Multilingual abstracts

Please see Additional file 1 for translations of the abstract into the six official working languages of the United Nations.

Introduction

Malaria is a parasitic infection spread by the female Anopheles mosquito and is responsible for 660,000 deaths globally and over 219 million cases of infection annually [1]. High-income countries (HICs) have been able to eliminate malaria, while many parts of low- and middle-income countries (LMICs) are still struggling to deal with malaria and vector control [1]. In 2012, out of the 104 malaria endemic countries, 79 countries are in the malaria control phase, ten are in the pre-elimination phase, ten are in the elimination phase, and five are focusing on the prevention of re-introduction [1]. A number of factors account for the existing malaria burden in developing countries including climate change, infrastructure, emerging drug and insecticide resistance, massive population and demographic shifts, and costs of containment and therapy. In Sub-Saharan Africa, the wide-scale implementation of insecticide-treated bed nets (ITNs) has been one of the main strategies to reduce malaria morbidity and mortality [2].

There are existing interventions for malaria prevention, which include indoor residual spraying (IRS), ITNs, intermittent preventive therapy (IPT), presumptive treatment, and education. Previously, malarial chemoprophylaxis with chloroquine (CQ) was generally recommended for pregnant women in malaria endemic regions, however, with the recent increase in Plasmodium falciparum resistance to CQ [3–5], the World Health Organization (WHO) recommends the use of sulfadoxine-pyrimethamine (SP) for the intermittent preventive treatment in pregnancy (IPTp). In this paper, we aim to evaluate the effectiveness of community-based interventions (CBIs) and their impact on the prevention and control of malaria.

Methods

We systematically reviewed literature published before May 2013 to identify studies evaluating the effectiveness of the community-based delivery of interventions for the prevention and control of malaria as outlined in our conceptual framework [6]. We included randomized controlled trials (RCTs), quasi-experimental trials, and before-and-after studies, in which the interventions for the prevention and control of malaria were delivered within community settings and the reported outcomes were relevant. A comprehensive search strategy was developed using appropriate keywords, medical subject headings (MeSH), and free text terms. Searches were conducted in PubMed, Cochrane Libraries, Embase, and WHO Regional Databases. Studies that met the inclusion criteria were selected and double data abstracted on a standardized abstraction sheet. We excluded studies in which interventions were delivered in antenatal or immunization clinics, or if social marketing strategies in combination with facility-based interventions were evaluated. Studies were also excluded if the interventions were given to the displaced population or if the efficacy/effectiveness of a particular brand of bed nets, drugs, or diagnostic tools was evaluated. Studies reporting only entomological indices and parasite prevalence were also excluded. Quality assessment of the included RCTs was done using the Cochrane risk of bias assessment tool [7]. The outcomes of interest are outlined in Table 1. We conducted a meta-analysis for individual studies using the software Review Manager 5.1. Pooled statistics were reported as the relative risk (RR) for categorical variables and standard mean difference (SMD) for continuous variables between the experimental and control groups with 95% confidence intervals (CIs). We also attempted to qualitatively synthesize the findings reported in the included studies for other pragmatic parameters identified in our conceptual framework including intervention coverage, challenges/barriers, enabling factors, aspects related to integrated delivery, monitoring and evaluations and equity. The detailed methodology is described in a separate paper [6].

Table 1 Outcomes analyzed
Full size table

Review

We identified 1,146 titles from the search conducted in all databases. After screening the titles and abstracts, 187 full texts were reviewed, of which 42 studies (17 RCTs, 10 quasi-experimental trials, 13 before-and-after studies, and two case control studies) were included in the review (see Figure 1). The characteristics of the included studies are summarized in Table 2. Of the 42 studies, four studies could not be included in the meta-analysis as these did not report poolable data [8–11], while for studies reporting multiple evaluations of a single intervention, we pooled the results from the last reported survey [12, 13]. From the 17 RCTs included in this review, randomization was adequate in six studies, allocation was concealed in six studies, and adequate sequence generation was done in four studies. None of the studies blinded participants due to the nature of the interventions, while all studies provided insufficient information on selective reporting which limited us from making any judgment (see Table 3).

Figure 1
figure 1

Search flow diagram.

