Introduction

Aedes aegypti and Aedes albopictus mosquitoes are competent vectors for several arboviruses, including dengue (DEN), yellow fever (YF), chikungunya (CHIK) and Zika (ZIK), which significantly impact human health. The year 2023 witnessed the effects of the El Niño phenomenon leading to a global surge in dengue cases, with simultaneous outbreaks worldwide [1]. In the same year, an increase in autochthonous dengue cases was recorded in the European mainland, with even higher numbers recorded in 2024 [2]. Over the past five years, the increase in dengue cases has been particularly pronounced in the Region of the Americas, with Brazil having the highest burden [3].

In sub-Saharan Africa (SSA), arboviruses have been reported to circulate silently and actively [4], contributing to acute non-malarial febrile illnesses, with dengue and chikungunya being among the most common [5,6]. Although there is increasing evidence of Aedes-transmitted disease in SSA, based on sero-surveys [7] and outbreak investigations [8], information on their epidemiology remain fragmented. Surveillance in this region largely focuses on outbreaks, and deaths caused by YF continue to occur despite the availability of an effective vaccine [7]. The World Health Organization (WHO)’s 2023 report on the global dengue situation lists Africa among the top four regions most affected by arboviral diseases. Outbreaks have been reported in 15 of the 47 African countries [1], and dengue is potentially endemic in at least 34 African nations [9]. The same year, Burkina Faso experienced a dengue epidemic with 70.433 probable cases and 709 recorded deaths [10]. Mathematical models further suggest that the burden of dengue in Africa is comparable to that in the Americas, while the disease receives considerably less attention [9]. The discrepancy between reported cases and model prediction may be due in part to the nonspecific symptoms of Aedes-transmitted diseases, which often mimic malaria leading to misdiagnosis in the absence of arbovirus-specific diagnostic tests.

Besides vaccines for YF that are existing since a long time, recently new vaccines are becoming available against Aedes-transmitted diseases, such as dengue and chikungunya [11]. Further investigation is needed into vaccine efficacy and their contribution as a component of control programmes. Moreover, proper clinical management of human cases remains a challenge. Thus, preventing transmission through effective surveillance and vector control (VC) strategies is crucial to reduce the disease burden and to avert severe public health consequences. The presence of Aedes mosquitoes in SSA is well documented, with Ae. aegypti and Ae. albopictus prevalent in urban and peri-urban environments [12]. However, there is growing evidence of their genetic and phenotypic diversity in SSA, compared to Asia and Latin America, necessitating the acquisition of continent-specific knowledge on their bionomic features [13]. Taken together with context-specific knowledge of resistance patterns to larvicides and adulticides, this is fundamental for the development of VC strategies [14–16].

Evidence on the effectiveness of control methods against Aedes mosquitoes in SSA remains scarce. Systematic reviews of Aedes mosquito control interventions have largely focused on studies from Latin America and Asia [17–19], with little to no data on VC tools used in the African context. A recent systematic review on household-level interventions against Aedes includes examples from three countries in SSA [20] but, overall, this region is underrepresented in the literature compared to research on Anopheles mosquitoes, where SSA has historically played a significant role.

As a comprehensive synthesis of what has been conducted so far in SSA is lacking, and taking into account the relevance of adapting VC tools to local conditions, our review aims to assess evidence on the effectiveness of interventions targeting Ae. aegypti and Ae. albopictus in field settings in SSA. We also seek to understand how these have been deployed and evaluated, with the goal of informing future programmatic efforts for vector-oriented preventive control.

Methods

Search strategy and inclusion criteria

We used the SPICE framework to frame the study question, defining the study’s setting, population/perspective, intervention, comparison and evaluation criteria (Table 1) [21].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of the key elements of the scoping review question following the SPICE framework.

https://doi.org/10.1371/journal.pntd.0013203.t001

The search strategy included keywords identified based on (i) the specific mosquito populations targeted by the interventions, Ae. aegypti and Ae. albopictus, including alternative names, and (ii) the geographical context of the interventions, i.e., the SSA countries listed by the WHO Regional Office for Africa (S1 Table). Boolean operators were used to combine the keywords appropriately. A decision was made for a broad and sensitive search strategy without restricting the search by study design in an attempt to include the different VC study designs for which we anticipated a lack of standardized and consistent terms for a literature search. The search was conducted across PubMed and ISI Web of Knowledge. Additionally, a snowballing technique was applied, reviewing reference lists of relevant articles and systematic reviews to identify any further studies that met the inclusion criteria, but were not captured in the initial systematic search. Google Scholar was consulted for the first 100 records and compared to the systematic search as validity check.

