Utilization of Insecticide Treated Nets and Health Seeking Behavior for fever among Children under Five in Southwest Nigeria

Background Malaria remains a burden globally with the WHO African region accounting for 94% of the overall disease burden and deaths in 2019. Nigeria alone accounted for 27% of total malarial cases and 23% of deaths. Insecticide-treated nets (ITN) use and receiving early care for fever represent the most cost-effective means of malaria prevention and control. In this study, we assessed the utilization of ITNs in households with under-ve children in Southwest Nigeria, the health-seeking behavior of their caregivers for fever, and associated factors. Methods was


Background
Malaria is an infectious disease caused by species of the protozoan parasite, Plasmodium (predominantly P. falciparum) and, transmitted by Anopheles mosquitoes. Globally, there were an estimated 227 million cases of malaria in 87 malaria-endemic countries and an estimated 407,000 people died from malaria in 2019(1). Malaria is the third leading cause of death in children under ve years old (U5) years worldwide, after pneumonia and diarrheal disease, and accounts for almost one out of ve deaths in children less than 5 years (2). The proportion of total malaria deaths among children U5 was 67% in 2019(1).
The prevalence of malaria is highest in sub-Saharan Africa, which accounted for 94% of cases totaling about 215 million cases in 2019(1). Nigeria (27%), the Republic of the Congo (12%), Uganda (5%), Mozambique (4%), and Niger (3%) accounted for 51% of all cases globally(1). Sub-Saharan Africa also accounted for 94% (384,000) of malaria deaths. Nigeria alone accounted for almost 23% of malarial deaths globally(1). In 2009, the Nigerian Federal Ministry of Health estimated that malaria was a direct contributor to 11% of overall maternal mortality, 25% of infant mortality, and 30% of under-ve mortality. Children in regions of steady transmission are at increased risk of progression to severe disease compared to adults (3). Malaria also contributes to malnutrition in children, which indirectly leads to the death of half of all children under the age of ve throughout the world(4) (5). Southwest Nigeria has the highest malaria prevalence in children under 5 (U5) (50.3%) according to a previous study (6). However, the 2015 Malaria Indicator Survey (MIS) and the 2018 Nigeria Demographic Health Survey (NDHS) reported a prevalence of 32.1% and 28.9% respectively(7) (8).
Nigeria has implemented 4 national malaria strategic plans (NMSP) to date. The fourth NMSP (2014-2020) aimed to achieve pre-elimination status and reduce malaria-related deaths to zero by 2020. Two objectives of this plan were to provide at least 80% of the targeted population with appropriate preventive measures and to treat all individuals with con rmed malaria in the appropriate health facilities and with effective antimalarial medicines (7). Preventive measures to be implemented include sustained mass distribution of long-lasting insecticide nets (LLIN). Insecticide-treated nets have been established to prevent malaria by reducing the number of infective bites by mosquitoes. ITNs are cost-effective and evidence suggests that they result in a 44% reduction in the incidence of severe malaria episodes in persons who used ITNs compared to those who did not according to a systematic review (9).
Effective malaria case management is important for preventing morbidity and mortality among children under ve. This includes seeking early treatment (within 24 hours) at a health facility. In this study, we analyzed data to assess the utilization of ITNs among children U5 in the Southwest region of Nigeria and factors associated with the use of ITNs, health-seeking behavior of caregivers for fever in children U5, and associated factors.

Study Design
We conducted secondary data analysis on the data set obtained from the 2018 Nigerian Demographic Health Survey (NDHS 2018). The 2018 NDHS is a nationally representative survey conducted by the National Population Commission (NPC). ICF International provided technical assistance through the DHS program which is funded by the United States Agency for International Development (USAID). Other agencies were involved in providing technical and nancial support for the successful implementation of the survey.

Study Setting
Southwest Nigeria is one of the 6 geopolitical zones in Nigeria comprising of the following states: Ekiti, Lagos, Osun, Ondo, Ogun, and Oyo. The area lies between longitude 2 0 3 1 and 6 0 00 1 E and latitude 6 0 2 1 and 8 0 37 1 N with a total land area of 77,818km 2 . It has 2 distinct seasons which are the rainy season (April-October) and the dry season (November-March). The predominant language spoken in the region is Yoruba. However, several other ethnic groups exist. The population in Southwest is estimated to be 27,511,892 based on the NPC Census of 2006 (10). The majority of the occupants are farmers, traders, artisans, etc.

