Using malaria diagnostic test results obtained from Demographics and Health Surveys (DHS), which are nationally representative surveys, we estimated malaria prevalence in each concerned country.

DHS data, used for quantitative analysis of ITN use, were from DHS 2017-2018 for Burkina Faso (malaria prevalence estimated with microscopy test was 17%) and from DHS 2017 for Senegal (malaria prevalence estimated with microscopy test was 1%). We used DHS data from 2010 to 2018 to estimate longitudinal progress concerning malaria prevalence in Burkina Faso and Senegal. Senegal (16 million inhabitants) is a west African country with a warm and tropical climate. The northern part of Senegal is very hot (dry Sahelian plain). Burkina Faso (20 million inhabitants) in the center of west Africa, is a landlocked Sahel country. The incidence of malaria in Burkina was 400 per 1,000 population at risk and in Senegal 400 per 1,000 population at risk, in 2018 (World Banque. Country Report) [3]. DHS survey is a cross-sectional, nationally representative survey carried out in developing countries. Firstly, each country was divided into small geographic areas (clusters) and in each cluster, three strata were created: towns, cities and rural or urban areas. In the second degree, households were selected (a two-stage sampling procedure) [2].

DHS surveys also collect socio-economic data, which we used in this study: urban or rural residence, economic level of the household, possession and use of an ITN, education level in the household, age of head of household, number of rooms in the house and number of household members. The construction of wealth quintiles is based on some considerations of the household population. Information such as water supply, type of vehicle, type of flooring, radio, television, refrigerator, electricity, domestic servants, ownership of agricultural land, sanitation facilities, or country-specific items was used for determining wealth level.

ITN ownership was measured with the number of ITNs in a household and ITN use was measured with the number of under five children number sleeping under ITN the night before the survey.

The prevalence of malaria was estimated using the population category of children aged 6-59 months in all selected households. During DHS surveys, a child infected with malaria was determined by several methods such as microscopy or rapid diagnostic test (RDT). For a rapid diagnostic test (RDT), the result of which, available after 15 minutes, was immediately communicated to the parents or persons responsible for the child. Positive cases received drugs considered to be first-line treatment in the country. To validate the results of the rapid malaria test, a microscopy test was performed. Blood smears were dried, fixed with methanol and packed. Results from microscopy testing were used for the present study [12].

We added information from a synthesis of official literature. The literature used was from the health ministries of each country and international authorities and agencies. We used methods proposed by Judith Green and Nicki Thorogood and we used a criteria grid. We selected official reports from 2010 to 2018, which accurately report national malaria programs in Senegal or Burkina Faso [13, 14]. The literature selected were WHO reports, President’s Malaria Initiative reports, United Nations Children’s Fund (UNICEF) reports [1, 12, 15, 16] and documents from the Ministries of Health of Senegal and Burkina Faso. The criteria grid used for this synthesis of the literature on malaria programs included eight important sub-groups: funding sources, entomological monitoring, use of ITNs, use of insecticide, malaria case management, health system organization, communication and surveillance.

Statistical analyses were performed using IBM SPSS version 20.0 and SAS studio. We performed a linear regression to investigate the relationship between the number of ITNs and the number of persons in a household, adjusted for the age of household head, the economic level of the household, the highest education level in the household and the area of residence (rural or urban). The normality and homoscedasticity of regression models were verified by observing the shape of points, curves in plots, and scatter plots of residuals. Figures representing R-squared in each model were shown with their confidence intervals at 95%.

We estimated the mean number of persons using ITNs among groups (urban or rural areas, wealth level, highest education level in the household and age of household head) and in each country using analysis of variance which is an important tool for multiple comparisons of groups.

The ANOVA provides much information, but it does not provide detailed information about differences between specific study groups. It was used to select a model that fits the best situation according to socio-economics factors [17].

We evaluated the importance of socio-economic factors in our linear regression model for the relationship between the number of ITNs in a household and the number of household members by computing the R-squared of each model. We must note that this evaluation was performed considering only areas of intermediate malaria endemicity (prevalence 5-40%) and high endemicity (prevalence >40%). Areas of low malaria endemicity (malaria prevalence <5) were not considered in this estimation.

Semi-partial correlation (correlated to R-squared) provides a solid estimate of the relative importance of each predictor in a multiple regression model. Studies have demonstrated the unique contribution of each independent variable in a multiple regression model. The square of semi-correlation represents how much the R-squared value decreases in a model when a predictor is removed. We used the decrease in the R² value when a co-variable is dropped from the model to assess the importance of a factor in the model for each country (Senegal or Burkina Faso) [18].