Full size image
Table 2 Characteristics of the included studies
Full size table
Table 3 Quality assessment of the included RCTs
Full size table

Twenty-five of the included studies evaluated the impact of the community-based distribution of ITNs or impregnated bed sheets, 14 studies evaluated IPT delivered in community settings, two studies focused on community-based education for malaria prevention, and one study evaluated environmental management through drain cleaning. Community education on malaria preventive measures was one of the components of interventions in most of the studies. All the studies were conducted in African countries except for three studies, one each from China, India, and Thailand. Interventions were non-integrated in 23 [10, 12–14, 16–18, 20–22, 28–33, 37, 39–42, 45, 47] of the studies, while in 17 [8, 9, 19, 23–27, 34–36, 43, 44, 46, 48–50] studies intervention was integrated with routine community-based antenatal care (ANC), primary healthcare (PHC), child health days or measles, and polio campaigns. The primary comparison was between the community-based delivery strategy versus routine or facility-based care. We also attempted to conduct a subgroup analysis to determine the relative effectiveness of integrated and non-integrated delivery strategies, according to the type of intervention and whether the evidence was from RCT/quasi-experimental studies or pre-post studies, where possible. The results are summarized in Tables 4 and 5.

Table 4 Results for the overall and subgroup analysis according to the type of study and treatment
Full size table
Table 5 Summary of evidence according to the type of intervention
Full size table

Quantitative synthesis

Table 4 depicts the impact of the overall community-based delivery of interventions and the subgroup analysis according to the type of study and intervention. Overall, community-based delivery of interventions to prevent and control malaria resulted in a significantly higher ownership (RR: 2.16, 95% CI: 1.86, 2.52) and usage (RR: 1.77, 95% CI: 1.48, 2.11) of ITNs in the intervention group as compared to the control group (see Figures 2 and 3). Ownership was defined as households having at least one net at the time of the survey, while usage was defined as having slept under an ITN the previous night or having an ITN hanging over the bed. The usage rate of ITNs among people who owned an ITN was around 66%. Community-based delivery strategy was also associated with significantly lower malaria incidence (RR: 0.70, 95% CI: 0.54, 0.90), parasitemia (RR: 0.56, 95% CI: 0.42, 0.74), and malaria prevalence (RR: 0.46, 95% CI: 0.29, 0.73) in the intervention group (see Figures 4 and 5). Anemia prevalence also reduced significantly (RR: 0.79, 95% CI: 0.64, 0.97) with sensitivity analysis after removing Eriksen 2010 (which reported concurrent interventions in both groups due to the introduction of a national campaign during the study period) (see Figure 6). We found non-significant impact on mean hemoglobin, splenomegaly, birth outcomes (low birth weight [LBW], prematurity, stillbirth/miscarriage), anthropometric measures (stunting, wasting, and underweight), and mortality (all-cause and malaria-specific). These findings are based on limited number of studies pooled with a high level of heterogeneity and hence should be interpreted with caution. The subgroup analysis for integrated and non-integrated delivery showed significant impacts on all outcome indicators except for malaria prevalence and splenomegaly which was non-significant for the integrated delivery subgroup, though this is based on a limited number of studies.

Figure 2
figure 2

Forest plot for the impact of CBIs on ITNs ownership.

Full size image
Figure 3
figure 3

Forest plot for the impact of CBIs on ITNs usage.

Full size image
Figure 4
figure 4

Forest plot for the impact of CBIs on parasitemia.

Full size image
Figure 5
figure 5

Forest plot for the impact of CBIs on malaria prevalence.

Full size image
Figure 6
figure 6

Forest plot for the impact of CBIs on anemia (a) with all studies included (b) after sensitivity analysis.

Full size image

Table 5 summarizes the evidence from the subgroup analysis according to the type of intervention. The community-based distribution of ITNs, impregnated bed sheets, and IRS resulted in significantly higher ITNs ownership (RR: 2.28, 95% CI: 1.95, 2.67), ITNs usage (RR: 2.49, 95% CI: 1.90, 3.27), and mean hemoglobin levels (SMD: 9.00, 95% CI: 8.80, 9.20). It also led to a significant reduction in parasitemia (RR: 0.58, 95% CI: 0.36, 0.94), malaria prevalence (RR: 0.42, 95% CI: 0.25, 0.70), anemia prevalence (RR: 0.49, 95% CI: 0.38, 0.62), and all-cause mortality (RR: 0.79, 95% CI: 0.64, 0.96). The community-based delivery of IPT significantly reduced parasitemia (RR: 0.54, 95% CI: 0.37, 0.81) and malaria-specific mortality (RR: 0.33, 95% CI: 0.20, 0.55). Community education and cleanliness campaigns alone did not show any significant impact on the outcomes measured.