As this review focuses on studies evaluating the effectiveness of VC tools in the field, we selected records that (i) were published from 01/01/2000 onward, (ii) were primary study reports or review articles, (iii) focused on evaluating one or more Ae. aegypti and/or Ae. albopictus VC intervention(s), (iv) described an intervention implemented in the field in SSA, and (v) provided a measure of entomological and/or epidemiological outcomes. Studies exclusively reporting on laboratory experiments, proofs of concept, or those focused solely on community acceptability were excluded. For the purposes of this scoping review, opinion articles and commentaries were also excluded. No a priori restrictions were applied concerning language or study design. A checklist with exclusion criteria was used to assist reviewers in screening and selecting relevant papers (S2 Table). This checklist resulted from calibration based on the screening of the first 20 records retrieved from the initial literature search.

Results

Selection and key characteristics of sources of evidence

The original search identified 3924 records. After removing duplicates, 2707 records were screened by title and abstract, of which 33 remained for full-text screening and eight studies were finally included for data extraction and analysis (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flowchart of the scoping review.

https://doi.org/10.1371/journal.pntd.0013203.g001

Four studies originated from West Africa, with three from Burkina Faso and one from Côte d’Ivoire. The remaining four studies were conducted in Ethiopia, Gabon, Kenya, and Sudan. As two of the records from Burkina Faso reported on two consecutive analyses of the same intervention, this review comprises 8 studies evaluating 7 interventions. The earliest paper was published in 2005, while the other seven were published between 2016 and 2022.

VC intervention evaluations were as follows: three cluster-randomized controlled trials (RCT), three before-and-after studies, one retrospective cohort analysis, and one case-study. Two of the three before-and-after studies were carried out during outbreaks, specifically during a dengue outbreak in Sudan in 2010 and a chikungunya outbreak in Ethiopia in 2019. The remaining (6/8) referred to studies realized outside of Aedes-borne disease outbreaks.

Most of the interventions (5/7) focused on non-chemical methods targeting the larval or pupal stages of mosquitoes, followed by chemical methods targeting adult stages (4/7) and chemical methods targeting immature stages (2/7). There were no studies evaluating non-chemical methods targeting adult Aedes mosquitoes, such as mass trapping or Wolbachia. Community participation was included in four of the seven interventions. Integrated VC strategies, combining both chemical and non-chemical VC measures, were employed in the two studies performed during outbreaks. Entomological endpoints were assessed in seven studies, and epidemiological outcomes in only two. Two studies including community participation, assessed also knowledge and behavior, with one study integrating both qualitative and quantitative methods (Table 2). The key characteristics of the included studies are presented in Table 2 and more extensively in the Supplementary material (S3 Table), detailing study objectives, methodology, interventions, indicators and main findings.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Key characteristics of the included studies.

https://doi.org/10.1371/journal.pntd.0013203.t002

Results of individual sources of evidence

Assessment of the quality of evidence for VC effectiveness evaluation

Based on the paper by Wilson et al [22] that elaborates on the essence of a good study design for evaluating the effectiveness of VC interventions, a critical assessment on the quality of evidence of the included papers was performed (Table 3). Specifically, this focused on study design, implementation and adherence to the intervention, epidemiological and entomological outcomes. None of the studies scored blue/good on all aspects evaluated.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Quality of evidence assessment.

https://doi.org/10.1371/journal.pntd.0013203.t003

Only three studies were cluster-RCTs, among them one only included one intervention and one control cluster. All but two studies employed an intervention allocated to clusters and in four cases the intervention was randomly allocated to an intervention group and compared to a contemporaneous control group. In six studies, randomization was also applied for the site selection for entomological indicators monitoring. Sample size calculation was specified in six studies. In five studies the duration of follow up was considered appropriate being at least one transmission season or a one-year period if the transmission is perennial. In only three studies, the entomological sampling was performed more than once during the post-intervention phase. Budget constraints were mentioned in one paper as a limiting factor to have an adequate follow-up period. Blinding was not possible as expected, with the exception of the only study with a qualitative component in which interviewers were blinded to adopter status.

All but three of the papers put in place quality controls of implementation through supervision, random checking, and evaluation of proper application of insecticides. The same studies also assessed intervention adherence through self-reported assessment combined or not with observations, the use of qualitative methods, and use of communication techniques for community motivation and engagement in VC activities.