Like almost all African countries, Senegal and Burkina Faso have national malaria programs supported by international and other external funds [32-35]. Most of the funding for the fight against malaria in both countries comes from international partners such as the United State National Institute of Health, the United Kingdom Department for International Development, the European Commission, the Wellcomtrust and the Bill and Melinda Gates Foundation.

The big question is whether those external funds are adequately distributed according to the needs of the national malaria programs in Burkina Faso and Senegal [33, 35]. Head, who investigated malaria research funds in African countries from 1997 to 2015, found that Burkina Faso is in the group of the highest-funded African countries concerning research on malaria (more than $ 30 million) and Senegal is in the second highest-funded group of countries ($ 5-30 million) [36].

It is crucial that the use of funds responds to the real needs of a country. A study by Barrenho demonstrated the importance of effective coordination among different donors of malaria funds if a good level of performance in relevant policies is to be achieved [37].

Entomological research is an important part of research for implementing regional policies against malaria. In Senegal, entomological monitoring is conducted several times a year in each district [22] and in Burkina Faso, it is conducted once a year in some districts. This does not necessarily indicate that one country’s entomological monitoring is more effective than the others. It can be a difference in strategy based on the country's context and stage of intervention in the fight against malaria. Burkina Faso may have a few highly relevant sentinel sites, a system that can be as effective as covering each district. As malaria has almost been eliminated in Senegal, a lot of active surveillance and control, including vector surveillance, is necessary. It is important for Senegal to have an entomological surveillance site in each district, but it is not necessary for Burkina Faso if fewer sentinel sites can adequately cover the whole country. Studies have found that several changes occur in mosquito behavior, mosquito abundance, composition, and dominant vectors, sometimes due to policies implemented in the fight against malaria such as the use of ITNs, the use of insecticide, or chemoprevention. Entomological monitoring is an important element in the initial stages of the implementation of a malaria intervention policy and for surveillance [38].

Both countries have applied a policy of free distribution of ITNs (1.8 ITNs per person every 3 years) throughout the population. Senegal does more with the additional free distribution of ITNs every year in targeted districts and to targeted populations, such as children under five.

Knowing that Burkina Faso is more populated than Senegal and the proportion of the poorest households in Senegal (31.0%) is triple the proportion of the richest in Burkina Faso (8.8%), this is particularly important for household ownership of bed nets and use. It must be noted also that wealthy people may have access to good housing, good sanitation and environmental management, aid conditioner, etc. All factors are determinants to be considered when providing interpretation. Quantitative data used in this paper is about the possession of ITN and the use of ITN. Senegal and Burkina Faso have the same proportion (51%) of households in which all children under 5 sleep under an ITN. In general, the wealth level of households has a positive relation with ITN possession but a negative relation with ITN use. This must be considered in the management of the fight against malaria. As mentioned before, although studies have demonstrated that the poorest houses are potentially at higher malaria risk than the richest houses, this should be relatively discussed with the fact that wealthy people have more access to a clean environment, access to other means of protection from malaria and may not need necessary bed nets compared to the poorest household. In Senegal, there is a significant difference between the mean number of ITNs owner in households in rural and urban areas, but in Burkina Faso, there is no significant difference. A study by Thwing explained that the difference between ITN numbers in urban and rural households in Senegal was due to the fact that Senegal also applied free ITN distribution to children under five. As rural households had more children under five than urban households, it was logical that there were more ITNs in rural areas than in urban areas in Senegal [39].

For both countries, in rural areas, there is a little more use of mosquito nets than in urban areas (not a significant difference). Due to suitable environments for mosquito breeding sites, malaria is mainly transmitted in rural areas and can influence the use of ITN. The study of Mfueni found that there is more risk of malaria in rural areas than in urban areas [5].

Several studies have demonstrated that the use of ITNs can greatly reduce malaria risk by reducing the rate of mosquito bites and the density of mosquitoes [23]. Wealth level is an important factor in ITN possession in African countries [30]. We found that poor households had the highest mean number of ITN use in both Senegal and Burkina Faso. Studies have demonstrated that the poorest houses are potentially at higher malaria risk than the richest houses [31].

The possession of ITNs does not necessarily equate to the use of ITNs. The use of ITNs in a population is strongly associated with the education of the population on the prevention of malaria. Senegal and Burkina Faso have implemented communication policies appropriate for the local context and level of education. We found that in Senegal and Burkina Faso, there is no significant difference in the use of ITNs in relation to the highest education level in the household. The wealth level of households is the most important factor for ITN possession in Burkina Faso (R²=0.096) but not in Senegal (R²=0.166) for the model of the relationship between ITN possession with some socio-demographic factors as shown in Table 3.

