Social and behavioural considerations for responding to Anopheles stephensi in Africa: a scoping review

Abstract

Background

Anopheles stephensi has been identified in multiple African countries over the past decade. Unlike malaria vectors endemic to Africa, invasive An. stephensi shares characteristics with Aedes aegypti, including laying eggs in artificial containers near human dwellings, particularly in urban areas. These characteristics pose a threat to malaria control and elimination and shape needed interventions, including the role of individuals, households, and communities. Due to the urgency to develop strategies for the African context, this scoping review sought to identify social and behaviour change (SBC) considerations for responding to An. stephensi in Africa, by synthesizing information currently available, leveraging lessons from Anopheles and Aedes control, to inform evidence-based guidance.

Methods

The review included core interventions promoted for malaria (insecticide-treated nets, indoor residual spraying, and care-seeking for fever) and larval source management (LSM) interventions. A systematic search of four online databases (PubMed; EMBASE; Global Health; and Global index medicus) used intervention-specific terms defined a priori. Peer-reviewed articles published from 2000 to 2021, with a description of individual, household, or community behaviour and/or participation in one or more core or LSM interventions were included. For LSM interventions, included articles related to control of malaria and Aedes-borne infections.

Results

Of 3,306 articles screened, 42 met inclusion criteria. Twenty-seven were malaria control studies with two on An. stephensi. Fourteen publications targeted Ae. aegypti and one did not specify vectors. In low transmission contexts, malaria-related prevention behaviours tend to be practiced at lower rates, underscoring the need to address the factors driving these behaviours and effectively tailor SBC to the needs of higher risk groups. It is also critical to create early and meaningful partnerships with affected communities. Finally, highly specific guidance for governments, communities, and households is paramount, particularly for LSM interventions, as missing any specific step in the necessary behaviours for these interventions will reduce their potential effectiveness.

Discussion

The dearth of An. stephensi-specific articles underscores the need for further research and documentation as countries respond to the threat. This review highlights the importance of tailoring SBC to groups experiencing higher risk, like those moving between higher and lower transmission areas or working in construction sites. Learnings from Aedes control provides valuable insights into LSM for An. stephensi and opportunities for integrated approaches to SBC, including multi-sectoral collaboration, partnership with community leaders and civil society.

Background

Anopheles stephensi, a dominant malaria vector in South Asia and the Arabian Peninsula, has been confirmed in Africa in recent years, and has the potential to threaten malaria control and elimination efforts on the continent. First identified in Djibouti in 2012, An. stephensi has since been identified in sites across Ethiopia, Sudan, Somalia, Nigeria [1] and most recently in Kenya [2] and Ghana [3]. Similar to primary malaria vectors in Africa, such as Anopheles gambiae and Anopheles arabiensis, An. stephensi may use natural habitats with stagnant water to lay eggs. However, unlike such endemic vectors, An. stephensi also shares breeding habitat similarities with Aedes aegypti, a vector of diseases such as dengue, Zika, chikungunya, and yellow fever. These similarities include a preference for using artificial larval habitats, such as plastic tanks, containers, abandoned tires, and cisterns near human dwellings.

Because of its ability to thrive in both rural and urban environments, An. stephensi poses new challenges in urban areas, and areas of rapid urbanization, in Africa. Sub-optimal urban infrastructure and insufficient provision of piped water means that many urban households depend on containerized and stored water for home use. Additionally, urban expansion requires an abundance of stored water for brick curing, building, and construction, creating numerous highly favourable larval habitats for An. stephensi [4]; and for locations where An. stephensi is already present, there is a higher risk to construction workers working in close proximity to these stored water sources. Displaced populations living in temporary shelters in urban areas and using open water containers where An. stephensi can proliferate may also experience greater risk of exposure to this vector [5]. This threat is compounded by the rapid growth of many African cities, with over 40% of the African population now living in urban settings [6]. Modeling studies have predicted potential consequences of the spread of this vector in Africa. One model estimates that an additional 126 million people in Africa could be at risk of malaria if the vector spreads unabatedly [7]. Another model predicted that the annual number of Plasmodium falciparum malaria cases in Ethiopia alone could increase by 50% if no additional measures were taken to control this vector [8].

In response to the increased number of countries detecting An. stephensi and given the threat and potential impact to the malaria landscape, the World Health Organization (WHO) launched an initiative to “Stop the spread of Anopheles stephensi in Africa” in September 2022, and updated it in 2023 [6]. The initiative highlights five priority areas including: increasing collaboration, strengthening surveillance, improving information exchange, developing guidance and prioritizing research. Better understanding of the behaviour and ecology of An. stephensi can help to guide selection of vector control interventions to address it. A vector alert released by WHO in 2023, underscores the importance of enhanced entomological surveillance to detect this mosquito, investigations of vector bionomics to understand vector behaviour across different settings, as well as insecticide-resistance testing to help guide appropriate vector control interventions [1]. The vector alert highlights, in areas where invasive An. stephensi has been identified, that the immediate focus should be on managing vector larval sites in urban and peri-urban areas. This includes removing larval sites where feasible, modifying larval sites to prevent mosquito access, and where removal and modification are not feasible, treatment with biological or chemical larvicides [1]. Where appropriate, introduction or reinforcement of insecticide-treated net (ITN) access and use, indoor residual spraying (IRS), and prompt care-seeking for fever can be considered. Malaria cases should also be monitored, particularly in urban and peri-urban areas, as increases in cases could indicate An. stephensi presence [1].

In addition, the 2023 WHO vector alert highlights the importance of social and behaviour change (SBC) activities to the success of the An. stephensi response in Africa. To date, SBC has been used to understand and influence a range of malaria-related behaviours [9, 10] such as proper use and care of ITNs [11–14], as well as prompt and appropriate care-seeking [15]. Studies assessing the impact of SBC on entomological indicators have also shown the effectiveness of household-level behaviours in the context of Ae. aegypti. For example, effectively covering water storage containers has contributed to reductions in larval count in containers [16–20] and removal of standing water along with education campaigns has been associated with lower vector densities [21, 22].

To address the threat of An. stephensi, it is imperative to act early. This means countries where An. stephensi has been identified, and all others on the continent that are at risk of invasion, should include SBC in their response. While the importance of integrating SBC in the An. stephensi response has been established [1], affected countries will need more detailed, evidence-based guidance to integrate SBC activities across planned interventions. Due to the urgency to develop strategies for the African context, this scoping review sought to synthesize information currently available on social and behavioural considerations for interventions likely to be included in the response to An. stephensi in Africa and whose implementation relies on individual, household, and/or community level behaviours. In light of limited literature on the role of human behaviour on An. stephensi and that An. stephensi’s bionomics share many similarities with the non-malaria vector Ae. aegypti, lessons from Aedes control were also leveraged to inform evidence-based guidance for SBC responses to An. stephensi in Africa.

This review of evidence informed the development of data-driven recommendations and guidance for individual, household, and community-level behaviours that have the potential to support the mitigation and control of An. stephensi in Africa as part of integrated strategies to control malaria and other vector borne diseases. The full guidance, Social and Behaviour change guidance for Anopheles stephensi in Africa, is available online [23]; this article presents the results of the scoping review.

Methods

Search strategy

A systematic search of four online databases (PubMed; EMBASE; Global Health; and Global index medicus) was conducted using intervention-specific search terms defined a priori with guidance from a Johns Hopkins University Library informationist. Each intervention’s search terms were developed based on their relevance to the immediate response for addressing An. stephensi in sub-Saharan Africa as well as the level and type of additional information needed to inform the SBC guidance. These intervention-specific search terms are listed below in Table 1. Following the systematic database search, a manual search of references was performed on articles that met the eligibility criteria to identify additional articles included in the final synthesis of the findings.

Table 1.