Regarding the outcomes, only two studies considered epidemiological indicators related to disease incidence and/or infection seroprevalence, providing direct measures of VC interventions impact on human health. In contrast, entomological indicators were consistently used across all studies, offering insights into mosquito population dynamics. Trap-based sampling, a method frequently employed to monitor adult vector density, was used in only one of the three studies assessing adult Aedes mosquitoes. This study employed both indoor and outdoor light traps to collect adult mosquitoes. Beyond traditional larval indices, two studies looked at the pupal/person index for immature Aedes abundance. Only one study made use of an entomological indicator based on human biomarkers for exposure to Ae. aegypti mosquito bite.

Discussion

Given the increasing global importance of Aedes-transmitted diseases, this study aimed to review the literature on the effectiveness of implemented VC methods targeting Aedes mosquitoes in SSA. We found only eight studies published since 2000, using various study designs, vector control strategies, and impact estimates. The heterogeneity of included studies limits the strength of conclusions that can be drawn for general recommendation of interventions.

Among the eight studies, three came from Burkina Faso, identified in the WHO Dengue Global Reports as the SSA country most affected by dengue [1,3]. The earliest study, published in 2005, evaluated a chemical intervention carried out in 1997 in Abidjan, Côte d’Ivoire [30]. Arboviruses represent a public health threat in several SSA countries, as evidenced by a major dengue outbreak in 2017 in Côte d’Ivoire [31], and in 2010 in Port Sudan with 3765 cases [23], and by a chikungunya outbreak in 2019 in Dire Dawa, Ethiopia [24], causing 41162 suspected and 16 confirmed cases [32]. Co-circulation of arboviruses has been described, causing simultaneous outbreaks as for example in Gabon [33], and subsequent outbreaks of dengue, chikungunya, Rift Valley fever, YF, and Crimean–Congo hemorrhagic fever in Kenya [34]. If model predictions are correct and the dengue burden in Africa is approximately equivalent to that in the Americas [9], this implies an extremely important gap in knowledge on arbovirus circulation and the huge need for VC effectiveness estimates in African countries. Most of the studies (5/8) were published from 2018 onwards, limiting our ability to assess trends over time.

The VC methods evaluated were varied, including both chemical and non-chemical approaches targeting immature stages, as well as chemical methods for adult Aedes. The majority of included interventions (4/7) investigated environmental management strategies including community-based interventions. Environmental management activities varied both in terms of the activities performed and in terms of places and people involved: households versus schools or adults versus children. The community was empowered through education but played a varying role. Often the community was a recipient of the VC intervention, and rarely involved in the intervention’s decision-making pathway. In a previous review by Bowman et al. on dengue VC effectiveness, interventions based on community participation performed significantly better in reducing vector larval development sites [17]. The strongest evidence comes from Cuba where intervention promoted by community groups for environmental management contributed not only to reducing vector indices, but also impacted dengue transmission [35]. More recent studies from Latin America and the Caribbean region also showed that environmental management, especially within integrated strategies, can lead to long-term beneficial effects when community participation goes beyond mere health education campaigns [36]. Conversely, from the examples included in our review, although community-based interventions showed a beneficial increase in knowledge and self-reported beneficial behaviors, they did not lead to a corresponding shift in practical behavior and/or impact on entomological indices. Indeed, in Kenya, the improvement in self-reported behavior against the Aedes mosquito did not match with the behaviors observed by the study teams and the community-based intervention did not show an impact on entomological indices [27]. Likewise in Burkina Faso, the community-based intervention had no effect on the absolute number of Ae. aegypti breeding sites [25]. Additionally, the integrated VC strategy implemented in Sudan in 2010 highlighted the challenges of maintaining household adherence to regular cleaning activities. It underscored the need for continuous health education programs to effectively promote and sustain community participation [23]. In this regard, it is important to keep in mind the complex nature of community-based interventions [37]. The qualitative approach that was presented in only one of the studies may be beneficial to explain how such interventions succeed or fail and what implementation processes are conducive.

It is worth noting that, in our review, community involvement focused only on the control of immature stages of Aedes, missing the opportunity to target interventions against adult mosquitoes that might enhance VC effectiveness by directly impacting the reduction of virus transmission through the reduction of adult mosquito density and longevity [38]. This has also been noted by Montenegro-Quiñonez et al. in their review on Aedes VC at the household level, in which they similarly identified only a minor role of the community in interventions against adult mosquitoes [20]. The involvement of the communities could enhance the acceptability and uptake of a VC activity against adult Aedes mosquitoes with promising effects [36], for example mass trapping, and increase the sustainability of an integrated Aedes VC strategy [39].