Qualitative synthesis

Interventions delivered in community setups reported great potential to improve coverage, access, and adherence to ITNs and IPT, as these were delivered via community volunteers who were easily accessible and trusted resource persons and who could make regular home visits and follow up with their patients [34]. Delivering intermittent preventive therapy during childhood (IPTc) through community health workers (CHWs) has shown several advantages as CHWs are community residents and can not only deliver effective and timely treatment, but also remind mothers/guardians if they forget to attend treatment. Thus, operationally, delivery using CHWs was less restrictive and more convenient for parents and guardians [16]. Furthermore, CHWs also contributed to improving recognition and referral of seriously ill patients, provided advice on hygiene and nutrition, and encouraged women to attend antenatal clinics and to immunize their children [12, 13, 16]. However, whether to give incentives to CHWs to encourage sustainability is still debated as some mass drug delivery systems have been successful without incentives whilst others have employed financial incentives of some kind [16]. Studies suggest that financial payment and a strengthened drug supply may contribute to program success, however, incentives must reach the CHWs in a timely and efficient manner to avoid demotivation [16]. Besides incentivizing, CHWs require proper training, facilitation, and linkage with health units coupled with the provision of reference manual for easy and quick referencing to deliver timely treatment [12, 13, 34].

The integration of CBIs for malaria with existing ANC and immunization campaigns is reportedly more feasible and acceptable, and has reported improved coverage of IPTp to pregnant women [36]. The integration of malaria control programs with such promotional campaigns has resulted in a major increase in treatment coverage and ITNs distribution at a very low cost [23, 24, 36, 43]. Several features of ITNs distribution and mass measles vaccination campaigns favor the sustainability of an integrated approach. These include high coverage and low cost, as well as the fact that ITNs require replacing at the same intervals as measles vaccination campaigns take place [24, 27]. However, effective integration requires careful planning to ensure that each component of the package is not jeopardized by the other [37]. Some of the other strategies shown to achieve high and equitable coverage include mass free distribution and social marketing [10, 14, 24, 27, 30, 37, 41].

Factors enabling the delivery of CBIs mainly involved community empowerment, intensive social mobilization, and education [24]. The implementation of environmental management activities at the community level requires empowering local residents, developing a sense of ownership, and improving environmental responsibility among the population [17]. Providing incentives, social marketing, and subsidization of the costs of ITNs have also been reported as powerful tools, especially for programs seeking coverage for vulnerable groups such as children and pregnant women [30]. However, household ownership should be followed up to ensure usage. Studies have reported that barriers to ITNs use were not only the costs and access to ITNs, but also fear about insecticides and a lack of knowledge about malaria and ITNs [27]. Other reasons for not using an ITN included discomfort, problems with hanging up the nets and lack of space, low awareness of its need, and seasonal variations in use [51]. Community education together with other interventions for malaria prevention can have a substantial impact on increasing the usage of ITNs [15]. Educational interventions based on lectures and theoretical case studies without any follow-up training have been shown to be less effective than multifaceted interventions involving other strategies [18]. School-based interventions involving school teachers delivering educational messages through activities such as role-playing, poetry recitals, slogan chanting, song composition, and dramatization have reportedly been acceptable and effective [15], however, these require a well-established school health system [15].

Discussion

Our review findings suggest that community-based delivery of interventions to prevent and control malaria is effective in improving ITNs ownership and usage, and reducing malaria incidence, parasitemia, malaria prevalence and anemia. However, this strategy did not have a statistically significant impact on birth outcomes, anthropometric measures, all-cause mortality, and malaria-specific mortality. These non-significant findings could be attributed to the concurrent national level malaria control measures being implemented in many of the African countries which could have led to more effective malaria control measures in both the intervention and control clusters since most of the studies included in our review are from Africa. Furthermore, the non-significant findings could also be attributable to the limited number of studies included for some of the outcomes (see Table 4). Our findings suggest a decrease in anemia prevalence, however, the mean hemoglobin remained non-significant. This could be due to various other causes of anemia coexisting with malaria in the study population (including infections, infestation, and malnutrition), the limited number of studies included, and a high level of heterogeneity in the pooled analysis. Such interactions and effect modifications should be considered when interpreting these findings.

