Background
Yemen is one of the six countries in the Eastern Mediterranean region that continues to have areas of high malaria transmission and contributes to the majority of malaria cases in the Arabian Peninsula [
1,
2]. A regional strategy with the ‘malaria-free Arabian Peninsula’ initiative by the year 2020 was proposed by the Eastern Mediterranean Regional Office of the World Health Organization (WHO, EMRO) in 2004 to maintain the malaria-free status in the Gulf states and support malaria elimination from Yemen [
2,
3]. The current updated Yemeni National Malaria Control and Elimination Strategy (NMCES) for 2014–2018 supports this initiative with the overall aim to eliminate the disease from the country by the year 2020 and includes collaboration with the Gulf Cooperation Council countries for funding, specifically the coordinated Saudi-Yemeni cross-border vector control and surveillance activities [
2,
4].
Two mainstream vector control intervention tools, indoor residual spraying (IRS) and insecticide-treated nets (ITNs) in the form of long-lasting insecticidal nets (LLINs), form the vector control component of the NMCES (2014–2018) [
4]. Based on the present malaria stratification in Yemen, universal coverage with LLINs, alone or in combination with IRS, is implemented in three altitude-based epidemiological strata; stratum 1 (0–600 m above sea level), stratum 2 (601–1000 m above sea level) and stratum 3 (1001–1500 m above sea level) [
4]. In Yemen, LLINs were first introduced in 2006 [
5,
6], and prior to 2011, their distribution only targeted the vulnerable population groups, children under 5 years of age and pregnant women. Distribution of LLINs is free of charge to all ages through mass distribution campaigns, which is currently the only distribution channel for this intervention in Yemen [
4]. One LLIN for every two people, as recommended by the WHO [
7], is distributed with the aim of protecting rural populations within the targeted malaria-endemic areas at altitudes between 0 and 1500 m above sea level, which include the highest burden governorates (Al Hudaydah and Hajjah) [
4].
The
‘proportion of households owning at least one ITN’ and ‘
proportion of children under 5 years or pregnant women who slept under an ITN the previous night’ were the previously recommended two principal indicators to measure the ownership and use of ITN as a malaria prevention tool [
8]. However, these indicators are limited by not identifying if actual use was due to inadequate ITNs within a household or due to behavioural factors [
9,
10]. Eisele et al. [
9] reported that only by achieving intra-household universal access of ‘
two people per ITN’ can surveys interpret actual use among children under 5 years and pregnant women, following which behaviour change communication (BCC) programmes can then reduce the gap between ITN use among these vulnerable populations within households with access to ITNs.
Following revision of the indicators by the ‘Survey and Indicator Guidance Task Force’ of the RBM Monitoring and Evaluation Reference Group (MERG) in 2011, additional ‘‘
new core indicators were proposed: the proportion of households with at least one ITN for every two people and the proportion of population that had access to ITN within the household” [
10,
11]. Two malaria indicator surveys (MIS) were conducted in Yemen during 2009 and 2013, both of which did not include assessing the proportion of population that had access to LLIN within the household and the use among the population with access [
12,
13]. Therefore, this study assessed the universal LLIN coverage by applying the indicators approach developed and recommended by the MERG and identified the factors associated with not using LLINs among people with access to LLINs (one LLIN for every two people) in universally covered malaria-endemic areas of Al Hudaydah in the Tihama region of Yemen.
Discussion
The present survey aimed to assess ownership and use of LLINs in malaria-endemic areas of Al Hudaydah targeted for universal coverage. Despite the ownership of at least one LLIN by 90.6% of the total households surveyed in the present study, only about a quarter of these households had one LLIN for every two members, the target for universal coverage [
7]. Therefore, the LLIN coverage in the present survey exceeds that reported by the 2013 Yemen MIS, where 13.0% of households owned at least one LLIN and 1.7% of households had at least one LLIN for every two people in Al Hudaydah [
13]. It is noteworthy that the latter MIS was conducted before the mass distribution campaign that targeted Al Hudaydah which might explain the variation between the findings of the two surveys. The inadequate availability of LLINs among these surveyed communities is still evident, where only half of the de facto population had access to LLINs, a finding that could be interpreted by the fact that this survey was conducted 3 years following the mass distribution campaign. This finding is similar to a study in Southwestern Ethiopia where half of the population (51.9%) had access to LLINs indicating that there is still a wide access gap in these malaria endemic settings targeted for universal coverage [
22]. A much lower proportion of population with access to LLINs has been reported in Congo, 3 years following a mass campaign [
23]. In the latter study, in a 2-month pre-distribution survey, the proportion of households with at least one LLIN for every two people was 4.1% and the population with access to a LLIN was 22.2% [
23]. A decrease in coverage and ownership of LLINs for any reason over time has been reported following distribution campaigns with varying rates in a number of post-distribution surveys in other countries [
24‐
27].
