Does Socioeconomic Environment Is A Risk Factor For Malaria In Senegal ? A DHS Data Analysis of Malaria Trends From 2010 To 2016 In Senegal

Analysis of the evolution of malaria will help address the determinants of malaria elimination in this country. The aim of this study is to analyze the evolution of malaria in Senegal from 2010 to 2016. Methods This article uses data from the Senegalese Demographic and Health Surveys (for 2010-2011, 2012-2013, 2014, 2015 and 2016. To assess the factors associated with the positivity of the RDT, a multivariate logistic analysis was conducted to account for the effect of confounding factors. Adjusted odds ratios were calculated with their 95% condence intervals. The dependent variable was the result of the Malaria rapid diagnostic test.


Introduction
The ght against malaria and its possible eradication have been priorities of the international community since the 1950s and 1960s [1,2]. However, Africa still remains the epicentre of this endemic. With the scarcity of resources following the halt of the global malaria eradication plan (UNICEF and other donors having withdrawn their nancial contributions), the disease has been completely neglected for decades, resulting in the resurgence of malaria in areas where it had been reduced with outbreaks leading to high mortality [3,4]. The WHO African Region accounted for 93% of global malaria deaths in 2017; however, it accounted for 88% of the 172 000 fewer malaria deaths compared to 2010 [5].
However, it should be noted that in recent years, some countries in sub-Saharan Africa have signi cantly reduced the prevalence of malaria and in some regions are in a state of elimination, although other countries in sub-Saharan Africa remain countries with high malaria endemicity.
In recent years, Senegal has seen a marked decline in malaria incidence, which fell by more than 50% between 2009 and 2015. In fact, parasite prevalence fell from 3-1.2% and all-cause mortality from all causes from 72% of live births to 33‰ among children under 5 years of age between 2009 and 2014. These convincing results have enabled Senegal to reach the Roll Back Malaria targets in 2015 [4]. However, several studies have shown that the risk of malaria infection could be correlated with factors related to the level of development of countries. For example, the risk of malaria infection has been correlated with the level of household wealth. [6,7]. Studies have also shown that the quality of housing construction is an important risk factor for malaria [8]. Some of these factors identi ed in the scienti c literature might suggest that despite health interventions, malaria elimination would be di cult to achieve in sub-Saharan Africa. The aim of this study is to analyze the evolution of malaria in Senegal from 2010 to 2016. representative at the national level, at the regional level, for urban and rural areas, and at the level of the 14 regions of Senegal. The DHS sample is drawn stratum by stratum. Thus, in accordance with the DHS methodology, the sample is based on a strati ed, two-stage, areal sample drawn in accordance with the DHS sampling methodology [9]. For the DHS in Senegal, at the rst level, the survey covers 400 clusters (Primary Survey Units PSU) which are drawn from the list of Enumeration Zones established during the General Census of Population and Housing, Agriculture and Livestock, using a systematic draw with probability proportional to size, the size of the UPS being the number of households [9]. A count of households in each of these clusters provides a list of households from which a second-stage sample of 22 households per cluster was drawn, in both urban and rural areas, with a systematic draw with equal probability. In 2010-2011, 7902 households were surveyed; in 2012-2013, 4131 households were surveyed; in 2014, 4231 households were surveyed; in 2015, 4511 households were covered by the DHS.

Data
During these surveys, malaria parasitemia tests were carried out. Malaria parasitemia tests had been performed on children aged 6-59 months. Two tests for the diagnosis of malaria were performed: a Rapid Diagnostic Test [e.g. Rapid Diagnostic test (RDT)], the results of which were communicated to the parents/caregiver, and a thick drop. Children who tested positive on the RDT were referred to a health service by the survey laboratory technicians according to the protocol in effect. This study uses the results of the RDTs.

Dependent variable
The dependent variable was the result of the Malaria rapid diagnostic test. This variable is a binary DHS variable which has 2 modalities "Positive" and "Negative".