Resistance tests for insecticide are performed in Senegal and Burkina Faso. A policy of indoor residual spraying has not been applied in Burkina Faso, whereas Senegal has implemented indoor residual spraying in districts where the incidence of malaria is higher than 50 per 1,000 persons. Studies have demonstrated the effectiveness of indoor spraying in the fight against malaria. A study by Pluess using randomized comparison found that indoor residual spraying significantly reduces the prevalence of malaria in unstable settings [40, 41]. In Senegal, there has been a reduction of immunity in the population. The risk of a rapid resurgence is therefore highly possible if all areas are not under control. A policy of indoor spraying is an effective tool for rapidly decreasing mosquito vector capacity in targeted areas. It must be noted that the use of ITNs in both countries has also contributed to a decrease in vector capacity.

Insecticide resistance research is being conducted in the two countries to determine which insecticide can be used against local malaria vectors. The study of insecticide resistance in the fight against malaria is very important and can influence malaria control in several ways. Resistance to insecticide may produce a gap between the entomological studies and the efficacy of epidemiological studies [24].

First line treatment is the same in the two countries, but we found a great difference between Senegal and Burkina Faso concerning the diagnosis of cases. As malaria has almost been eliminated in Senegal, more targeted case management in high-transmission areas is necessary. Home-based diagnosis is performed and health workers with the objective of rapidly decreasing malaria transmission give malaria treatment to positive cases. Senegal introduced home-based management of malaria cases in 2008 with rapid diagnostic tests and artemisinin-based combination therapy as a treatment for positive cases [42].

A study by Landier demonstrated the importance of early diagnosis and treatment in the reduction of malaria transmission. In the case of P. falciparum malaria cases, treatment within 48 hours from the onset of fever is a preventative action against malaria transmission. Without early diagnosis and treatment or with the incomplete treatment of victims following bites from infected mosquitoes, gametocyte remains in those individuals for several days after clearance of asexual parasites. They remain infected and can transmit malaria [25].

Senegal’s national malaria program is decentralized, but the opposite is the case in Burkina Faso. A centralized program is a system in which the federal government makes the decision for all areas in the country. Decentralization allows Senegal to implement a more effective malaria program according to the context of each district, considering, for example, dominant vector species, the quantity of rainfall and temperature, and the education level of the population. As a country close to eliminating malaria, Senegal must closely monitor all districts to avoid a resurgence of the disease and maintain very low transmission. A study published in 2016 demonstrated the benefits of local contextualization of policies implemented in the fight against malaria [43].

Local context is more widely used in communication policies for the national malaria program of Senegal and Burkina Faso. The most appropriate methods of communication for reaching a targeted population must be tailored for different regions. The culture of each region or its socio-economic conditions must be considered if the best results are to be achieved from malaria programs [44]. In Senegal, research policies on the feasibility of interventions and their acceptability are integrated into the national malaria program [28-31]. Several studies converge, finding that it is better to work within a local than a national context to eliminate this significant public health problem in Africa.

In Senegal, a demographic and health survey with epidemiological centers for surveillance in each district is organized every year. In Burkina Faso, routine data are collected from all districts and data from surveys (questionnaires) by mobile phone. We can observe that malaria surveillance implemented in Burkina Faso is more passive than active, while in Senegal (a country close to malaria elimination), malaria surveillance is more active than passive. Active surveillance contributes to the reduction of malaria transmission and will facilitate malaria elimination. Testing and treating during active surveillance enable the identification of people who do not go to hospitals and facilitates the early detection of malaria cases, which reduces malaria transmission. A study by Singh found that active surveillance is an effective element in policies implemented for malaria elimination [45].

Although the strength of our study was to combine information from a synthesis of the literature with quantitative data, there are some limitations. Due to the unavailability of data in the two countries, the quantitative analysis only focused on the relationships between the number or use of ITNs with the number of people in a household (with socioeconomic and demographic factors controlled for). Therefore, the results of the analysis show and focus only on the vector interventions while information on other policies was collected during DHS surveys but is not available. The results of this study highlight the ITN net interventions with the correlation to household size. It is the partial segment of the malaria control (elimination) policy that cannot conclude all other confounding factors related to malaria. It lacks, for example, additional quantitative analyses of the availability and quality of diagnostic tests, treatment and other interventions. We hope that future studies can address this limitation. Considering that the prevalence of malaria and the use of ITNs were estimated on children under five years old, it is necessary to take a step back and consider confounding elements between the wealth level of households and the use of ITNs (number of children in the household, rural/ urban areas, quality of housing, number of rooms). We would like to use more demographic factors in our models. Another demographic information associated with the household in our DHS data was the sex of the head of the household (apart from the age of the head of the household). We did not use the sex of the head of the household in our statistical models because almost all the households involved in the data used had male household heads.