Intervention-specific search terms

Intervention	Search terms
Core malaria interventions
Use of insecticide-treated nets (ITNs)	("Insecticide-Treated Bednets"[Mesh] OR "Insecticide-Treated Bednets" [tw] OR "Treated Mosquito Nets" OR "Treated Mosquito Nets" [tw] OR "ITNs") AND ((urban) OR (city)) Filter: 2000–2021
Indoor residual spraying (IRS)	((indoor residual spraying) AND ((urban) OR (city))) Filter: 2000–2021
Prompt care seeking for fever	(Care-seeking) OR (sought treatment) OR (health-seeking) OR (treatment seeking) AND (Fever OR Malaria) AND ((urban) OR (city)) Filter: 2000–2021
Larval source management (LSM)—full review interventions
Household larviciding	((((An stephensi) OR (Anopheles stephensi)) OR (Aedes aegypti)) AND (larvicid*)) AND (water container) Filter: 2000–2021
Community larviciding	((Anopheles [MESH] OR anopheles [tw]) AND (larvicide OR larvicid*)) Filter: 2000–2021
LSM—interventions included in previous Zika review
Finding and removing standing water regularly	((((((((An stephensi) OR (Anopheles stephensi)) OR (Aedes aegypti))) AND (breeding sites OR stagnant water OR standing water)))) AND (remov* OR fill OR drain OR find)) Filter: 2016–2021
Covering water storage containers	((Aedes aegypti) AND (lid OR lids) OR (cover)) Filter: 2016–2021 (((An stephensi) OR (Anopheles stephensi)) AND ((lid OR lids) OR (cover)) Filter: 2000–2021

Interventions included in this review were identified through a preliminary desk review and consultations with the U.S. President’s Malaria Initiative (PMI). An intervention was included if it (1) has the potential to be effective against An. stephensi based on published evidence of the intervention’s efficacy against An. stephensi or other mosquito vectors with similar bionomics; (2) is likely to be included in the immediate An. stephensi response in areas where An. stephensi has been identified and/or those at elevated risk for invasion in Africa; and (3) has an individual, household, and/or community component.

The prioritized interventions that met these criteria were then categorized into three groups that underwent tailored search strategies based on their relevance to the immediate response for addressing An. stephensi in sub-Saharan Africa and the level and type of additional information needed to inform the SBC guidance. The first group of priority interventions was made up of three core malaria interventions that are already widely used in the African context and have the potential to be effective against malaria transmitted by An. stephensi. This included ITNs, IRS, and prompt care-seeking for fever. However, there is already substantial evidence to support the effectiveness of these interventions, and comprehensive SBC guidance available to promote associated individual, household, and community-level behaviours in rural areas. Therefore, the present review focused on unique considerations for addressing An. stephensi in urban contexts. The unique threat of An. stephensi in urban areas led to a targeted search that focused on identifying SBC considerations for these interventions in urban areas.

The second group of priority interventions consisted of larval source management (LSM) interventions that were likely to be used to address An. stephensi in Africa. This group underwent a full review, described below. The category included the regular application of larvicide, by vector control technicians or trained household members, to larval sources. This intervention was separated into two different interventions, one at the household level and another at the community level, given that there were likely to be different SBC considerations for the application and acceptance of larvicide in household water containers compared to community larval sources.

An additional set of LSM interventions included in this review were two interventions that recently underwent a literature review for inclusion in the SBC response to Zika in Latin America: regularly finding and removing standing water in and around the home and in communal areas, and covering infrequently accessed water storage containers with effective lids to prevent mosquitoes from entering [24]. Given the strong overlap between the larval habitats of Ae. aegypti and An. stephensi, these interventions underwent a targeted review that focused on identifying updates to the literature over the past 5 years since the original review [24].

Eligibility criteria

Following the search in the above databases, the articles were screened for inclusion based on the eligibility criteria shown in Table 2. For LSM interventions that underwent a full review (see Table 1 for search specifics) both urban and rural settings, as well as relevant vectors, Aedes and Anopheles were included in the search. To minimize outdated information, only studies published from 2000–2021 were included for the full review interventions and the core malaria interventions (ITNs, IRS, and prompt care-seeking), while the interventions that were included in the previously mentioned Zika SBC review [24] were restricted to the period since the previous review was done (2016–2021). One exception to this timeframe was the additional review carried out to identify An. stephensi-specific literature from 2000–2021 on covering water storage containers. Publications were also limited to the English language.

Table 2.

Inclusion and exclusion criteria

Criterion	Inclusion	Exclusion
All interventions
Language	Published in English	Not published in English
Study focus	1. Included a description of the intervention of interest AND 2. Included a description of community, household, or individual behaviour and/or participation in the intervention	Does not include intervention of interest or a description of community, household, or individual behaviour and/or participation in the intervention
Availability	Full text is available	Full text was not available
Core malaria interventions
Date published	Published between 2000–2021	Published before 2000
Data collection setting	Promotion in urban settings as defined by the author of the study	Any setting that is rural or not classified as urban
LSM—full review interventions
Date published	Published between 2000–2021	Published before 2000
LSM—interventions included in previous Zika review
Date published	Published between 2016–2021 to update findings from the previous literature review*	Published before 2016
*For covering water storage containers an additional search from 2000–2021 was carried out for An. stephensi specific literature

Study selection

All search results were imported into Endnote version 20.4. Following the removal of duplicate articles, study titles that clearly did not meet all inclusion criteria were excluded, the remaining articles were further screened by abstract and full text review. Any discordance during the selection process was discussed by reviewers and resolved through consensus.

Synthesis of results

Key characteristics of publications that met all inclusion criteria were summarized including the intervention, study design, study setting, disease focus (i.e., dengue, Zika, or malaria), target vector (i.e., Anopheles or Aedes), and key SBC findings, conclusions, and/or recommendations. Following the summary of selected publications, a thematic analysis was performed to identify key themes and findings for each of the interventions as well as cross-cutting SBC considerations for An. stephensi control in Africa.

Results

A total of 3306 articles were identified from the search. Following title, abstract, and full text screening a total of 32 publications met the inclusion criteria. Ten additional articles were identified for inclusion through manual reference searching, for a total of 42 articles. The article selection process is aligned with the PRISMA guidelines [25] and is shown in Fig. 1. Common reasons for record exclusion following the abstract screening included the study not describing any individual, household, or community behaviour and/or participation in the intervention (n = 912), the study not describing the intervention of interest as defined in this review (n = 632), the study was a duplicate article that failed to be excluded previously using Endnote software or during the title screening (n = 47), was not a full text peer-reviewed article, such as conference abstracts (n = 20), or the study was not published in English (n = 1). An additional 41 articles were excluded following the full text review for providing no relevant SBC guidance.

Fig. 1.

Process of identification, screening, and inclusion of articles for the scoping review according to PRISMA guidelines

Of the 42 articles identified, a majority [23] focused on countries in Africa, followed by eight publications in Latin American countries, and seven in Asian countries. The remaining four were literature reviews that encompassed more than one region. Eighteen articles were specific to urban areas only, while 14 were restricted to rural areas. The remaining 10 articles included both urban and rural areas in their study sites.

Table 3 summarizes the study designs, target vector, and frequency of each intervention included in the review. The single intervention with the most articles, at 11 publications, was community larviciding. Prompt care-seeking for fever was the focus of four publications, while ITNs, household larviciding, and covering water storage containers had three publications each. IRS and finding and removing standing water had the fewest with two publications each. The remaining articles (n = 14) covered more than one of the interventions in this review, typically multiple LSM interventions. Only two of the 27 malaria control studies were specifically focused on An. stephensi as the disease vector, while fourteen publications, covering dengue or Zika, targeted Ae. aegypti. One article did not specify a target vector because the disease outcome of interest was general febrile illness [26].

Table 3.

