Menu
Menu
Menu
Menu
Background: Poor people bear a disproportionate burden of malaria and prevention measures may not reach them well. A study carried out to examine the socio-economic factors associated with ownership and use of treated bed nets in Cross River and Bauchi States of Nigeria took place soon after campaigns to distribute treated bed nets. Methods. A cross-sectional household survey about childhood illnesses among mothers of children less than four years of age and focus group discussions in 90 communities in each of the two states asked about household ownership of treated bed nets and their use for children under four years old. Bivariate and multivariate analyses examined associations between socio-economic and other variables and these outcomes in each state. Results: Some 72% of 7,685 households in Cross River and 87% of 5,535 households in Bauchi State had at least one treated bed net. In Cross River, urban households were more likely to possess bed nets, as were less-poor households (enough food in the last week), those with a male head, and those from communities with a formal health facility. In Bauchi, less-poor households and those with a more educated head were more likely to possess nets. In households with nets, only about half of children under four years old always slept under a net: 54% of 11,267 in Cross River and 57% of 11,277 in Bauchi. Factors associated with use of nets for young children in Cross River were less-poor households, fewer young children in the household, more education of the father, antenatal care of the mother, and younger age of the child, while in Bauchi the factors were a mother with more education and antenatal care, and younger age of the child. Some focus groups complained of distribution difficulties, and many described misconceptions about adverse effects of nets as an important reason for not using them. Conclusion: Despite a recent campaign to distribute treated bed nets, disadvantaged households were less likely to possess them and to use them for young children. Efforts are needed to reach these households and to dispel fears about dangers of using treated nets.
In 2011, a household survey on prevention and treatment of childhood illnesses formed part of a programme to support evidence-based planning of health services in two states of Nigeria [10, 11]. The stratified, last stage random, cluster sample of enumeration areas from the 2006 census comprised 90 clusters in each state (Bauchi and Cross River): ten sites in each of three focus LGAs and 60 among the remaining LGAs, to give state-level representation. The cluster in each community comprised contiguous households radiating from a random starting point, to collect data on about 100 children under four years old. There was no subsampling within the cluster. Between July and September 2011, trained fieldworkers administered a questionnaire to mothers of children aged less than four years. The questionnaire asked about the mother’s most recent pregnancy and childbirth care and outcomes, and about childhood illnesses and treatment and related issues. It asked, for each child under four years old whether that child always slept under a treated bed net in the malaria season. The field teams also administered a questionnaire to each household about demographics and socio-economic status, which included a question about possession of any treated bed nets. They interviewed key informants in each community to get information about access to health services. Trained teams returned to the same communities in January 2012 and conducted separate male and female focus group discussions in each community. The participants for the separate male and female groups were drawn from among the households included in the household survey. Each group comprised some eight to 12 participants. The facilitators used a guide that presented the findings from the household survey about access to bed nets in each state and, based on this evidence, invited discussion about the perceived reasons for a lack of ownership and use. The trained reporters took notes during the discussions, and afterwards, together with the facilitators, prepared reports on the discussions. The Ministry of Health in each state gave formal ethical approval for the study (Cross River – reference number CRS/MH/CSG/E-H/018/Vol.1/23, dated 23 June 2011; Bauchi – reference number MOH/ASS/166/V.1, dated 16 June, 2011). The field team leaders sought consent for the survey from leaders in each community, and interviewers sought verbal consent from the head of each household, as well as from each individual respondent. Interviewers did not record any names or identifying information and were trained not to proceed with any interview unless they could do so without being overheard. Different operators entered the data twice with validation to minimize keystroke errors using Epi Info. Analysis relied on CIETmap open source software [12] that offers a user-friendly interface with the popular statistical programming language R. All estimates were weighted proportional to the population in each state, including rural and urban characteristics, and allowing for the over-sampling in the three focus LGAs in each state. The analysis handled the findings from the two states separately. There is no intention that the two states together represent the situation in the whole of Nigeria, and the overall project under which the survey was conducted focuses on supporting evidence-based health planning at state level [10, 11]. Bivariate and then multivariate analyses examined associations between potential determinants and the outcome of interest using the Mantel Haenszel procedure [13], adjusted for clustering [14]. The multivariate analysis started with saturated models of potential determinants, and backwards elimination, based on the cluster adjusted Mantel Haenszel Chi square, continued until only variables significantly associated with the outcome remained. The odds ratio (OR) with the cluster adjusted 95% confidence interval (CI) serve to describe associations in the analyses. A raster map of bed net coverage, created using CIETmap, combined the population relevance of each sample site with space (using inverse-distance weighted interpolation) to provide a population-weighted extension of each colour in the map legend [15]. The analysis examined associations with two outcomes: whether the household owned treated bed nets, and among households with treated nets, whether children under four years old always slept under a treated bed net during the malaria season. The equity-related variables at household level included: sex of the household head (male-headed or female-headed household), education of the household head (less than or more than junior secondary education), access to safe drinking water (‘safe’ sources including taps, bore holes with pumps and tube wells), whether the household had enough food in the previous week (as an indicator of absolute poverty), household construction (with good construction meaning zinc roof and concrete walls, as opposed to thatch/mud/timber), crowding (more than two people per room), occupation of the main breadwinner (lower or higher paying occupation), and perceived relative financial situation of the household (above or below the community average). At community level variables included: urban or rural location, electricity in the community and presence of a formal health facility in the community. The analysis considered additional factors in relation to whether young children always slept under a treated bed net: age and sex of the child, education of the parents, whether the mother had four or more antenatal visits in the last pregnancy, and number of children under three years old in the household (split between two or fewer and three or more). Due to interaction in the Bauchi model for bed net use, an additional variable combined maternal education and antenatal (ANC) visits (mother having some formal education + four or more ANC visits against all other combinations). A secondary analysis examined factors related to ownership and use of treated bed nets, excluding those LGAs in Cross River State not covered by the distribution campaign before the household data collection. Two of the authors conducted a thematic analysis of focus group responses on three topics: problems getting treated bed nets, why children do not sleep under nets even when the household has them, and what could convince people to use bed nets. The two investigators read through the focus group reports to identify common themes emerging for each topic, and extracted relevant quotes.
N/A
