Use of insecticide treated nets among women in Ghana: a multilevel modelling of the 2016 malaria indicator survey

Malaria is among the leading causes of morbidity and mortality particularly among women and children in sub-Saharan Africa and remains a public health challenge. Over the years, provision of insecticide mosquito nets (ITNs) for individuals and households have been on the forefront of various organizations. Focus has been on proportion of population with access to ITNs at the expense of usage proportion. This study examined the factors associated with the use of ITNs among women in Ghana. This study used a cross-sectional data obtained from the 2016 Ghana Malaria Indicator Survey (GMIS). A sample of 4,267 women aged 15–49 who possessed at least one mosquito net was included in this study. Use of ITNs was the outcome variable. Descriptive statistics and multilevel binary logistic regression were employed. Results were provided in tables with a signicance level set at P < 0.05.


Background
Malaria continues to be among the leading causes of morbidity and mortality particularly among women and children in sub-Saharan Africa. In spite of the attention given to malaria prevention and control and the resources invested into same (2.7 billion dollars in 2018 [1]), malaria infection remains a major public health challenge. Globally, more than three billion people are at risk of malaria. In 2018, there were estimated 228 million cases of malaria resulting in 405,000 deaths with children under ve accounting for 67% of deaths worldwide [2](WHO, 2020). Africa carries a disproportionately high share of the global malaria burden. In 2018, the region was home to 93% of malaria cases and 94% of malaria deaths [2]. Ghana is not spared by this public health menace. Malaria is responsible for approximately 1800 admissions at healthcare facilities and 10 deaths for every 100,000 population in Ghana [3]. In the rst quarter of 2020 alone, the country recorded more than a million cases of malaria with 54 deaths [4].
While indoor residual spraying, use of coils and other mechanisms continue to play a major role in malaria prevention and control [5], current policies in many of sub-Saharan Africa emphasize the use of insecticide treated nets (ITNs). In fact, ITNs are one of the most effective methods for preventing malaria, having averted an estimated 68% of malaria cases between 2000 and 2015 [6]. In 2007, the WHO recommended full coverage of ITNs for populations in areas at high risk for malaria transmission [7]. This has led to a massive scaling up of ITN distribution programmes aimed at providing enough nets for all households. These programmes are assessed by the proportion of households owning at least one ITN, and the proportion of people using ITNs.
Since 2010, an average of 54% of households in sub-Saharan Africa own at least one ITN, and 33% of the population reported using one the night [8]. In Ghana, the proportion of the population with access to an ITN within their household varies widely by region and wealth quintile [8].
While access to an ITN within the household is a good predictor of ITN use, it is noteworthy that contextual factors could also play a signi cant role. Researchers have attempted to explore the reasons for ITN non-use while others examined the determinants of ITN use among women or caregivers. The reasons for ITN non-use include discomfort [9], uidity of sleeping arrangements [10], and little perceived need to use a net when mosquito density is low [11] [9]. Age, education, household wealth [12] [13] [14], household size, distance to health facility, ideational factors [15] [16] [17], knowledge about e cacy of mosquito nets to prevent malaria, source of knowledge of mosquito nets and socioeconomic class [18] [19] are among the factors found to determine mosquito net use.
Since individuals are nested within communities, considering individual factors alone does not provide adequate explanation for ITN use. This notwithstanding, only few studies considered the effects of contextual factors on ITN use. A recent Ghanaian study [20] for instance explored how several factors affect ITN use. However, this study like others, did not consider wider contextual factors like community socio-economic disparity in their analysis. The purpose of this study thus, is to examine the effects of individual and contextual factors on ITN use among women of reproductive age in Ghana.

Study design
The dataset used in this study is a cross-sectional data obtained from the 2016 Ghana Malaria Indicator Survey (GMIS). It provides information on malaria prevalence and indicators in the country. The sampling frame used for the 2016 GMIS was strati ed and is the frame of the 2010 Population and Housing Census (PHC) conducted in Ghana. A two-stage sampling frame was used to select the sample for the 2016 GMIS. Each region was separated into urban and rural areas. In the rst stage, 200 Enumeration areas (EAs), including 93 EAs in urban areas and 107 EAs in rural areas, were selected with probability proportional to the EA size and with independent selection in each sampling stratum. The resulting lists of households then served as a sampling frame for the selection of households in the second stage. EAs that were large were segmented and only one segment was chosen for the survey. In the second stage of selection, a xed number of 30 households was selected from each cluster to make up a total sample size of 6,000 households and replacements for nonresponding households was not allowed. Detailed description of the sampling can be found at https://dhsprogram.com/pubs/pdf/MIS26/MIS26.pdf [21].

Data collection
Information from women aged 15 and 49 years who were either permanent residents or visitors present in the households a night preceding the survey was obtained through the administration of questionnaire. There were three types of questionnaires (Household, Woman's and Biomaker) and were written in English and three local Ghanaian languages (Akan, Ewe, and Ga). The questions were programmed into tablet computers and so allowed the use of computer-assisted personal interviewing (CAPI) for the survey. Issues that were captured included background characteristics, malaria prevention, malaria in children, malaria knowledge, among others. More details can be found elsewhere [21].