Evidence from the subgroup analysis according to the type of interventions suggests that community-based distribution of ITNs, impregnated bed sheets, and IRS can effectively increase ITNs ownership, usage and mean hemoglobin levels, and effectively reduce parasitemia, malaria prevalence, anemia prevalence and all-cause mortality. The community-based delivery of IPT was found to be effective in reducing parasitemia and malaria-specific mortality, while community education and cleanliness campaigns alone did not show any improvement in the outcomes measured. The non-significant impact of the various types of interventions could also be attributable to the limited number of studies included for some of the outcomes in the subgroup analysis and a high level of heterogeneity. We did not find any conclusive evidence on the relative effectiveness of integrated and non-integrated delivery strategies from our quantitative synthesis due to limited data in each subgroup. However, the qualitative synthesis of the data from the included studies suggests that high coverage could be achieved at a lower cost with the integration of CBIs with existing ANC and immunization campaigns. There is a need for large-scale effectiveness trials to explore the contextual parameters associated with scaling-up CBIs for malaria prevention and management.

Merely half of the countries with ongoing malaria transmission are on track to meet the World Health Assembly’s (WHA’s) target of 75% reduction in malaria cases by 2015. The WHO recommends prompt and effective treatment with artemisinin-based combination therapies, use of ITNs by people at risk, and IRS with insecticide as the key interventions to control malaria. The past decade has witnessed some tremendous expansion in the implementation of malaria control programs with an increase in international disbursements from less than USD $100 million in 2000 to USD $1.71 billion in 2010, mainly targeting the African Region [1]. Treatment of malaria among children and pregnant women has also expanded coverage in many countries recently. However, millions of people still continue to lack access to preventive therapies, diagnostic testing, and quality-assured treatment with an emerging challenge being the resistance to artemisinins and mosquito resistance to insecticides. Some of the other reported factors affecting the delivery, access, and use of interventions to prevent malaria include unclear policy and guidance; general healthcare system issues, such as stock outs and user fees; health facility issues stemming from poor organization and leading to poor quality of care; and poor healthcare provider performance. Key determinants of coverage include education, knowledge about malaria, socioeconomic determinants, and employment status [52–55].

With the recent increase in attention geared towards community-based delivery and its ever-expanding repertoire of services, systems must be identified to ensure training, support, and incentives for CBIs. Lack of supplies including ITNs and antimalarials have also been reported as a barrier to program success, and hence routine supply of these commodities needs to be guaranteed [16]. Another major barrier to sustainability of such programs is the issue of understaffing at health units [34]. There is also a need to address the challenges of fast growing cities and enhance the health conditions of its inhabitants [17]. Scaling-up environmental activities will require resources for initial massive cleaning and structural repairs that is not possible without donor support [17]. Community involvement remains an essential component of malaria control measures as these interventions require implementation at the household level and the disease is more prevalent in settings with limited access to health facilities. Building community ownership for creating demand for ITNs and increasing trust on CHWs is pivotal for any community-based program to be successful and improve health behaviors.

Interventions such as the distribution of ITNs, net impregnation, IPTp, and IPTc have the potential to be integrated with existing programs such as ANC, immunizations, deworming campaigns and child health days, and could provide useful models for evaluation. The WHO recommends seasonal malaria chemoprophylaxis to be delivered in integration with existing community-based programs, however, a single deployment strategy has not yet been devised [1]. Community case management (CCM) that integrates the management of childhood diarrhea, pneumonia, and malaria is one of the strategies that has received government support and has the potential to improve access. This program utilizes existing CHWs to treat children during home visits and has led to an improvement in ITNs usage and timely malaria treatment for children [56–58]. Similarly, a sharp increase in the number of distributed ITNs have been observed during child health days [59, 60]. Many countries in Africa have successfully implemented various integrated models of delivery, however, these are yet to be formally evaluated for effectiveness. A recent analysis carried out in Malawi, Rwanda, Kenya, and Senegal to determine the cost of providing integrated CCM concluded that this is associated with lower costs provided it is used by sufficient numbers of patients to justify the costs of training, equipping, managing, and supervising the CHWs who provide the services [61, 62]. However, simultaneous efforts should also be concerted to prevent over diagnosis and drug resistance.

Conclusion

Collaborative partnerships between governments and donors for the establishment of healthier environments for malaria prevention could play a crucial role in building an ideal platform for malaria specific interventions. Such programs could be articulated in collaborative partnerships between the government and various institutions such as engineering, waste management, education, and public health. We conclude that community-based strategy to deliver malaria specific interventions including ITNs, IRS and IPT, in combination with community education and sanitation, can be effective in reducing the overall burden of malaria morbidity and mortality, especially in malaria endemic areas.