Summary of study designs, target vectors and countries

Intervention	Study design				Target vector				Countries
	Quantitative	Qualitative	Mixed Methods	Literature Review	Anopheles stephensi	Other Anopheles	Aedes aegypti	N/A
ITNs 3 articles	[28]	[29, 30]				[28–30]			Nigeria, Uganda, Tanzania
IRS 2 articles		[31, 32]				[31, 32]			Mozambique, Tanzania
Prompt care seeking for fever 4 articles	[26, 33–35]				[34]			[26, 33, 35]	Bangladesh, India, Sudan
Household larviciding 3 articles	[36]			[37, 38]		[37, 38]	[36]		Argentina, Multi-country review
Community larviciding 13 articles	[39–44]	[45–48]	[49–51]			[39–47, 49–51]	[48]		Kenya, Tanzania, Ethiopia, Burkina Faso, Malawi, Haiti, Rwanda
Finding and removing standing water 6 articles	[52]	[53, 54]	[55–57]				[52–57]		Honduras, Pakistan, Colombia, Vietnam, Mexico, Ecuador, Brazil, Uruguay, El Salvador
Covering water storage containers 3 articles	[58, 59]	[60]			[59]		[58–60]		Guatemala, Honduras, El Salvador, India, Brazil
Multiple interventions 13 articles	[17, 61–68]	[69, 70]	[71]	[24, 27]		[62, 63, 65, 67, 70]	[17, 24, 27, 61, 64, 66, 68, 69, 71]		Laos, Thailand, Nicaragua, Mexico, Tanzania, Kenya, Malawi, Brazil, Taiwan, Guatemala, Cuba, Multi-country review

A quantitative design was used in 23 studies, 10 studies implemented a qualitative-only design, while five studies used a mixed-methods design combining the use of cross-sectional surveys, observational checklists, focus group discussions (FGDs), and in-depth interviews (IDIs). The remaining four studies were literature reviews including one systematic review on the effectiveness of biological, chemical, and educational dengue fever prevention programmes on entomological indicators [27].

The results below present relevant SBC findings and recommendations for more effective promotion, uptake, and use for each of the interventions included in the review. A summary of key findings from the scoping review as well as associated SBC recommendations generated by the authors based on review findings and SBC best practices, are included in Table 4.

Table 4.

Key findings from the literature by intervention and associated SBC recommendations developed by the authors of this review

Intervention	Key findings	Recommendations
Cross-cutting	• Malaria-related behaviours tend to be lower in urban areas and may require different types of approaches to address the factors that influence those behaviours	• Increase risk perception of malaria in urban areas as well as knowledge of malaria prevention, mosquito breeding sites prevention, malaria symptoms and treatment • When new or less familiar malaria interventions are introduced with a population, communicate the rationale for the change and the importance of continuing use of all core malaria interventions • Emphasize response efficacy (confidence that interventions work to prevent malaria, reduce mosquito breeding sites, or treat malaria) of all interventions • Identify and focus SBC with higher risk and vulnerable segments of urban populations, such as low socioeconomic status families, mobile and displaced populations, and construction workers
ITNs	• ITN usage is less consistent in mobile populations such as fishing communities, pastoralists, refugees, and immigrants [29] as well as among pregnant women in urban areas compared to pregnant women rural areas [28] • Low-income areas are more likely to rely on ITNs for malaria protection while wealthier areas are more likely to use other prevention measures (e.g., housing modification, insecticide spray and long-sleeved clothing) [70] • Factors for long-term habitual use include previous positive experience of ITN use, perceived community social support, and appreciation or awareness of other benefits from ITN use [29]	• Ensure access to ITNs through push and pull mechanisms, including urban areas where ITN use may not be a community norm • Promote benefits of ITN use and care beyond malaria-prevention [e.g., sleeping more comfortably, protection against other insects) • Promote ITN use and care as a community norm, including in urban areas
IRS	• In urban areas, IRS refusal is more likely among those with more education, citing they did not perceive sufficient impact in terms of reduced mosquitos or malaria [32] • Communities (including urban areas) that had experienced more rounds of IRS had fewer doubts and rumors [31]	• Emphasize the importance of high IRS coverage for promoting both familial and community health Ensure clear and consistent communication with community leaders particularly during spray operator selection and spray implementation. This is particularly important in urban settings or others where IRS may not be a norm
Prompt care seeking for fever	• Reasons for delaying seeking care for fever include lack of time, self-medication, expectations that the fever would resolve itself, and use of a home remedy [33] • Migrant workers in urban areas (i.e., construction sites) may not be aware that malaria testing and treatment is available in public health facilities, often at no cost [34] • Caregivers are more than twice as likely to consult a health care provider for what they perceive as a serious fever versus a mild one [35]	• Messages should emphasize that fever is a serious symptom, especially in young children, and that early management is best even if the fever is perceived to be mild • Engage high risk and vulnerable populations in urban areas about how, when, where and why to seek care for fever promptly
Household larviciding	• Household members have access to larval habitats within property that conventional vector control technicians may not	• If provided with larval source management tools, household members can be key to overcoming application and coverage barriers
Community larviciding	• Community education on mosquito biology, awareness, and involvement in larval source management have been key factors in the success of larval source management programmes	• Engaging with community members in multiple ways, including education and approach awareness through to implementation is essential to programme success
Finding and removing standing water regularly	• Interventions at the household level can better empower individuals to take action in and around their home [69], especially if they understand why removing standing water on a weekly basis reduces mosquitoes [24] • This intervention is more effective when communities and households are provided with specific instructions to focus their search on the most prolific breeding sites in their context so as to not dilute their efforts [24] • Community level interventions often involve significant municipal and community coordination and cooperation [69], though successful engagements can result in increased efficacy [68]	• Inform those household members responsible for water collection and storage as they are most appropriate to learn about, find, remove, and/or treat containers • Provide households with information on the mosquito life cycle and demonstrate how to identify larvae and standing water that might breed mosquitoes. Conduct active mosquito searches in homes and communities • Promote very specific actions for how to remove standing water tailored to the most important breeding sites in the location. For example, indicate whether a specific type of container should be emptied, turned over, covered, or discarded • Seek collaboration between health and other sectors
Covering water storage containers	• To be effective, lids should be properly used, maintained, and tight fitting [17] • High levels of efficacy are observed for covering long-term water storage containers but not for frequently used containers due to misuse and wear of the lids [24, 58] • For large tanks, repair and proper lid fitting may be done by professionals [60] • Incorrectly covered containers can favor production of mosquito larvae [19, 24]	• Provide highly specific guidance for the type of lid and/or water storage container to ensure the intervention is implemented effectively and does not create new breeding sites • Ensure access to effective lids where this behaviour is promoted and encourage maintaining lids in good condition • Where appropriate, train professionals who maintain large tanks in mosquito proof covering

Complete and more detailed recommendations can be found in Social and Behaviour Change Guidance for Anopheles stephensi in Africa [23], developed in part from the findings of this scoping review

Core malaria interventions

Insecticide-treated nets

Of the four studies that included ITN use in urban areas, all were based in African countries and targeted Anopheles mosquitoes for control of malaria. Three studies focused only on ITNs (in Uganda, Nigeria, and Tanzania), while one qualitative study presented SBC findings on the use and acceptance of LSM interventions as well as ITNs in Tanzania.

In Uganda, a qualitative study found that ITN usage was more consistent among settled communities compared to mobile populations such as fishing communities, pastoralists, refugees, and immigrants [29]. In Nigeria, a retrospective cross-sectional survey of pregnant women in Nigeria found ITN use among pregnant women in urban areas to be 12% lower than in rural areas [28]. A qualitative descriptive study exploring mosquito control perceptions and practices among residents in four study sites across Dar es Salaam, Tanzania, found that respondents in low-income areas relied mostly on ITNs for malaria protection and considered them as being part of the culture, while in wealthier areas other prevention measures, including housing modifications and personal protection such as insecticide spray and long-sleeved clothing, were more frequently reported [70].

Across studies, protection from malaria was the primary driver for the use of ITNs [29, 30, 70], however, other individual and structural factors were also found to influence ITN usage. In Uganda, other prominent factors for long-term ITN use were previous positive experience of ITN use prior to the distribution, perceived community social support, and appreciation or awareness of other health, economic, or social benefits arising from ITN use [29]. Similarly, ITN affordability, effectiveness, convenience of use, and ready availability were cited as reasons ITNs were listed as the first-choice malaria prevention measure in low-income, urban areas of Tanzania [70].

In addition to identifying factors that enhanced uptake and consistent use of ITNs, reasons for lower uptake of ITNs in urban areas were also described. The two qualitative studies in Tanzania identified low perceived response efficacy of ITNs and the inability of ITNs to protect from outdoor and/or nuisance biting; these were frequently mentioned as reasons for respondents in urban areas to rely on other prevention measures besides ITNs [30, 70].