Independent variables
The study considered explanatory variables related to socio-economic and demographic factors. These included individual factors, household factors and factors related to means of prevention.
With respect to (A) individual factors: (1) Age was analyzed by 1-year age group. (2) sex was a 2-modality binary variable. (3) The variable "Child works on a farm or with animals" was also binary (B) Household factors: (1) Household wealth -The wealth index, a measure of relative economic well-being based on household assets, was ranked by quintiles (lowest, second, middle, fourth, highest). (2) Residence -The place of residence has been dichotomized into "urban" or "rural". (3) The construction material of the houses with the following variables-The variable "Main wall material (wall)" was generated from an EDS variable with 14 modalities : No walls, Cane / palm / trunks, Dirt, Bamboo with mud, Stone with mud, Adobe uncovered, Plywood, Cardboard, Reused wood, Cement, Lime stone / cement, Bricks, Cement blocks, Adobe covered (Adobe is sun-hardened brick, made from mainly clay soil, diluted and mixed with straw or chopped dry grass), Wood planks / shingles, Other. It has been recoded into 2 modalities "improved material" and "unimproved material". The "improved material" modality corresponded to the modalities Cement, Limestone / cement, Bricks, Cement blocks, Adobe coated, Wood planks / shingles. The variable "Main roof material" was generated from an EDS variable with 12 modalities: No roof, Stubble/palm leaves, Grass, Rustic carpet, Palm/bamboo, Wood planks, Cardboard, Metal, Wood, Scale/Cement bre, Ceramic, Cement and Roof shingles. It has been recoded into a binary variable with two modalities "improved material" and "unimproved material". The "improved material" modality corresponded to the Metal, Wood, Cement Fibre/Calamine, Ceramic, Cement and Roo ng Shingles modalities.
With regard to (C) factors relating to means of prevention. The following variables were concerned. (1) The variable "Existence of windows with mosquito nets in the house" had 2 modalities "yes" and "no". (2) The variable "Child sleeps under mosquito nets all night" was generated from a variable "Children under 5 years who slept under mosquito nets last night". This variable had 4 modalities "No", "All children", "Some children", "No nets in the household". The variable was recoded into a binary variable with two modalities "yes" and "no", the yes modality corresponding to children having "slept under a net last night". (3) the variable "Use of mosquito nets as a means of prevention outside sleeping rooms" was also binary.

Analysis
The analysis was carried out using STATA/SE 15. As noted above in the section on data source, a two-stage sampling design was adopted. To account for the multi-stage sampling design of the survey, all data were weighted to account for disproportionate sampling and non-response.

Participants
In our study, a secondary analysis of the Senegal DHS (2010-2016) data was performed. Participants from urban and rural areas were selected from all 14 administrative regions of Senegal. The study focused on the diagnosis of malaria by RDTs in children under 5 years of age. The population size for our study was 408 969 children across 5 years. The ow diagram of the study population is represented in Fig. 1.

Socio-demographic characteristics Table I : Distribution of individual and household characteristics by year
In the descriptive analysis, the variables were presented in terms of frequency and percentage of data. Inter-group comparisons were made using the Chi2 test. The signi cance level was set at 5, and 95% con dence intervals (CI) were used. Variables where p was less than 0.2 in the bivariate analysis were selected for the multivariate analysis. [10].
To assess the factors associated with the positivity of the RDT, a multivariate logistic analysis was conducted to account for the effect of confounding factors. Adjusted odds ratios [ e.g. Adjusted odd ratio (ORa)] were calculated with their 95% con dence intervals. To handle complex sampling (multi-stage sampling, weighting and strati cation), the identi cation variables for weights, strata and primary sampling units (PSUs) were de ned before using the svy command (survey pre x STATA). The logistic model was tted with the independent variables mentioned above.