Outcome variable
This study concentrated on women between ages 15 and 49 years who possessed at least one mosquito net. A question which asked about sleeping under a treated mosquito net in the night preceding the survey was categorized as using ITNs while those who did not sleep under ITNs were categorized as not using mosquito nets. This was coded as a binary variable with a value of 1 and 0.

Independent variables
The variables considered included the following: age, education, household wealth, knowledge about causes of malaria, exposure to malaria messages, knowledge about the e cacy of mosquito nets, knowledge about coverage of malaria under National Insurance Scheme (NHIS), number of household members, place of residence, region and socioeconomic disadvantage. Age was categorized into 15-24, 25-34 and 35+ years. Education was de ned as no education, primary and secondary or higher. Household wealth was recoded into poor, middle and rich. Knowledge about causes of malaria was de ned as 'Yes' for women who reported that mosquitoes cause malaria and 'No' for those who reported otherwise. Exposure to malaria messages was de ned as 'Yes' for those who heard or saw messages about malaria 6 months preceding the survey and 'No' for those who were not exposed to messages of that nature. Knowledge about the e cacy of mosquito nets was measured as 'Yes' for women who agreed that the chances of getting malaria were the same whether mosquito nets were used or not and 'No' for those who either disagreed or claimed that they did not know. Knowledge about coverage of malaria under National Health Insurance Scheme (NHIS) was de ned as 'Yes' for those who agreed malaria was covered under NHIS and 'No' for those who disagreed or reported that they did not know. Number of household members was categorized into less than 5 and 5 or more. Socioeconomic status was obtained by applying principal component analysis technique to measure the proportions of individuals who were uneducated and poor within the communities. The resulting values was categorized into tertile 1 (least disadvantaged), tertile 2 and tertile 3 (most disadvantaged). Coding of independent variables were done based on previous works [22] [20].

Descriptive statistics
Percentage distribution of all the explanatory variables against the outcome variable were presented in a table. We used Chi-Square test to obtain p-values to indicate the level of signi cance of each variable.

Modelling approaches/ model t and speci cations
The study adopted the multilevel binary response models to estimate the odds and associated credible intervals of associations with mosquito net usage among women. There were four models in all. The rst model had no covariates (Null model) and only showed how the outcome variable reacted. The second model included the independent variables except socioeconomic status (a community level factor). The third model then introduced residence as a random coe cient to allow for urban-rural differentials by communities. Finally, model four then added community socio-economic status to attain the full model. Analysis was done using MLwiN command in Stata version 14.0 and models were tted using Markov Chain Monte Carlo (MCMC) estimation after using second order penalized quasi-likelihood (PQL2) to obtain sensible starting values for the model parameters [23]. No multicollinearity was identi ed amongst the covariates. Results were weighted and accounted for complex survey design.

Estimation technique
A two-level women-within-region variance components model was written as: Where Net_use ij is the binary response for whether a woman i in the region j uses ITNs. β 0 is the intercept for the log-odds of using ITN in the average household. U j is a household level random effect assumed normally distributed with a zero mean and constant variance σ 2 u . There was no woman-level residual error appearing in the linear predictor.
The between region variance is now a function of residence in the random effect part of the model.
var (u 0j +u 7j rural ij ) = σ 2 u0 +2σ u07 rural ij +σ 2 u07 rural 2 ij The nal model then introduces a community level explanatory variable, socio-economic status, to check whether certain regional-level variations in rural-urban areas may be explained.

Descriptive analysis
More than half (55.9%) of the respondents were users of ITN. Most of the respondents (60.1%) who used ITNs were aged 25 -34 years. Usage of ITNs was around 70 percent for women who were not educated and the poor. About 57 percent of women who had the opinion that malaria was caused by mosquitos and 54 percent who had exposure to malaria messages made use of ITNs respectively. Those who indicated the chances of getting malaria is not the same whether mosquito net is used or not used mosquito nets the most (56.4%). Most (60.6%) of the women who indicated malaria could not be treated as well as those who knew malaria to be covered under NHIS (58%) made use of mosquito nets. Moreover, close to 60 percent of the women who lived in households of more than ve members used ITNs.

