Background
The World Health Organization’s first aim towards achieving a world free of malaria is to ensure “universal access to malaria prevention, diagnosis and treatment” [
1]. Malaria prevention using insecticide-treated bed nets (ITN), and treatment with artemisinin-based combination therapy (ACT), are cost effective, and have reduced disease incidence and prevalence in the African population [
2,
3]. For pregnant women, intermittent preventive therapy in pregnancy with sulfadoxine–pyrimethamine (IPTp-SP) is a cost effective intervention to prevent adverse effects of malaria in pregnancy for both women and the unborn child, and reduces disease burden in stable transmission areas [
4‐
7]. However, the lack of equitable access (availability, affordability, and acceptability) [
8] and coverage affect utilization of the interventions in poor and rural communities [
9,
10]; poor communities have a disproportionately higher burden of malaria infection than wealthier communities [
11]. Other key factors affecting utilization depend on the quality of the health system and traditions of behaviour in the community [
12].
Although the prevalence of
Plasmodium falciparum parasitaemia in children under 5 years old in Malawi has declined between 2010 and 2014 according to the 2014 Malawi Malaria Indicator Survey (MMIS), the prevalence is three times higher in rural than urban areas [
13]. Given that the burden of malaria is high in rural communities, determinants of community utilization of the efficacious interventions need to be identified and specifically addressed to improve and maximize their effectiveness in malaria control. While national-level surveys provide statistics on the access and utilization of malaria control interventions, these aggregated survey estimates may mask substantial variations in intervention coverage that exist in rural communities [
14].
In this study, data from repeated cross-sectional malaria indicator surveys were analysed to quantify access and utilization of adequate prompt diagnosis and treatment of reported fever in children and women in a rural Malawian community. ITN usage (before the ITN distribution campaign), and an evaluation of access to and utilization of adequate IPTp in pregnant women are described and compared to 2014 MIS. Finally, factors associated with ITN and IPTp utilization and care-seeking for fever in the rural community are investigated.
Discussion
The study findings suggest suboptimal utilization of malaria control interventions in the study area. ITN ownership and usage, prompt care-seeking for diagnosis and effective treatment of malaria, and IPTp-SP rates in these communities are lower than both national estimates, and universal coverage targets. This highlights that national coverage aggregations mask substantial inequities in intervention coverage [
10] and are not ideal for identifying and targeting high burden and low intervention marginalized communities. District-based and fine-scale surveillance may improve identification of these marginalized communities. The Malawi National Malaria Control Programme conducted free nationwide ITN distribution campaign in April 2016, and increased availability of free diagnosis and treatment of malaria in both public and private health facilities. Recently, provision of uncomplicated malaria diagnosis and treatment for children below 5 years old by community health workers has increased availability and access in remote communities. MMP also implemented a community-led behaviour change programme to improve care-seeking and ITN use. An evaluation of access and utilization of the interventions following their scale up and MMP community-led intervention will be published.
There were delays in care-seeking for fever for both children and women, with some not seeking care at all (Table
2), similar to other studies in Africa [
24]. Only a small proportion of participants (19.5%) sought care within 24 h of symptom onset. Untreated symptomatic or asymptomatic cases are parasite reservoirs for continued transmission [
25]. Although distance to a health facility has been reported in the same area to affect care-seeking [
22,
23], distance was not a significant factor in the current study. Distance to facility may have been an important factor in previous reported studies which had a wider variation in distance (communities located 8 km or more were compared with those close to the hospital) than the current study (less than 3 km on average). In the current study, an estimate of the distance between the household and a primary health facility was a Euclidean distance calculated based on GPS locations. This estimation method may not reflect the true distance of the actual path between the household and health facility which is determined by geographical (i.e. terrain) and man-made barriers. The lack or delay in accessing diagnosis and treatment of fever may be more related to socio-cultural factors which were not investigated in this quantitative analysis. MMP is currently conducting qualitative research and implementing a malaria behaviour change communication (BCC) strategy [
26] within the community to understand and improve, respectively, socio-cultural factors affecting care-seeking.