In Yemen, replacement campaigns are planned 3 years following the mass campaigns, as recommended by the WHO [
7]; however, there are no additional continuous distribution channels currently in place for the provision of nets for additional members following the mass campaigns. Furthermore, there is no monitoring on the durability or loss of the net for replacement which might lead to the reduction in the coverage level [
26,
28‐
30]. The use to access ratio in this study was 0.37. It is noteworthy, that analysis of data from 93 household surveys in 44 countries, assessing ownership, access and use, have reported that a use to access ratio of less than 0.60 is considered poor and reasons for non-use of the available nets should be investigated [
20]. Although increased use of ITNs/LLINs among people with access was reported [
10,
22,
25,
31,
32], ownership has not been consistently translated to use [
33,
34]. The low usage rate of LLINs among people with access could also be attributed to behavioural factors. Kilian et al. [
35] reported that multi-channel BCC campaigns influenced the use of LLINs, with a significant increase of LLIN use by vulnerable populations. BCC activities are included as an important strategic component in previous and current Yemen’s NMCES (2014–2018) towards achieving malaria control and elimination. A communication for behavioural impact (COMBI) plan (2009–2012) was developed in 2009, an approach for BCC, particularly aimed to increase use of LLINs [
5]. However, BCC activities in Yemen continue to be under key challenges such as ‘weak activities’ with limited allocation of resources that include both financial and staff for implementation [
4,
6].
Despite the low access of children under 5 years to the LLINs (13.7%), a higher proportion (42.5%) of those having access actually slept under them during the night preceding the survey. On the other hand, low access and usage rates were observed among pregnant women (16.4 vs. 20%). MIS 2013 reported that 26.2% of children under 5 years and 29.0% of pregnant women slept under LLINs, respectively [
13]. As mentioned previously, it should be noted that proportions of children and pregnant women using LLINs in the MIS were calculated from those households with at least one LLIN (not universally covered). In addition, the different sample sizes used in the two surveys could have contributed to such variations. During the present survey, only 10 of 61 pregnant women had access to LLINs. Therefore, it is rather difficult to compare the proportions of using LLINs in the present survey with the 2013 MIS findings [
13].
In pursuit of unveiling the barriers to LLIN use in Al Hudaydah, the association of several factors with not using LLIN among individuals with access was assessed. Proportion of population not using a LLIN during the night preceding the survey in Al Mansuriyah was three times higher than those in Ad Durayhimi. Local cultural or behavioural factors might affect LLIN usage as reported in other previous studies elsewhere [
35,
36], and such factors need to be further investigated.
Age was significantly associated with LLIN usage, where children under 5 years were the highest of all age groups having slept under LLINs during the night preceding the survey. As documented in literature, this vulnerable age group usually represents the priority household category to use a net [
37‐
41]. Nevertheless, more than half of children under 5 years are still not using LLINs, highlighting the continuous vulnerability of this age group to malaria infection risk in the surveyed areas targeted for universal coverage. Although the proportion of individuals not using LLINs increased with age, with a higher proportion of LLINs non-use among older children aged 5–15 years, there was no significant difference compared to children under 5 years. The latter older age group encompasses children of school-age and studies have reported low use among this age group [
40,
41]. In a recent study in Malawi, school-age children (11–15 years) showed significantly lower bed net use as a result of low access to bed nets within the household [
42]. In another study in Malawi, the highest malaria prevalence was observed in school-age children (6–15 years), and this highlights the importance of bed net use among this age group as one of the preventive measures against the disease [
43]. In the present study, adolescents over 15 years and adults (age groups 16–45 and >45 years) were significantly less likely to use LLINs compared to children under 5 years, which is in agreement with the findings of previous reports conducted in Liberia [
44] and Nigeria [
45]. However, other studies reported a higher use of bed nets among older age groups [
24,
32,
46]. Both the school-age children and adults are usually identified as asymptomatic parasite carriers within malaria-endemic areas and, therefore, contribute to malaria transmission [
43].
The poor physical condition of the LLINs inside households, such as the presence of apparent holes, affects their use. In the present study, having three or more damaged nets was significantly associated with a higher risk for non-use of LLINs. This is in line with the findings of several previous studies elsewhere [
36,
38,
47,
48], while Kilian et al. [
26] reported that the increasing number of net holes was not a determinant factor of decreased use in Nigeria. In the present study, such bad physical condition of the observed LLINs could have been due to their use since 2013. Studies have shown that with increasing years, poor fabric integrity is evident and thus the serviceable life of a LLIN is reduced leading to the decrease of use [
28,
47,
49]. The life span of LLINs can vary in different regions, as shown in studies investigating the durability of LLIN. In Rwanda, a study monitoring the durability of LLIN reported a high number of damaged nets; from five to nine out of ten remaining LLINs were damaged, 2 years following distribution campaign [
28], while in Nigeria the net serviceable condition varied in the three states surveyed, Nasarawa, Cross River and Zamfara, with ‘
an estimated median net survival of 3.0, 4.5 and 4.7 years’, respectively [
26]. Therefore, there is a need for regular monitoring of the physical integrity of the distributed LLINs to determine whether the nets are still in serviceable condition during the duration of the recommended 3 years life span and until the targeted year for the replacement campaigns.
The type of house structure was significantly associated with non-use, where living in huts was significantly associated with a higher risk of not using the LLINs compared to living in other more typical houses in the area. Being rectangular in shape, it is possible that there were difficulties in hanging the LLINs in the circle-shaped huts, which may discourage their use. In this context, inadequate space and house structure in the form of huts have been reported as factors affecting the use of bed nets [
36,
50].
It should be noted that risk assessment of not using LLINs in the present survey is limited by the small sample size that might affect the study of some variables such the presence of pregnant women inside households.
Authors’ contributions
SMAA, MAKM and AMA designed the project. SMAA and AMA conducted the field work. SMAA and MAKM analysed the data and interpreted the results. SMAA drafted the manuscript. SMAA, MAKM, AMA and RA revised the manuscript. All authors read and approved the final manuscript.