Measures of association (random effects) Estimation/ model speci cation results
Model 0 shows a 0.215 log-odds of using ITN in the average region. This translates into a probability of 0.553. This indicates that almost 55% of the women used ITNs. The between-region variance is estimated to be 0.284.
Women's propensity to use ITNs was 0.079. This means that the expected correlation in the propensity to use ITNs between two women from the same region is 0.080. Detailed graphical statistical MCMC diagnostics are shown in appendix. At the second model (Model 1), it was realised that the between-region variance was estimated to be 0.072 compared to 0.284. This suggests that even though all the predictors were level 1, they explained away relatively more variation between regions than within regions. It was also found in model two that there was a greater community-level variation in the probability of using ITNs in rural areas than in urban areas. Since the residual between-community variation decreased in model three for rural areas compared to urban areas, some community-level variation in rural areas is explained by the differences in socio-economic status. The intra-community coe cient indicates that close to 26% of the variable in odds of using ITNs were linked to community level factors. It was found that women aged 15-24 were less likely to use ITNs compared to those 35 and above. This nding is similar to Adedokun and Uthman (2020) [22] but contrast reports from Malawi [16] and Ghana [31] [14] where older women were found to be less likely to use ITN compared with their counterparts aged 15-24 years. It could be reasoned that older women are more likely to own ITNs. Furthermore, in an Ethiopian study [18], women aged 26 years and above were about three times more likely to own ITNs compared with those less than 26 years. This could possibly explain why younger women in this study were less likely to use ITNs. The nding also suggests that young women may prove stubborn and not comply to sleeping under ITNs. One other reasoning could be that young women might prefer other means such as using mosquito spray and repellent.
This study reported that those with low education are signi cantly more likely to use ITN. This may probably be because of the perceived vulnerability which is higher among the poor or those with little or no education. This nding supports the ndings of Diema Konlan, et al. (2019) [31] where it was identi ed that caregivers with increased knowledge on malaria and ITN were less likely to use ITNs and vice versa. They further explain that as caregivers become more educated on risk factors of malaria and signi cance of ITNs, they tend to practice better environmental hygiene which reduces their utilization of ITNs. Thus, it is likely that community with low income and education as documented in this study may sleep under an ITN once it is available to them. However, this nding did not support an Ethiopian study [32] where higher education was reported to be signi cantly associated with use of ITN. Differences in the results may be due to the fact that the Ethiopian study was conducted in an urban setting [33].
Similarly, those within the poor wealth category were found to be more likely to use ITNs compared to the rich.
This nding corroborates earlier studies in sub-Saharan Africa [34], Ghana [35] and by the Ghana Statistical Service [36]. Contrasting ndings however indicated no associations of wealth with ITN usage [37] [38] [39]. However, while other studies [40] [41] indicated higher wealth status as associated with high utilization and low wealth status as associated with low utilization, one study found lower wealth index to rather be associated with high utilization [35]. It could be argued that utilization of ITNs may be high among the poor probably because they are unable to afford more "comfortable" (but expensive) means of malaria prevention, like the usage of mosquito repellents. Koenker and Yukich (2017) [9] identi ed discomfort (due to heat) as one of the key reasons for ITN non-use among women. Thus, individuals and households who are nancially sound resort to other means of malaria prevention rather than using ITN.
Rural residents were found to be more likely to use ITNs in comparison to those at the urban areas. This result is similar to Ricotta, Oppong, Yukich and Briet (2019) [20]; Min, Maung, Oo, Oo, Lin, Thi and Tripathy (2020) [42] and Oresanya, Hoshen and Sofola (2008) [43]. Conversely, a meta-analysis of Demographic Surveys [44] revealed that children in urban households were more likely to use ITNs than those in rural areas. Ricotta et al (2019) [20] postulated that those in urban areas were more exposed to messages through mass media as compared to the rural folks who normally tap messages from health workers. This nding calls for improvement in efforts and emphasizes the need for unique targeting of interventions as well as messaging and education to each setting [20].
Among the respondents, least and moderately disadvantaged women were more likely to use ITNs than mostdisadvantaged women. This nding is similar to a study in Malawi where women with high socioeconomic status were more likely to use ITNs than women with low socioeconomic status [16]. The result is also in line with ndings from lower middle-income countries where there were poor care-seeking among caregivers of under-ve children belonging to the lowest socio-economic status [42] [45] [46] [47]. The results of the present study suggest that women who are most disadvantaged should be prioritized in the mass distribution of ITN as those in this category can hardly afford ITN or the free distribution of the ITN has not reached the right community. In contrast to this nding, a study done in Malawi found no association between wealth, which is a component of socioeconomic status, and ITN usage [48]. Moreover, Adedokun and Uthman (2020) [22] found a contrasting result where the odds of using mosquito nets increased tremendously for women who lived in the most socioeconomically disadvantaged communities.
Surprisingly, the study revealed that those who knew malaria was covered under NHIS were more likely to use ITNs as compared to women who did not know malaria was covered under NHIS. This result may be due to the fact that women who knew malaria was covered under NHIS may also have been exposed to enough information about malaria with its effects, hence, the more likely that they will use ITN [49]. Additionally, the results may be attributed to both perceptions of susceptibility to and severity of malaria which is shown to have a positive association with ITN usage [20] and not merely being covered under NHIS. Also, women who believed that chances of getting malaria was the same with or without the usage of mosquito nets were more likely to use ITNs than those who disagreed. This result is consistent with the work of Adedokun and Uthman (2020) [22] who attributed this outcome to the fact that these women may have found other advantages of using mosquito nets such as prevention of bed bugs and other insects.

Study Strengths And Limitations
The cross-sectional nature of the data set did not allow for provision of information on causality of non-use of ITNs. Since some questions relate to events in the past, there was the possibility for recall bias while conducting the interview with respondents. Aside these limitations, this study made use of a large data set which provided the opportunity for generalization of ndings and also to countries with similar characteristics. Information on the role of individual as well as contextual factors in mosquito net use among women in Ghana were provided.