For ITN ownership and usage, the findings suggest national ITN coverage estimates, which are reported every 2 years, underrepresent rural communities. Although this study was conducted in one rural community, household ITN ownership (35.3%) and under five ITN usage (33.5%) are lower than the 2014 MMIS estimates of 70.2 and 67.1%, respectively [
13]. The national surveys use self-reported responses to estimate ITN usage, similar to the current study. Low ITN ownership and usage highlight the gap in ITN coverage for poor communities similar to findings in Uganda and Tanzania [
27,
28]. Even within these rural communities, diminishing wealth was associated with lower ITN usage, highlighting health inequity for the poor [
29]. Household ITN ownership was previously reported to be associated with lower odds of parasitaemia in children in this community [
17], meaning that the children in households without ITNs are not only at higher risk of malaria infection, but also serve as a source of malaria parasites to surrounding communities. These poor households and communities are not identified in national surveys for targeted interventions. Evidence for improving household ownership of ITN through mass distribution campaigns has yielded mixed results in sub-Saharan Africa; in some countries, equity between the relatively wealthier and the less wealthy has increased, while in others, it decreased [
30‐
32]. Deliberately addressing inequity and specifically developing strategies aimed at improving community socio-economic status may improve malaria control for poor communities [
33]. For instance, fine-scale mapping of malaria burden and intervention coverage can potentially assist identifying and targeting marginalized communities. Surveillance using district-based health teams may also increase fine scaling mapping and identifying such communities. Most rural areas have community health workers who can be utilized for this role.
The proportion of women confirmed to have taken the recommended three or more doses of IPTp-SP (15.0%) was lower than those who took two doses (40.6%) and comparable to 2014 MMIS estimates (12.6%). The proportion is significantly lower than 31% reported in World Malaria Report 2016 for 20 African countries [
3]. Three SP doses are associated with higher neonatal birth weights, reduced risk of low birth weight and reduced risk of placental malaria, compared to two doses [
34,
35]. No quantitative variables were significantly (statistically) associated with completion of IPTp-SP doses. The lack of significant variation in the predictors and sparse data, account for this finding.
The implementation of malaria interventions through the health system were suboptimal. Although reported utilization of ANC services was high, the supply of IPTp-SP was inadequate (Fig.
1), as 14.9% of women who attended ANC did not receive it. This discrepant use of IPTp-SP needs further investigation from both health system and patient perspective. For those who sought care for fever, the number who had a finger prick for malaria test was 41%, far below the universal access to malaria diagnosis target. Health system barriers for IPTp-SP and diagnosis and treatment of malaria were not investigated in this study, although in other studies, these affect rural areas disproportionately to urban areas. Barriers include availability of medical supplies and understaffing, [
14] as most health care workers do not want to work in remote areas.
The survey relied on interviewee responses, which are standard in most national household surveys. However, documented information on IPTp-SP use in health passports was used as an additional more reliable source of information. Reported IPTp-SP and ITN use during surveys are susceptible to recall and social desirability bias [
36‐
38]. For IPTp-SP, there was no difference (statistically) between interviewee response and documented information, suggesting that interviewee responses were reliable. For ITN use, a previous study in Malawi validated the accuracy of care-giver response for evaluating ITN use in children; the study was however conducted in a community with higher ITN ownership (following an ITN distribution campaign) than the current study and also reported a low accuracy for people who reported no ITN use [
39].
Although this study was conducted in one rural setting, the situation may not be unique to this community. More fine-scale studies need to be conducted focusing specifically on rural regions, where the majority of people in developing countries reside, to improve malaria control.
Authors’ contributions
All listed authors were involved in designing the primary study. ANK, MGC, RSM and ZT implemented the study. MGC sampled the households for rMIS. ANK and MGC analysed the data. ANK wrote the first draft. All authors contributed in developing the manuscript. All authors read and approved the final manuscript.