Behaviours mentioned by focus group respondents that contributed to the lower perceived response efficacy of ITNs were shared by two or more people; ineffective usage habits; small ITNs which become un-tucked at night; and ITNs with holes large enough to allow in mosquitoes [30]. Furthermore, the same houses with well-screened windows, with doors that shut properly, and that use insecticidal sprays against mosquitoes, were reported to not use ITNs, suggesting that urban households with these measures felt protected from malaria without the use of ITNs [30]. While the majority of studies reported individual or behavioural factors contributing to variations in ITN use, authors from the Nigeria study, attributed the 12% urban and rural difference in ITN use among pregnant women to the prioritization of rural communities in previous ITN interventions and campaigns that led to greater access and sensitization for ITNs in rural areas compared to urban communities [28].

These findings describing variations in ITN uptake across high-risk groups and factors influencing ITN usage led to various SBC guidance and recommendations for promoting consistent and correct ITN use in urban areas. In Nigeria, the authors highlighted the need for increased outreach to pregnant women in urban areas, such as through messages delivered at antenatal care (ANC) and radio advertisements to remind about ITN use [28]. Conclusions from the qualitative study among male heads of households, female caregivers, and community health workers in urban areas of Uganda suggested that while the protection of malaria remains a powerful motivator, SBC strategies should promote consistent net use throughout the year and emphasize the non-malaria benefits of net use that provide a long-term rationale for consistent use, such as greater comfort during sleep, protection against other insects and debris from the roof, warmth, overall better health, and greater availability of time to work or perform household duties when malaria illness is avoided. Furthermore, the study suggests SBC campaigns could promote replacement options, emphasizing ongoing net care and replacement as a household responsibility, thus reducing dependence on free distributions [29]. Finally, following focus group discussions in urban areas of Tanzania where several sub-optimal ITN usage practices were identified, the study authors suggest the need to increase the frequency of distribution to increase access and minimize the necessity for bed net sharing and enable improved usage practices [30].

Indoor residual spraying

Only two studies were identified from the search that examined IRS acceptance and household behaviours in urban areas. Both were qualitative studies that assessed IRS for malaria prevention in both urban and rural areas in Tanzania [31] and Mozambique [32]. Reported barriers to IRS uptake included logistical or social barriers to removing household possessions, low perceived effectiveness of IRS, concerns about IRS activities being politically motivated, selection and performance of spray operators, or previous negative experience with IRS campaign, and a preference for ITNs [31, 32]. The cross-sectional study in Mozambique also found that refusal of IRS in urban areas was more likely among those who had secondary or higher education levels, finding that they were more skeptical of IRS because they did not perceive sufficient impact in terms of reduced mosquitoes or malaria [32]. Authors from the qualitative study in Tanzania, with urban study sites in Zanzibar which had begun to transition from blanket to targeted IRS, surmised that the refusal of IRS in this setting was likely a result of misunderstanding or lack of understanding of the process and doubts of the long-term effects of the programme [31]. They also noted that in communities (including urban areas) that had experienced more rounds of IRS, doubts and rumours were allayed [31].

To improve IRS uptake, the authors of the Tanzania study recommended a more comprehensive and streamlined information process to better engage and inform community members. They suggested holding general information meetings focused entirely on IRS where SBC materials could be distributed, followed by additional meetings specifically to answer questions among people with concerns or those who needed additional information [31]. Furthermore, results from the cross-sectional study in Mozambique suggested that SBC should emphasize the importance of high IRS coverage for promoting both familial and community health as well as specifying that IRS reduces the mosquitoes that cause malaria as a way to increase perceived efficacy without solely focusing on nuisance mosquitoes. Both studies identified the need for clear and consistent communication with community leaders particularly during spray operator selection and spray implementation [31, 32]. These studies also identified barriers to IRS acceptance and emphasized the need for effective community leadership and engagement from planning through implementation of IRS programming.

Prompt care-seeking for fever

Four studies were identified in the literature review with relevant SBC considerations for care-seeking for fever in urban areas; all were quantitative studies that utilized cross-sectional surveys for data collection. India was the location for two of the studies, of which one examined An. stephensi and malaria care-seeking knowledge among urban migrant construction site workers [34] and the other assessed malaria knowledge among patients seeking care for fever at a New Delhi hospital [33]. The two remaining studies took place in Sudan [35] and Bangladesh [26] and assessed care-seeking behaviours among urban caregivers for children under five with any febrile illness.

Three of these studies documented levels of knowledge related to care-seeking. One study among suspected malaria cases in an outpatient care department of a New Delhi (urban) hospital found low to moderate knowledge of malaria symptoms, transmission, prevention, breeding places of mosquitoes, and diseases transmitted by mosquitoes [33]. A cross-sectional study in an eastern city of Sudan found that female caretakers of children under 5 years of age had adequate knowledge of malaria which authors attributed to malaria being the most common febrile illness in the area [35].

A third study among migrant construction workers who lived in or near the construction site, in the city of Mangaluru, India where An. stephensi is a documented malaria vector, found that female workers, those with a preference for allopathic medicine, and those who reported suffering from malaria within the previous year had significantly greater odds of increased mean malaria knowledge score (cause, symptoms, and when and how to seek care for malaria) [34]. Male construction workers were significantly more likely to prefer care from public health facilities and allopathic health providers, and the latter was also significantly preferred among those WHO reported a history of malaria within 1 year. Given these findings, the authors reference health facilities as an opportunity for health education. However, knowledge that government hospitals offer free malaria diagnosis and treatment was low among the population surveyed. Knowledge of malaria preventive measures, such as ITN use, among these construction site workers, was also found to be unsatisfactory [34].

Reasons for delays in care-seeking for fever by urban residents were noted in the studies and included the preference for home management of fever or seeking medication in a pharmacy or other location first. Among 200 febrile patients WHO sought care in a New Delhi hospital, 83% did not seek care from a qualified health provider until 3 or more days after fever onset. Reasons cited for the delay were no time (36%), used a home remedy (11%), tried self-medication (31%), or expected the fever to resolve itself (17%) [33]. The authors also found low knowledge about malaria (symptoms, transmission, and breeding places) among the study population, and dengue seemed to be a more top-of-mind mosquito-borne disease than malaria [33]. A cross-sectional study in Dhaka, Bangladesh found that seeking care with a trained health provider for a febrile child (without respiratory symptoms) was more likely for those with higher socioeconomic status (OR 5.6), for younger children under age 2 (OR 1.6), male children (OR 1.5), and those with decreased consciousness (OR 5.3). This analysis did not consider timeliness of care-seeking, although the authors note that, while 80% of caregivers sought care in the first 2 days, 41% of them first sought care with an untrained provider. Finally, the authors note an overall preference for trained health care providers in urban Bangladesh, which may influence care-seeking behaviour and apply to other urban areas [26]. The cross-sectional survey in Sudan found that severity of fever was a primary factor in a caregiver's decision to seek care. Caregivers stated they would be more likely to consult a health care provider for what they perceived as a serious fever (91%) versus a mild one (40%), and moreover that they suspected the cause of a child’s illness to be malaria in only 41% of low-intensity fevers versus 79% of high-intensity fevers [35].

Based on these findings, the four studies provided some SBC recommendations, including increasing knowledge among urban dwellers of malaria prevention (specific to the local vector), symptoms, and treatment. When developing programmes and materials to increase malaria care-seeking knowledge in urban areas, recommendations were made around the use of local languages [34] and message content [33, 35]. Specifically, one study suggested that messages should emphasize that fever is a serious symptom, especially in young children, and promote its early management despite the perception that it is mild; that treatment should only be sought from a health facility and nowhere else; knowledge of preventing mosquito breeding sites [33, 35]; and use of ITNs [34]. Another study recommended that there should be a focus on the importance of early care-seeking for fever with qualified health providers (leveraging a potential preference for such providers), compliance with prescribed treatment, as well as information about potential danger signs and how to handle them [35]. A study also identified a need to focus SBC activities with specific, vulnerable, or often overlooked segments of urban populations, such as low socioeconomic families [26] or migrants and construction site workers, given that construction sites have the potential to create mosquito breeding sites [34]. The importance of involving labor employers and building contractors was also suggested to support effective SBC activities among migrant populations [34].