As emerged from our review, the same intervention can result in different impacts depending on the context of its implementation, be it in urban, rural, maritime or forest ecosystems. Behavioral differences in the Aedes mosquito population have indeed emerged. While in the Ethiopian city of Dire Dawa Aedes adult mosquitoes were prevalent in outdoor open empty barrels, in rural Burkina Faso they were mainly caught indoors. In Côte d’Ivoire, overall exophagic behavior prevailed (56%), but in the specific maritime area, mosquitoes showed more clearly endophagic behavior. Although Aedes species are generally considered highly anthropophilic, feeding and resting primarily inside human dwellings [40], a more diverse range of behaviors seem indeed to be present in SSA. Activity outside dwellings by Ae. aegypti has been reported from West Africa, as in Ghana [41], Senegal [40], and from East and Central Africa, as in the case of the Democratic Republic of the Congo (DRC) [42,43] and Kenya [44]. As previously highlighted by Badolo et al, it is possible that exophagic or endophagic preferences may not be exclusive and coexist in mosquito populations [13]. Preferences for oviposition sites by Ae. aegypti also varied in our included papers with indoor containers most infested in Port Sudan, Sudan, while outdoor containers containing rainwater prevailed in Dire Dawa, Ethiopia, and discarded containers in rural Burkina Faso. Prior knowledge of specific mosquito populations bionomics is indeed essential to optimize VC interventions, as is understanding the specific composition of the local vector population. More specifically, in the context of SSA, the subspecies Ae. aegypti aegypti, known as the domestic ecotype found outside Africa, coexists with Ae. aegypti formosus, which more traditionally occupies forest habitats with zoophagic tendencies and preferences for oviposition in natural rather than artificial sites [45], but which is increasingly being found in urban environments [46]. These divergent behaviours suggest that VC interventions effective in other regions, such as Asia and Latin America, where Aedes mosquitoes are typically endophagic and endophilic, might not be as effective if applied in the same way in SSA. This calls for a reflection on the possible limitations of strategies based on residual indoor spraying.

It is worth noting that none of the publications included in our review identified the presence of Ae. albopictus, despite it being known as an important vector of Aedes-borne diseases in Africa [12,47].

In our review, chemical methods against the adult population of Aedes in SSA (permethrin, deltamethrin and propoxur) alone or in combination with larvicides (temephos) were reported to decrease the entomological indicators. In two studies, one based on a chemical adulticide [30] and the other on a biological larvicide [28], researchers reported the short time effect of the intervention on the Aedes population, with mosquito densities returning to or even going beyond the starting point, i.e., the rebound effect, as described by Horstick et al [48]. None of the included studies reported on insecticide resistance, however, this is of primary importance when planning and evaluating interventions, given the potential negative effect of insecticide resistance on VC effectiveness. After a long period of inadequate monitoring of Aedes resistance to insecticides in SSA, recent studies, particularly from West Africa, are offering new insights on this topic [49]. Overall resistance to DDT has been described as widespread in SSA for both Ae. aegypti and albopictus, while resistance to pyrethroids (mainly permethrin and deltamethrin) appears to be more sporadic [46,50], which are often used for insecticide treated nets and indoor residual spraying. Resistance to pyrethroids have been reported in Burkina Faso [51], Benin [16], Ghana [41], Senegal [15], DRC [52], Capo Verde [53], Cameroon [54] and contrasting resistance patterns from the Republic of the Congo [55]. In Cote d’Ivoire, where the study based on the use of deltamethrin was carried out, field populations of Ae. aegypti were found to be susceptible to deltamethrin in Yopougon but not in Port-Bouët [56]. Conversely, in the Port Sudan area, the other study site included in our review, sensitivity to pyrethroids remained preserved though certainly requiring attention to ensure its future effectiveness [57]. Some resistance to carbamates (usually propoxur) has been described in SSA [46], while biological and chemical larvicides, Bti and temephos, seem to maintain a less problematic level of resistance [49]. Considering insecticide resistance that is occurring in several areas of endemic countries in SSA where many insecticides from the organochlorine, organophosphate and pyrethroid groups are also used for the control strategy of anophelines, besides Aedes [57], testing new sustainable and innovative technologies for vector control is becoming crucial. These include, amongst others, the Wolbachia-based methods, declared by the WHO in 2021 to be an “intervention of public health interest” [58], and based on Wolbachia’s ability to manipulate reproduction and reduce the transmission of vector-borne pathogens [59]. Large-scale use of Wolbachia-infected Ae. aegypti has been deployed in Australia, Asia and the Americas, coinciding with a reduction in the local transmission of dengue. There are currently no pilot studies in SSA [58].

