would require significant efforts to address other inequalities in accessing and utilizing health care services.
The 2015-2016 Demographic and Health Survey (MDHS) was conducted from October 2015 through February 2016 and sought to provide current estimates of basic health and demographic indicators of the population. The survey provided a comprehensive picture of the key social demographic and health challenges facing the Malawian population specifically focusing on maternal and child health. The Malawian Population and Housing Census conducted in 2008 served as the sampling frame for the 2015-2016 MDHS. This consisted of a complete list of all census standard enumeration areas (SEAs) defined as a geographic area that covers an average of 235 households. Hence, the census frame contained information about the location of the SEAs, the type of residence (rural vs urban), and the estimated number of residential households. In addition, Malawi has 3 main regions (North, Central, and South) divided into 28 districts. Using information from the sampling frame, each district was stratified into rural and urban denomination, which yielded 56 sampling strata. A 2-stage sampling approach was used for the 2015-2016 MDHS. The first stage of the survey involved a selection of 850 SEAs, including 173 SEAs in urban areas and 677 SEAs in rural areas with a probability proportional to the SEA size with independent selection in each of the sampling stratum. Within each of the selected SEAs, all households were listed; this listing served as a sampling frame for the selection of households at the second stage of the sampling process. In the second stage of selection, a fixed number of 30 households per urban cluster and 33 per rural cluster were selected with an equal probability systematic selection from the newly created household listing. All women aged 15 to 49 years, who were either permanent residents of the selected households or visitors who stayed in the households the night before the survey, were eligible to be interviewed. In about one third of all sampled households, all men aged 15 to 54 years, including those who were usual residents and others who stayed in the household the night before the interview, were eligible for individual interview. The women’s questionnaire gathered data from all eligible women, pertaining to their background characteristics, reproductive history, family planning, maternal and child health, breastfeeding and nutrition, marriage and sexual activity, fertility preferences, husbands, and background among others. Data obtained from the women’s questionnaire were used to extract a subset of data on children between ages 0 and 59 months. Women were asked whether any of their child had diarrhea in the last 24 hours or within the last 2 weeks prior to the survey. For each child with a known diarrheal episode within this timeframe, women were requested to report the place at which medical treatment or advice was sought for the last episode of diarrhea that the child had. Type of health care facility sought by the caregivers as the dependent variable was defined as follows: We isolated pharmacies, shops, and market as a single group because research has shown that in many Sub-Saharan African countries, informal pharmacies, drug shops, and markets are important channels for health care treatment.18,19 We wanted to test whether such assumption would hold true for children with diarrhea episode in the context of Malawi, yet we also strove to document the profile of children that were seeking care through those channels for diarrhea episode. Therefore, our dependent variable consists of 4 categories: (1) public, (2) private–non-pharmacy, (3) pharmacy, and (4) other (traditional healers, MACRO, etc). Independent variables included the following: Sources of drinking water was categorized into “improved” (piped into dwelling, piped into yard/plot, piped to neighbor, public tap/standpipe, or tube to well or borehole), “unimproved” (unprotected well, unprotected spring, and river or dam or lake or ponds), other sources (rainwater, cart with small tank, and other unspecified sources), and not de jure residents. Not de jure resident children were excluded from the final model because data were collected at the household level and assigned to individuals in the data. Therefore, most of the information regarding those children were missing. Summary measures including weighted frequencies and percentages for categorical variables were derived from the baseline characteristics of the study population. Univariate analysis was performed to document the association between socioeconomic and demographic characteristics of the study participants with the dependent variable, health care source. A bivariate and multivariate multinomial logistic regression models were fitted and relative risk ratio (RRR) with the associated 95% confidence interval (CI) were reported to investigate the relationship between dependent variable, health care source, and covariates. Besides, with regard to P value, as stated by the American Statistical Association (https://amstat.tandfonline.com/doi/pdf/10.1080/00031305.2016.1154108?needAccess=true), practice that reduce data analysis or scientific inference to mechanical “bright-line” rules (such as P < .05) for justifying scientific claims or conclusion can lead to erroneous beliefs and poor decision making. It is recommended that researchers should bring many contextual factors into play to derive scientific inference including the study design, the quality of measurements, the external evidence, and the validity of the assumption that underlines data analysis. Given the fact that we did not control for many contextual factors (road infrastructures for ease of access to health facilities, level of health professional training across different health facility types, etc) we reported also any P value <.10). This allowed us to consider any such factors that could be further investigated, while controlling for more contextual factors and external evidence. This was not required as data were extracted directly online from the National Demographic Health Survey of Malawi.
N/A