Larval source management

Household larviciding

These studies noted several behavioural barriers or challenges to effective household larviciding programmes. Recommendations are mixed on who should apply the larvicide within households. Aedes aegypti control programmes in Latin America have historically employed vector control technicians for household larviciding, to ensure safe handling of chemicals and proper dosing [24]. Under this model, significant resources are needed to maintain the operational effectiveness of the vector control programme, and the role of household members is limited to accepting the larvicide application and, to a lesser extent, maintaining it in between applications. However, a review article on achievements of Anopheles larval control in the tropical Africa region found that larvicide application (in particular, using Bacillus thuringiensis [Bti]-based granules or briquettes) could be safely conducted by trained household members or community volunteers because it did not require special equipment [38].

For household larviciding to be effective, identifying and accessing the larval habitats inside households and compounds is necessary, so that they can all be treated. However, breeding sites in the peri-domestic area can be challenging to locate, such as those located on elevated structures like rooftops or gutters, or in heavily cluttered areas around a home. Such challenges may be greater in urban settings due to greater housing density and multi-story structures. Access to breeding sites is also influenced by household permission for technician entry. Not allowing vector control personnel into the home poses another documented challenge to household larviciding, as some household members may be uncomfortable with unknown personnel entering their homes [36, 37, 63].

A review article on Anopheles larval control in Africa found two studies from urban Tanzania demonstrating that acceptance of household larval control programmes may be influenced by the reduction in or persistence of general nuisance biting mosquitoes, potentially risking dissatisfaction with larval control interventions where mosquito biting persists [38].

Multiple studies have suggested that community understanding, interest, and involvement are critical for the acceptability of household larvicide application and access to larval sites [37, 63]. In a review of the effectiveness and feasibility of bacterial larvicides on malaria vector control in sub-Saharan Africa across multiple studies, high levels of acceptability were observed in the five studies that monitored acceptance, even in urban areas, where there was early and meaningful engagement with communities [37]. Another study found that the use of technicians who are trusted by the community can help overcome barriers to gain entry into homes; when vector control staff were made up of women from the same community, most households (~ 97%) granted entry to their house premises, including in neighbourhoods where residents usually did not accept labor or municipal agents in their households [36].

One other article also focused on community engagement and support of larviciding programmes, noting successful experiences when creating true partnerships between the community and other sectors towards a shared goal of mosquito control being everyone’s responsibility. A quasi-experimental intervention on community mobilization for dengue control in Cuba found that building a partnership between the community and the government where the interests of all stakeholders involved (e.g., vector control staff; household members) are represented can be impactful in reducing household risk for disease. The article found improvements of 45–47% in household compliance with covering water containers and not removing larvicide from containers, and a 75% decrease in the absolute number of positive water containers over 2 years, suggested this partnership could be achieved by creating local task forces at the neighbourhood level with the goal of reducing community-based sources of mosquito larvae [71].

Community larviciding

Seventeen articles involved community larvicide interventions, defined as regular application of larvicide to larval sources, communal areas, and natural water bodies (not domestic, household containers), including six articles in which community larviciding was one of multiple interventions employed concurrently. The articles employed a range of qualitative, quantitative, or mixed-methods study designs and the studies took place primarily in African countries. Nine articles were reported from rural areas, while six were from urban areas, and two from both rural and urban areas. Of the articles including urban settings, six were from Tanzania (four from its Urban Malaria Control Programme), and one each from Kenya and Brazil. One article was a literature review of multiple interventions from countries in Africa, Asia, Central America, South America, and the Caribbean. Fifteen of the community larviciding articles documented in this review focused on malaria control, and one on dengue.

A key component of the effectiveness of community larvicide is community acceptance. Seven studies assessed the acceptance of community larvicide among community members. Of the studies that identified high community acceptance, one found that when combined with low knowledge on larviciding, the high acceptance rate would likely not be sustained, and that those who perceived biolarvicide as effective in reducing malaria were five times more likely to accept it [50]. Another study documented that few community members were aware of the larvicide programme and that those WHO were not previously aware of it were 60% less likely to have a favorable opinion of larviciding as a malaria intervention [47]. In urban Tanzania, the community larvicide application and surveillance was delegated to modestly paid community members, known as Community-Owned Resource Persons (CORPs), leading to high community acceptance and support [40] and possible reduced malaria transmission [44]. One study identified the need to combine the use of larvicide for malaria control with other vectors as well by introducing larvicide to breeding sites infested with Culex and Aedes mosquitoes. With that, mosquito nuisance would be largely reduced and although this addition would not directly control the malaria vector itself, it would increase the community’s perception of the success of larvicide application in reducing mosquitoes. The authors claim that improving the experience of communities by reducing nuisance biting can eventually lead to higher acceptance of future larvicide campaigns for malaria [45]. A study conducted in Taiwan noted the importance of considering both the vector ecology and the subculture of the community and emphasizing that mosquito control is everyone’s responsibility to enhance community cooperation and acceptance in community-based vector control programmes [68]. Two studies discussed other needs for community acceptance: willingness to pay for the larvicide by community members in rural and urban areas [39] and the vision of collective responsibility from community members and local governments [70].

In addition to community acceptance, effective implementation of larvicide interventions requires that community members are aware of the intervention, understand the process, and understand the people involved. Community involvement, in conjunction with other aspects of a larvicide application programme in urban Tanzania played a key role in achieving reduced larval abundance and malaria Transmission by 31% [40]. In that study, paid community members applied the larvicidal product. Another study in urban Tanzania, which also involved paid community members and measured malaria prevalence with randomized cluster sampling at multiple times, found that larvicide application was most effective in reducing malaria prevalence during dry seasons and when combined with ITN use or housing improvements [42, 44]. In addition to the reduction of mosquito abundance and nuisance, leading to the perceived effectiveness of Bti use, one article from Rwanda suggested cooperation with farming cooperatives to present Bti activities to rice farmers as is done for introducing agricultural chemicals or fertilizers, emphasizing safety, and leveraging trust between farmers and the cooperatives to increase community confidence in Bti [49].

Other studies covered the use of SBC and social engagement strategies to build the necessary relational capital for a community-based approach [48], effectively locating larval habitats when they are few and fixed (e.g., dry season) while also investigating existing community practices [41], active community participation in LSM activities to increase awareness of malaria as a health problem [46], conducting community-led education workshops, and using trained local volunteers to conduct larvicide application [67]. Additional factors involved in effective interventions include perceived safety and effectiveness of microbial larvicides, including an understanding of how the malaria transmission cycle may be logically interrupted by larviciding [51].

Despite the successes documented in this review, several studies noted challenges associated with community larvicide interventions. Challenges included lower accuracy of larval counting and identification of larval habitats by community-based larvicide teams, presumed to be due to insufficient dipping, examination, or training in mosquito identification, although it is notable for community members to collect such monitoring data [63]; low levels of participation in community-led LSM activities, due to concerns about labor intensiveness, time requirements, lack of financial incentives, and concern about health risks [46, 66]; and lack of community resources and commitment for community members to conduct the LSM activities [62]. Another study pointed out a concern expressed by focus group participants that community members may over-rely on the larvicide and reduce their use of ITNs or IRS upon the introduction of community larvicide interventions [49]. It is important to note that larvicide interventions require careful consideration of the operational settings, disease epidemiology, and effectiveness of the intervention [43]. At the programmatic level, key informant interviews with malaria programme officials in an assessment of a larvicide programme in rural and urban Tanzania noted they faced challenges due to insufficient technical knowledge to identify habitats of malaria (and non-malaria) vectors, how to apply the larvicide, and the safety of the larvicide, as well as inadequate supply of larvicide or funding for activities, some resistance from community members, and late or inadequate involvement or collaboration from other governmental and non-governmental sectors [47].