Study Population
The study focused on children U5 who were de facto members of households that owned at least 1 ITN in Southwest, Nigeria. Furthermore, a subset of children U5 who had a fever in the 2 weeks before the survey was also studied.

Data Collection
The 2018 NDHS was conducted from August to December 2018. The sampling frame used for the survey was the Population and Housing Census of the Federal Republic of Nigeria (NPHC) which was conducted by the NPC in 2006. The sampling for the survey was done using a strati ed sampling method conducted in 2 stages. Strati cation was achieved by dividing the 36 states and the Federal Capital territory into 2 strata: urban and rural. A total of 41,668 households were selected for the sample, of which 40,666 were occupied. Of the occupied households, 40,427 were successfully interviewed, yielding a response rate of 99%. Women aged 15-49 years old were interviewed in selected households and the women's questionnaire included questions related to children in these households. We obtained household ownership of at least one ITN by merging the Household member recode (PR) and the Household Recode (HR) data set. We further extracted data from the merged data set for children U5 who were de facto members of households that owned at least 1 ITN in Southwest Nigeria and had slept in the house the night before the survey. This resulted in a total of 1,887 children or 2,281 after applying sample weights. Assessment of health-seeking behavior for fever in children U5 was done using the children recode (KR) data set. We rst extracted data on children U5 living in Southwest Nigeria, resulting in a weighted count of 4,153, and then extracted data on children who had fever 2 weeks before the survey. This resulted in a total of 322 children or 360 after applying sample weights.

Data Analysis
Analysis was conducted using SPSS (version 23.0). The dependent variable for net utilization was whether a child U5 slept under an ITN the night before the survey in households that had at least 1 ITN.
The dependent variable for fever health-seeking behavior was whether a caregiver of a child U5 who had a fever within the past 2 weeks preceding the survey sought advice or treatment. This includes those who sought advice or treatment from the following sources: public sector, private medical sector, shop, market, itinerant drug seller, and community-oriented resource person. We excluded advice or treatment from traditional practitioners. Independent variables for net use were classi ed as child and household factors.
Child factors included the age and gender of a child. Household factors were type of place of residence, wealth index, age of household head, gender of household head, number of nets in a household, number of de facto household members, and state of residence. In addition to child and household factors, maternal factors were included as independent variables for fever health-seeking behavior. This includes women's age group, religion, highest education level, literacy level, and marital status. Descriptive statistics were conducted on dependent and independent variables. Chi-square was used to test the association between child factors, maternal factors, household factors, and outcome variables. The level of signi cance was set at 5% and 10% for ITN use and seeking advice or treatment of fever respectively.
Multivariate analysis was done using logistic regression to test the strength of association between the independent and dependent variables at a 95% con dence interval. The Complex Samples analysis method was used to account for the sampling technique and weighting factor applied in the survey.
Results are presented as tables and gures.