Evolution of malaria cases from 2010 to 2016
The malaria prevalence rate varies from 3.01% in 2010 to 0.87% in 2016 (Table II).
The evolution of malaria cases shows an increase in Survey 2 which corresponds to the years 2012-2013.
This results in a signi cant and steady reduction until 2016 (Fig. 1).  This study showed a prevalence of malaria in Senegal of 0.87%, which places Senegal among the countries with elimination potential.
Analysis of the evolution of malaria in Senegal has shown that there has been a slight increase in the prevalence of malaria in Senegal from 2010 to 2013 from 3.01 to 3.65%.
From 2013 onwards, there will be a gradual reduction in the malaria prevalence rate to 0.87% in 2016.
Despite its proximity to some countries in the West African sub-region, Senegal has been able to make signi cant progress in reducing its malaria prevalence rate. These countries bordering Senegal, although having similar Sudano-Sahelian climates, have higher malaria prevalence than those of Senegal. In Mali the prevalence of malaria according to the RDT is 19% [12], in Guinea Conakry 26% of children are reported to have been treated for malaria [13], in Burkina Faso the prevalence of malaria is 17% [14].
Senegal's low prevalence could be attributed to Senegal's health policy in the ght against malaria. In fact, many interventions are implemented in Senegal, such as chemo-prevention of seasonal malaria, universal coverage with impregnated mosquito nets and intermittent preventive treatment for pregnant women.
Chemo-prevention of seasonal malaria using sulfadoxine-pyrimethamine plus amodiaquine, administered monthly during the transmission season, is recommended for children living in areas of the Sahel where malaria transmission is highly seasonal. Chemo-prevention of seasonal malaria's recommendation is currently limited to children under ve years of age, but in many areas of seasonal transmission, the burden of older children may justify the extension of this age limit. The implementation of chemoprevention started in Senegal in 2008 and was generalized in all districts in 2010. Evaluations of its effectiveness in Senegal had shown that the introduction of seasonal chemo-prevention of malaria in children was effective and associated with an overall reduction in malaria incidence [15,16]. Studies have shown that malaria morbidity is higher in children under 7 years of age [21]. This study showed that the age of the child was a factor associated with the evolution of malaria. Children aged 4 Burkina Faso and Pakistan, compared to children living in the richest wealth quintile, those in the poorest wealth quintile were more exposed to malaria [25,23].
Other protective factors against malaria identi ed by this study were the construction characteristics of the houses. These were houses with improved wall and roof materials with aOR 0.45 [0.24-0.85] and 0.48 [0.25-0.93] respectively. The results of our study were similar to those of a systematic review of studies evaluating the relationship between modern housing and malaria infection (n = 11 studies) and clinical malaria (n = 5 studies). The latter study showed that residents of modern homes were less likely to be infected with malaria than residents of traditional homes [8]. The 2015 study concluded that future research should assess the protective effect of certain house characteristics and incremental housing improvements associated with socio-economic development.
As in our study in Nigeria, also the nding was that children who lived in houses built entirely of unimproved materials were more likely to be infected with malaria (aOR = 1.

Limits
Some limitations were identi ed in this study. These include differences in the speci city and sensitivity of the diagnostic methods (RDT and microscopy) and the way the sampling strategy was conducted, which may be biased in favor of certain individuals in the population. This study used only the results of Rapid Diagnostic Tests.

Conclusion
This study was able to conclude that the factors associated with the evolution of malaria in Senegal are the age of the child, the level of wealth of the household and the type of dwelling in the house. Health policies in Senegal should consider the protection of the 4 to 5 age group that is vulnerable to malaria infection. Educating Senegalese communities about the malaria-protective role of modern housing and encouraging the use of improved materials for house design (roof and wall) in addition to other malaria prevention strategies could help strengthen malaria control measures for children under ve years of age.
It should be noted that the level of development of countries, by in uencing better living conditions for communities, remains an important prerequisite for the elimination of malaria in the African sub region and in Senegal in particular.

Declarations
Availability of data and material Data supporting the conclusions of this study are available from MEASURE DHS. Data are available from MEASURE DHS upon request and with the permission of http://www.dhsprogram.com.