To overcome certain challenges and effectively conduct the interventions, several recommendations were found in the studies reviewed. A cross-sectional survey in Tanzania led to recommendations that education should focus on community benefits and safety of larviciding to gain acceptance from community members, especially for groups that demonstrate less trust in the intervention [50]. In a study conducted in Malawi, crop farmers, low-income households, and individuals with lower levels of formal education were unlikely to have high knowledge of mosquito LSM methods. This led to the conclusion that raising community awareness of these methods, especially among these populations, is imperative [65]. Studies pointed out recommendations for the overall impact of larvicide programmes: intensify training efforts (e.g., identifying habitats of important vectors), adopt technical principles for handling larvicides, improve financing for local implementation, and improve public engagement to increase community awareness and participation [47, 62]. During training sessions, it is important to communicate about LSM interventions and follow a learning-by-doing approach [51]. Other key recommendations include devising inclusive policy tools, especially those that enhance collective action, resource mobilization, and capacity strengthening to empower communities [39], and collaborating with other sectors, such as agriculture or waste management [49, 62].

Finding and removing standing water regularly

Twelve studies included in the review had relevant SBC findings for finding and removing standing water regularly as a vector control measure. The studies spanned African, Latin American, and Southeast Asian settings. There were only two studies that assessed finding and removing standing water as a standalone intervention; both identified Ae. aegypti as the target vector. One of these articles was a quantitative study in Honduras [52] and the other used a mixed-methods approach combining cross-sectional surveys, focus group discussions, and in-depth interviews in Pakistan [55]. The remaining ten articles addressed regularly finding and removing standing water as part of a mixed package of vector control interventions targeting Ae. aegypti in seven studies and Anopheles mosquitoes in three. The majority of studies that included finding and removing standing water were quantitative (n = 8), while two were literature reviews, one was a qualitative study, and a final publication utilized mixed methods.

Interventions aimed at finding and removing standing water can occur at both the household and community levels. At the household level, these interventions generally focus on identifying committed household members to lead in finding water collection and storage containers and removing them from the households. A household and school-based cluster randomized-controlled trial in Kenya identified the need for household and community engagement for long-term change, including a focal point for container management within the household, finding a 50% increase in knowledge and self-reported behaviour of caregivers who were engaged in the study intervention. This study found that caregivers would be the most appropriate to find and remove containers since they are the ones primarily responsible for water collection and storage activities, however the authors also concluded that commitment from all members of the household is needed to maintain water containers free of larvae [64]. For interventions to be effective, household members must be aware of what stagnant water containers entail and why it is important to remove them. In addition, interventions at the household level may have the potential to empower individuals to take actions in homes more than it would at the community level as removal of tires and discarded containers at the community level often involves significant municipal and community cooperation for successful implementation [69].

Another study in Malawi also identified areas of empowerment in community-level interventions to remove standing water. This study found that promoting collective action from existing local organizational structures and individuals as well as education, led to community empowerment. Authors suggested that the intervention could be complemented by community workshops by trained local unpaid volunteers, such as health animators and village-level committees [67]. Community-level efforts have the potential to reduce levels of mosquitoes in breeding sites, as observed in a Taiwan study that promoted community-supported efforts on Aedes source reduction through a broad set of interventions that included discarding water storage containers that were not being used anymore and reported that Ae. aegypti became nearly extinct in four of the intervention villages [68].

Community engagement and inter-sectoral involvement were discussed as ways to promote this LSM-based intervention at the household and community levels. Community engagement can increase the efficacy of removing standing water by including activities such as active mosquito searches in homes and communities and promoting awareness of the mosquito life cycle. However it is recommended that instructions should be specific enough for community members to understand and effectively locate the highest density breeding sites to be able to weekly remove the specific targets [24, 53, 56, 57]. In addition to these efforts, community members need to be aware of the willingness of community leaders, government organizations, the private sector, and community-based civil society organizations to collaborate and further engage in vector control activities including the reduction of larval habitats in communal areas [54]. The success of such recommendations is made possible upon political will and commitment, community sensitization and participation, financial resources, and inter-sectoral collaboration, leading to community empowerment [62].

A study in Honduras found that a promising approach to train community members in breeding site identification and removal around their homes is to first train local stakeholders such as community leaders and government officials, who in turn train community members in the key practices using instructional print materials [52]. A study in Pakistan used a positive deviance intervention in which community members in an intervention area developed interactive activities (sketches, flip chart presentations, storytelling events) to improve community dengue knowledge and practices. Knowledge regarding dengue Transmission, prevention practices, and symptoms improved at 2 months and 4 months of follow-up, with statistically significant differences between the control and intervention groups. Authors suggest that community-led design of SBC activities leads to greater engagement and interest [55].

A control-intervention study in both Mexico and Nicaragua that promoted community engagement in pesticide-free LSM, including finding and removing breeding sites, noted that vector control programmes benefited from engagement with broader sectors of the community. Collaboration with schools, and operational integration with local and municipal services (e.g., water supply; waste collection) were noted as contributing to the effectiveness of vector control measures, while also enabling each community to implement the intervention in its own way. The interventions resulted in a relative risk reduction of dengue virus in children of 29.5% [61].

Finally, a systematic review of LSM interventions for dengue prevention found that to expect success in community or household-level initiatives to find and remove standing water, clear communication about how the actions need to be performed to be effective against the targeted mosquito life stage is critical. The review noted that there can be misperceptions on how to manage water storage containers effectively and that the behaviours to be promoted must reflect the local community’s skills, time, and resources. For example, a plan to remove trash from premises required cooperation between households, community leaders, and government officials to create and sustain trash removal infrastructure [27].

Covering water storage containers

Covering water storage containers was the sole focus of three studies, however relevant SBC findings for the intervention were identified from a total of nine articles. All but one article assessed covering water storage containers in the context of Ae. aegypti in primarily Latin America settings (n = 5) with one publication each in Southeast Asia and Africa. In total, four of the studies underwent a quantitative-only study design, followed by two qualitative-only studies designs, one mixed-methods study, and two literature reviews including one systematic review assessing Ae. aegypti mosquito reduction interventions for the prevention of dengue across various settings [27].

Only one identified study assessed the effectiveness of covering water storage containers for malaria, which was specifically for An. stephensi control in India [59]. This pre-post intervention quantitative study in rural India found the intervention village, where tanka (household underground water storage tank) lids were replaced with improved mosquito-proof polyvinyl chloride plastic (PVC) lids resulted in “complete absence of mosquito breeding” with reductions of 94.9% and 97.9% of An. stephensi larvae in the post-intervention and at the 1-year follow-up, respectively [59]. Furthermore, this study found that the active engagement of community members in the lid replacement process increased knowledge of the importance of proper lids and mosquito-proofing of underground tankas [59].

The importance of effective covering was noted in multiple studies. A quantitative study assessing the effectiveness of this intervention for Aedes larval control in urban areas found that using an effective lid versus no lid was associated with a 94% decrease in odds of Ae. aegypti larval presence in long-term water storage containers in three Latin American countries; an effective cover was defined as having no observable holes that completely sealed the container, and had no water accumulated on it [58]. Likewise, a study in suburban villages of Laos and Thailand conducting cross-sectional household entomological surveys, found that household containers with lids were significantly less likely to be infested with Aedes larvae than those without lids, with reductions of 70% in Laos and 90% in Thailand [17]. The authors did not collect data on the condition or quality of the lids, but did note that based on other studies, a partial fitting lid is ineffective and recommended that to be effective, lids should be properly used, maintained, and tight-fitted [17].

A literature review of Zika prevention behaviours across Latin America and the Caribbean [24] cited a study in Thailand [19] that found a similar effectiveness of correctly covering containers with lids to reduce larval infestation in jars storing drinking water (AOR: 0.1–0.25). However, the study authors noted that covering containers alone was not enough. Rather, they found that partially or incorrectly covered containers could actually favor the production of mosquito larvae with 34.9% of larvae identified in incorrectly or uncovered jars compared with 7.8% in containers covered correctly [19, 24]. Therefore, the authors noted that highly specific guidance for the type of lid and/or water storage container is needed to ensure the intervention is implemented effectively and does not create new breeding sites. Such guidance includes ensuring the lid makes a very tight seal, does not touch the water, does not accumulate rainwater, and is of a material that is durable with respect to exposure to heat and sunlight [23, 24].