Results
Description of Sample Characteristics and utilization of nets for children U5 Table 1 is a description of sample characteristics of children U5 in households with at least one ITN in southwest Nigeria. The majority of children in this study were within the age group 12-35 months (41.3%, n=942) and 36-59 months (40.3%, n=918). About half of the children were males (51.7%, n=1,179) and a majority were urban residents (72.7%, n=1,659). The majority of children lived in households classi ed as rich under the wealth index (44.5%, n=1016). The proportion of households with males as heads of the household was higher than that of females (85%, n=1,940). The highest proportion of children were from Lagos (24.9%; n=568) and Oyo state (23.9%, n=544), n=522). Half of the children were from households that had only one ITN (50.2%, n=1146). The majority of children were from households with between 4 to 6 household members (65.6%, n=1497). About one-third of the children slept under an ITN the night before the survey (67.6%, n=1542, Table 1).    Health seeking behavior of caregivers in Southwest Nigeria for fever in Children U5 Table 4 is a summary of the health-seeking behavior of caregivers of children U5 who had a fever within 2 weeks preceding the survey. Figure 1 is a summary of where treatment was rst sought for children U5 with fever. Of the 4,153 children in Southwest Nigeria, 8.7%, (n=360) had a fever in the last 2 weeks preceding the survey. Of these, advice or treatment for fever was sought for 72.4% (n=261). The proportion of children for whom advice or treatment was sought for the same day or the next was 43.6%.
More than a quarter of respondents (27.6%, n=60) took their children to government health centers, while 17.7% (n=53) took their children to chemists/private medical stores. Table 5 is the result of bivariate analysis of independent variables of seeking advice or treatment for fever in children U5. Factors associated with seeking advice or treatment for fever were the age of the child, literacy level of the mother, highest educational level, sex of household head, wealth index, and state of residence (p<0.1). Table 6 is the result of multivariate analysis of factors associated with seeking advice or treatment for fever in children U5. Age of child, the religion of caregiver, and mother's educational level were associated with seeking advice or treatment for fever (p<0.05). Caregivers of children less than 12 months old and within 12-36 months were more likely to seek advice or treatment for fever than caregivers of children aged 36-59 months (aOR= 2.     (13). More importantly, this is higher than that reported in the 2015 NMIS report (39.2%), indicating an increase in the utilization of ITNs by children U5 (7). However, utilization by children U5 in southwest Nigeria was lower than the NMSP target of 80% (7).
Predictors of utilization of ITNs for children U5 were the age of a child, state of residence, number of nets in a household, and number of household members. Children aged 35 months and below were more likely to sleep under an ITN than older children. This is consistent with ndings from other studies in Myanmar and Malawi(14)(15). Children within this age group are more likely to sleep with their parents thus making it more likely for the child to sleep under an ITN. Children in households with a higher number of nets were more likely to sleep under an ITN than those with a single net. This implies that the more nets a household possesses, the more likely it is for children U5 to sleep under ITNs. Children in households with fewer members were more likely to sleep under an ITN than children in households with more members.
This is similar to ndings in a previous study in Nigeria and Ethiopia(16) (17). There was no statistically signi cant association between type of place of residence and children U5 sleeping under an ITN.
However, the proportion of children who slept under an ITN was higher in rural than urban settings. Children from households within the poorest wealth index were more likely to sleep under an ITN than children than those from households within the richest wealth index. This may be due to rich households being located in well-developed areas of states and the feeling of safety from mosquitoes by respondents. Association between caregivers' socioeconomic status and use of bed nets by children under ve was also reported in a study in Ebonyi, Southeast Nigeria (18). In another study, children from wealthier households were more likely to use the nets than children from poorer households in Tanzania and Uganda (19). The WHO World Malaria Report 2020 indicated that ITN use was generally pro-poor or close to a perfect equality in West African Countries(1).  (21)(25). This could be due to the presence of these stores in great numbers in communities of developing countries. The utilization of government and private health facilities should be encouraged as it is one of the objectives of the NMSP. Less than half of the respondents sought treatment for children with a fever the same day or next. This is similar to the ndings in a study in Ethiopia which reported only 38.7% of children received treatment within 24hrs of fever onset (25). Age of child, caregivers' religion, and education were predictors of seeking advice or treatment for fever. Seeking treatment was higher for children less than 12 months old than children aged 36-59 months. This is similar to ndings from a study in Burkina Faso (26). A study in Myanmar revealed that seeking treatment or advice for treatment was higher for children <36 months than older children. This may be due to caregivers perceived severity of fever in younger children compared to older children. Odds of seeking advice or treatment was higher in Muslim caregivers than Christian caregivers. Caregivers with higher education were more likely to seek care than those with secondary education. This is contrary to ndings in a previous study in Mozambique where caregivers with higher education were less likely to seek care (23). Mothers of children from Osun, Ekiti and Lagos were less likely to seek care than those from Ogun state.
This study was a cross-sectional study. Thus, a causal relationship could not be established between dependent and independent variables. The survey responses could also be affected by social desirability bias and recall bias as women were asked questions on net use the previous night and fever in children within 2 weeks preceding the survey. Nevertheless, the results of this study revealed important ndings that could help future policy planning.

Conclusion
This study showed that utilization of ITN and seeking care for fever in children U5 in Southwest Nigeria was below the national targets. The governments of Osun, Ekiti, and Lagos should develop and implement strategies to improve net use and health-seeking behavior for children U5.