Finally, the integrated VC interventions presented in the included studies were evaluated only during outbreaks, so evidence on the effectiveness of integrated VC strategies against Aedes in non-outbreak settings in SSA is lacking. Ideally vector control programs should be integrated, have a pre-rainy season prevention activities, and a strategy for rapid response to disease outbreaks [13,60].

The critical appraisal of the quality of the VC effectiveness in the included studies was based on evidence-based VC considerations reported by Wilson et al. [22]. This assessment aimed to highlight the main strengths and shortcomings of the included articles and provide insights for improving the design and implementation of VC studies. Only three included studies adopted the cluster-RCT approach [25,27,28]. This methodology lends itself to the highest methodological quality in VC evaluation, as opposed to before-and-after studies, but as shown in Table 3, these RCTs had several shortcomings, which limits the generalizability for findings. In four papers [23,24,29,30], the lack of a control group hindered the assessment of impact. This limitation made it difficult to account for seasonal changes, such as variations in precipitation, which could either mask or exaggerate the observed effect [22]. Apart from blinding, which is notoriously difficult in VC studies, the other critical points in most studies were the post-intervention evaluation period, entomological sampling, and the type of outcome used for evaluation. We considered as appropriate a follow-up period of at least one transmission season or a 1-year period if transmission is perennial, indicated as the minimum follow-up duration by Wilson et al. Whether using chemical methods or targeting behavioral change in community-based-type interventions, follow-up periods with repeated measurements are needed to detect sustained behavioral change or maintained intervention effects. In relation to outcomes, almost all studies included entomological indicators. It is worth mentioning that these studies did not rely exclusively on the traditional Stegomyia indices only (HI, CI, BI), but in most cases included indices of adult Aedes abundance, the pupal indices, human biomarkers for exposure to Ae. aegypti mosquitoes. Stegomyia indices, which assess the immature mosquito stage (larvae and pupae), were developed over 90 years ago for YF, and their reliability and sensitivity as indicators of dengue epidemic risk have been widely questioned [17,61]. As reported by Wilson et al, measuring pupae and/or adult vector density is more appropriate for assessing transmission risk [22]. From a public health perspective, however, epidemiological outcomes are needed to demonstrate the effectiveness of interventions in protecting human populations [22], which were lacking in most of the included studies. Only the study performed during the dengue epidemic in Sudan examined both entomological and epidemiological indices, where authors identified a significant relationship between the reduction in dengue incidence and entomological outcomes. It should be noted, however, that in both outbreak studies [23,24], the traditional entomological indices, although reduced post-intervention, still indicated a high probability of arbovirus transmission. Providing evidence on effectiveness should include not only entomological indices [61], but also epidemiological information on disease occurrence, based on incidence data, or even better sero-prevalence seen the importance of asymptomatic infections.

Strengths and limitations

One of the strengths of the present study was the use of a broad search string, which was particularly advantageous given the current unavailability of MESH terms specifically related to the VC field. This approach enabled broad coverage in our scoping review. At the same time, the paucity of the included studies related to SSA published since 2000, not meeting the quality standards, as well as the heterogeneity of the results are among the limitations of our study. None of the systematic and scoping reviews cited in our discussion, except the one by Montenegro-Quiñonez et al [20], considered VC practices against Aedes implemented in SSA, a region that remains underrepresented in the literature.

Conclusion

In conclusion, there are only a limited number of studies evaluating the effectiveness of VC strategies against Aedes in SSA compared to what is published in Latin America and Asia. The paucity and heterogeneity in the study designs and evaluation procedures of the included studies, hinder the ability to draw unequivocal conclusions on Aedes control methods in SSA, revealing a critical research gap. Chemical interventions against adult or immature stages of Aedes (permethrin, deltamethrin, temephos), either alone or as part of an integrated VC strategy, seemed to have a positive, albeit temporary, impact on entomological indicators, however, no evidence was reported on epidemiological outcomes. Additionally, environmental community-based management interventions have been prominent in the studies and warrant further investigation on how to tailor Aedes VC strategies to local vector bioecology, but also to community and individual behaviors.

Supporting information