In addition to having an effective cover, the type of water storage container was another consideration, as high levels of efficacy were observed for long-term water storage containers but not for frequently used washbasins [58]. This is supported by the review in Latin America that reported covering long-term water storage containers was found to have moderate potential efficacy in reducing breeding sites if a tight-fitting, long-lasting lid is available; however, covering short-term water storage containers had less potential efficacy, as frequent lid use can result in wear and tear and render the lids ineffective or counterproductive [24].

Multiple studies reported that acceptance of covering water storage containers tends to be high [17, 24]. In Kenya, one study found that covering water storage containers was the most reported source reduction behaviour due to the desire of respondents to keep water clean, especially water used for drinking and cooking [64]. Furthermore, a quasi-experimental study in Cuba, found that creating local task forces of community working groups in which interests of all direct stakeholders were represented contributed to the successful implementation of the intervention such that, the number of uncovered water storage containers had decreased by 46.7% (P < 0.01) and the number of houses with unprotected artificial containers had decreased by 55.5% [71].

Additional considerations around access to and maintenance of lids were also identified. Focus group discussions in urban and peri-urban areas of Guatemala identified challenges with access to proper or effective covers, as most evident for water storage drums, reducing people’s perceptions of the feasibility of the intervention. Therefore, this study suggested households may opt to only cover water used for drinking or food preparation, as proper lids for those smaller containers may be easier to access or more affordable. Similarly, while covering containers could potentially be implemented at the community level, the authors suggest this intervention may be best limited to the household level given the perception that joint action to maintain covers on community water sources is difficult [69]. Finally, findings from a cluster non-randomized controlled trial assessing the effectiveness of an eco-bio-social strategy for Aedes control in sub-urban areas of Brazil, evidenced that for large, elevated tanks, household members tend to neglect and delegate responsibility of repair and proper lid fitting to professionals who repair water pump parts as they had little time to worry about vector control and rather focused on ensuring immediate livelihood. Therefore, this study recommended training for professionals working in the maintenance of large, elevated tanks in mosquito-proof covering [60].

Discussion

Summary of key findings and implications

This scoping review was carried out to identify social and behavioural considerations for interventions likely to be included in the response to An. stephensi in Africa, including peer-reviewed literature from malaria and Aedes control. The findings provided an evidence base to inform the development of the Social and behavior change guidance for Anopheles stephensi in Africa [23].

Several important themes emerged from the review, including: (1) In low malaria transmission contexts, malaria-related prevention behaviours tend to be practiced at low rates, so it is important to address the factors driving these behaviours and effectively tailor SBC to the needs of higher risk groups; (2) It is critical to create early and meaningful partnerships with affected communities; and (3) Highly specific guidance for governments, communities, and households is paramount, particularly for promoting behaviours for LSM interventions, as missing any specific step in the necessary behaviours for these interventions will reduce their potential effectiveness.

In areas that have historically had low or very low transmission rates, such as urban cities, perceived risk of malaria is often lower, and may correspond to lower levels of malaria-related behaviours, such as ITN use, IRS acceptance, and prompt care-seeking for fever [72]. Few peer-reviewed articles were identified specifically related to core interventions in urban settings, likely because malaria has historically been concentrated mostly in rural areas and therefore publications have also focused on these contexts. However, several important considerations were identified from the literature of relevance to the An. stephensi response.

It is important to increase malaria risk perception and emphasize why core interventions are critical. The literature suggests SBC strategies should promote consistent net use all night long and throughout the year and promote replacement options, emphasizing ongoing net care and replacement as a household responsibility [29]. For IRS, it is important to understand the unique barriers to acceptance and engage communities and leaders effectively from planning through implementation. Barriers may include logistical or social barriers to removing household possessions, low perceived effectiveness of IRS, perceived increase in other insects, physical side effects, concerns about IRS activities being politically motivated, selection and performance of spray operators, or previous negative experience with IRS campaigns [31, 32]. Providing information sessions, opportunities to answer follow-up questions, and active participation of community leaders can help to overcome barriers [31]. Where supplemental vector control interventions (such as LSM) are promoted, it is important to clarify that these are not sufficient to control malaria alone and to maintain promotion of ITNs and IRS in households where they are part of the existing strategy [49, 73].

The literature also highlights the importance of targeting SBC to groups of people at higher risk for malaria. This may include groups such as migrant or construction workers who move between higher and low transmission areas or whose risk may be increased due to their occupation, as well as other vulnerable populations. Ensuring access to appropriate prevention measures and promoting prompt care-seeking for fever is critical. When targeting these higher-risk groups, SBC interventions should consider local languages and effective use of informal communication channels. For those groups whose risk is tied to occupation, involvement of labor employers and building contractors can support effective SBC approaches [34]. For other vulnerable groups, such as pregnant women, increased outreach in urban and peri-urban areas where malaria-related behaviours may be lower, is important. This may include outreach through ANC visits [28].

A key theme across interventions was the importance of the quality and approach to engaging households and communities. This was particularly true for interventions that require entry into the peri-domestic space, such as IRS and several LSM interventions. Early and frequent engagement with communities and civil society, building trust in the technicians implementing interventions, and inclusion of community members in implementation activities were found to increase acceptance of interventions and ultimately impact. For example, access to breeding sites to carry out household larviciding is heavily influenced by household permission for technician entry and it is critical that household members are comfortable with personnel entering their homes [36, 37, 63]. Building trust and, when possible, employing community members can help to increase acceptance. For example, employing community members to apply larvicide and support surveillance leads to high community acceptance and support and in turn reduced larval abundance and transmission [40]. In some contexts, working with cooperatives (e.g., rice farming) to emphasize safety and leverage trust between farmers and the cooperatives can increase the community’s confidence in larviciding [49]. For finding and removing standing water, community-level interventions often involve significant municipal and community coordination and cooperation [69], underscoring the importance of effective coordination across stakeholder groups.

Related to effective community partnerships is ensuring high levels of understanding of interventions and why they are being implemented. Community education on mosquito biology, awareness, and involvement in LSM have been key factors in the success of LSM-driven vector control programmes. For example, multiple studies found that greater knowledge of larviciding increased support for its application [47, 50]. For interventions that may be novel or less familiar, it is important to ensure that the rationale for the new intervention is clear, as well as the importance of continuing to use existing interventions.

Another important theme that emerged from the literature was the importance of specificity, particularly for LSM interventions. For finding and removing standing water, articles highlighted the need for clear and specific instructions to support community members in locating the highest-density breeding sites and removing the specific targets at specified intervals [24, 53, 56, 57]. Specificity was perhaps most important for covering water storage containers. Effective covers for water storage containers can reduce An. stephensi and Ae. aegypti larvae [17, 58, 59]. However, to be effective, lids should be properly used, maintained, and tight-fitted [17], as incorrectly covered containers can favor production of mosquito larvae [19, 24]. High levels of efficacy were observed for long-term water storage containers but not for frequently used containers [24, 58], and households may be more likely to cover small containers as lids may be more accessible and can keep drinking and cooking water clean [64, 69]. For large tanks, repair and proper lid fitting may need to be done by professionals [60]. These findings underscore the necessity of providing highly specific guidance for the type of lid and/or water storage container, ensuring access to effective lids, and where applicable, appropriate training for professionals who maintain water sources. Several tools—for Ae. agypti and An. stephensi—have been developed by the U.S. Agency for International Development (USAID) partners to guide SBC that is specific enough for individuals, families, and communities to take effective actions for An. stephensi:

• Social and Behaviour Change Guidance for Anopheles stephensi in Africa: https://breakthroughactionandresearch.org/resource-library/sbc-guidance-anopheles-stephensi-in-africa/

• Effective Behaviours to Prevent Zika—Removing Standing Water: https://breakthroughactionandresearch.org/wp-content/uploads/2019/04/standing-water.pdf

• Effective Behaviours to Prevent Zika—Cover Water Storage Containers: https://breakthroughactionandresearch.org/wp-content/uploads/2019/04/water-containers.pdf

Research gaps and opportunities

There are also several important opportunities to fill gaps in evidence identified through this review. First, the review returned a dearth of An. stephensi-specific information. While the review identified important social and behavioural considerations, context-specific data is needed to inform intervention design, implementation, and evaluation across settings in Africa. Examples of important data include context-specific breeding sites, identification of higher risk and mobile populations, context-specific risk factors for exposure, as well as psychosocial and contextual factors that influence behaviours of interest across settings. Through research and implementation, there is an opportunity to continue building the evidence base for SBC related to An. stephensi interventions in Africa and to document best practices and lessons learned as countries respond.

Furthermore, while this review focused on individual, household, and community behaviours, it is important to consider the behaviours of a broader range of stakeholders and decision-makers as well as the context in which they operate. One critical group to consider is health care providers. While there were limited studies with relevant SBC considerations for care-seeking for fever in urban areas, none captured elements of health provider behaviour and its relationship with demand for services and perceived quality of care. This highlights an important gap within the context of An. stephensi. With the rapid rise in urbanization in Africa [74] and expectations that it will only continue to increase, community members in urban areas may be more likely to bypass the formal health structure or prioritize private facilities, including pharmacies [75, 76]. Therefore, it is important to address provider behaviour holistically in these settings accounting for environmental factors that impact quality of services (commodity availability, provider competencies, facility environment) and individual and interpersonal factors like provider self-efficacy, risk perceptions of malaria, attitudes, and norms. An emphasis on effective service communication [the use of SBC activities by health providers to influence malaria-related behaviours) can help improve the quality of provider-patient interactions in public and private facilities, increase the adoption and maintenance of malaria prevention and treatment behaviours in contexts with reduced risk perception, and contribute to creating a cycle of good provider-patient relations, which may lead to increased demand for, and use of, malaria control products and services [33–35].

In addition to considering SBC for a broader set of audiences, it will likely be important to consider SBC for a broader set of interventions in the future. This review focused on interventions that were considered most likely to be deployed immediately in the response to An. stephensi in Africa, however, there are several vector control interventions currently being evaluated for public health impact that could be deployed in the coming years if proven efficacious. This includes interventions such as spatial repellents, mass drug administration (MDA) of an endectocide to humans and/or livestock, and housing modifications. Endectocides, for example, have the potential to target indoor and outdoor biting mosquitoes as well as mosquitoes that feed on both humans and animals. Integrating relevant SBC considerations for these interventions in the future, and ensuring context-specific social and behavioural research, if and when they are considered for wider implementation, can help to ensure the effectiveness of these new interventions. Further, there are molecular vector control technologies being developed that can specifically target An. stephensi, including Oxitec’s Friendly™ An. stephensi [77], which carries a self-limiting gene preventing female offspring from surviving, leading to male-only production. While development is in early stages, with the first pilot release in Djibouti in 2024, the technology could play an important role in the An. stephensi response in the future. Community and stakeholder engagement, as part of an established framework for introducing genetically modified mosquitoes [78], is critical for the success of this molecular vector control approach and to inform future social and behaviour change interventions.

Finally, multi-sectoral engagement, including integrated approaches to policy development and structural changes, will be critical to sustainability and the integration with water infrastructure and management personnel emerged as a recurring theme in the articles reviewed. The dynamics of mosquito vector-human interactions are complex and the reliance of both species on water creates numerous opportunities for mosquito-borne diseases to thrive in settings without infrastructure considerations to prevent mosquito breeding. Effective and sustainable control of An. stephensi therefore necessitates a coordinated multisectoral approach due to the diverse infrastructure and livelihood linked habitats in peri domestic, urban, and rural settings that An. stephensi thrives in. Artificial water storage containers, tankers, construction and brick pits, and other water bodies commonly found in association with urban development are sites where An. stephensi larvae are commonly reported. Effective LSM for An. stephensi therefore requires collaboration across various sectors, including public health, civil engineering and urban planning, water and sanitation management, community engagement, and others. For example, urban planners and water management authorities may be engaged to ensure built environment and development infrastructure reduces, limits, or regulates stagnant water to minimize impact of An. stephensi and other vectors on public health. The WHO’s recent documentation of country experiences in An. stephensi control from India, Iran, and Sri Lanka highlights multi-sectoral approaches and recommends water and waste infrastructure developed with vector control in mind, putting in place and enforcing building standards and civic by-laws that prevent An. stephensi proliferation, and an integrated approach to Aedes and Anopheles vectors [79]. Without integrated efforts, isolated vector control measures, such as larviciding alone, may be less effective and sustainable. Additionally, global guidance generally reinforces that malaria control programmes should use LSM in conjunction with other malaria prevention measures, and not as a standalone approach.

Limitations

While this review provided valuable insights on social and behavioural interventions for responding to An. stephensi in Africa, there are several limitations which should be considered when interpreting the findings. First, it is possible that the search strategy did not identify all relevant articles. This review included peer-reviewed journal articles and may have missed important learnings from routine implementation that have not been published. The review focused on literature published in English from 2000 through 2021, and relevant articles published outside of this timeframe and in other languages could also have been missed. Second, the review focused on social and behavioural considerations for interventions most likely to be included in the response to An. stephensi in sub-Saharan Africa. It did not include interventions that have been used in other contexts, including larvivorous fish. Additional research is needed to better understand specific contexts where larvivorous fish could be used in the response in Africa, and considerations for sustainability and cost-effectiveness. If implemented, country experiences from Sri Lanka and India underscore the importance of household-level behaviours and acceptance and suggest effective SBC will be critical [79]. Third, this review uncovered a lack of studies directly focused on SBC for An. stephensi; as a result many recommendations are based on studies on Ae. aegypti. When translating lessons from Aedes control, it is important that malaria programmes assess and consider any contextual biological and social differences in disease burden and public perception specific to An. stephensi. Finally, the quality of studies was not evaluated as part of this review nor was meta-analysis feasible. Results were synthesized qualitatively to identify important themes across studies of relevance to SBC programmes.

Conclusions

This scoping review identified important social and behavioural considerations for the An. stephensi response in Africa and helped inform the development of evidence-based guidance. The unique characteristics of An. stephensi, including its ability to transmit malaria in urban and peri-urban areas and in dry season, and laying eggs in artificial water storage containers, shape the interventions needed to effectively respond and the role of individuals, households, and communities in that response. It is important to increase malaria risk perception in areas where transmission has historically been low, and tailor SBC to higher-risk groups, including mobile populations and those with occupational risk. Early and meaningful partnerships with communities is critical for building trust and increasing the acceptance of interventions. Evidence from Aedes control provides valuable insights, particularly for LSM interventions that may be introduced as part of the response in Africa. Perhaps most important is the need for specificity when promoting LSM behaviours, particularly finding and removing standing water and covering water storage containers. While the review identified valuable social and behavioural considerations, the limited number of An. stephensi-specific articles identified on this topic underscores the need for context-specific social and behavioural research and documentation of best practices as countries respond to the threat. Finally, there is an important opportunity for more integrated approaches to SBC, including multi-sectoral collaboration, effective partnership with community leaders and civil society, and where appropriate, increased coordination with Aedes control.

Abbreviations

Bti

Bacillus thuringiensis Var. israelensis

CORP

Community-owned resource persons

FGD

Focus group discussion

IDI

In-depth interview

IRS

Indoor residual spraying

ITN

Insecticide-treated net

LSM

Larval source management

PMI

United States President’s Malaria Initiative

SBC

Social and behaviour change

WHO

World Health Organization

U.S.

Agency for International Development (USAID)

Author contributions

AM, JH, GH, SC, AA, BH, SZ, and MY conceptualized the parameters for the literature search. JH conducted the search and screening. JH and AM extracted the results, and with GH, interpreted the results. JH drafted the methods section and JH, EM, AM and GH wrote the results. AM drafted the discussion and conclusion sections with contributions from SZ and SC. GH, AM, SC, BH, SZ, and MY reviewed and revised the manuscript. All authors reviewed and approved the final manuscript.

Funding

Breakthrough ACTION was funded by the U.S. Agency for International Development (USAID) and U.S. President’s Malaria Initiative (PMI) under the terms of Cooperative Agreement No. AID-OAA-A-17-00017. The funders were involved in the study design and preparation of the manuscript, and not involved in the data collection, analysis, or decision to publish.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable. As a review article, no study participants were recruited for participation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.