Family Vector Control Response Kit: Equipping Households to Prevent Amplication of Aedes-transmitted Diseases

Natural disasters, such as earthquakes, hurricanes, and oods create ideal mosquito breeding conditions while simultaneously disrupting government services that prevent outbreaks where Aedes-transmitted diseases, and possibly other vector-borne diseases (VBD), are circulating. UNICEF, the MENTOR Initiative, and Kenya Medical Research Institute (KEMRI) evaluated novel UNICEF Family Vector Control Response Kits distributed to households (HH) in Wajir Town, Northeast (NE) Kenya at the end of the rainy season (January/February 2019). This region regularly experiences epidemics of VBD during, and for several months after, rainy seasons. The premise behind this study was to determine if HHs can adopt the use of the kit contents based solely on pictures rather than needing a comprehensive education campaign. This was necessary for the following two reasons: 1. Successful disruption of a disease outbreak often requires transmission be stopped at the HH level as soon as possible, 2. In the beginning of most large-scale rapid response emergencies, it is challenging to enact and align comprehensive education campaigns with delivery of supplies. To stop transmission early, it is often faster, in UNICEF’s experience, to deliver supplies directly to HHs, especially if the global/regional/national supplies are strategically prepositioned in warehouses. The study aimed to determine key outcomes when HHs were given simple picture-based application instructions, without any other education, for different combinations of evidence-based insecticidal products, including adulticides, spatial repellents, larvicides, insecticidal window curtains, and personal protection. The products were those commonly expected to be found in retail markets or will become available through public health procurement channels. transuthrin effectively protected against indoor-biting and outdoor-biting Anopheles mosquitoes 24 . A 2019 cluster randomized placebo control trial conducted in Sumba, Indonesia with the spatial repellent showed a signicant impact on malaria infection (about 60% protective ecacy) among the highest-risk clusters 25 . More recently another randomized control trial provided conclusive evidence of the protective ecacy of transuthrin based spatial repellents against Aedes vectors of arboviral diseases 51 .

The concept of integrated intervention for disease prevention, or integrated vector management (IVM), is gaining increasingly broader acceptance, although to date, consensus is still to be reached regarding the details of how and what combination of approaches can be most effectively implemented to manage disease 7 . Best practices in vector control must be de ned for each setting (i.e. which tools or methods the community should employ), as well as what constitutes adequate or su cient coverage in order to impact the vector population and virus transmission. Different responses will also have different outcomes depending whether applied in an urban or rural setting or a natural disaster or refugee crisis situation. Horstick and Runge-Ranzinger 8 found that: (1) vector control could be effective, but implementation and coverage remained an issue; (2) single interventions were probably not useful; (3) combinations of interventions had mixed results; and (4) careful implementation of vector control measures may be most important. While the concept of vector control is reasonable, control must be early in an outbreak or strategically applied during inter-epidemic periods to prevent escalation in transmission. Community participation improved all interventions employed associated with it and so it is an indispensable component in any control programme 9 .
A key factor about vectors transmitting diseases is they are often found within and in close proximity to houses. Prevention measures, especially against Ae aegypti, are focused on that environment. Numerous control practices against eggs, larvae, and adults are supported by WHO recommendations 10,11 .
Individual components such as strong community engagement in elimination of larval habitats 12 , larvicidal treatment of water holding containers, adult control using coils, consumer aerosol insecticides, spatial repellents, and insecticide treated curtains have been shown to have bene t in reducing indoor mosquito populations [13][14][15][16] . However, there remains a vital need for robust, rigorously designed eld trials, with epidemiological and entomological outcomes, to improve the understanding of optimal implementation of existing tools and measure reduction in VBD transmission. Only then can strategyspeci c programmes be realistically developed and evaluated 7 .
Based on this background a strategy was developed that would capitalize on the strengths of UNICEF with regards to engagement in humanitarian and emergency situations in resource-limited settings, maintenance of a visible presence in communities, government advocacy, and operation of an unrivalled global supply capability. A component of the strategy was to improve the speed of response when an Aedes-borne disease was rst con rmed or highly suspected. Built on UNICEF's experience with other disease management it was thought that if the infected person could be isolated as infection loci then further transmission might be delayed or prevented for a short period 49 . That action would give a window of time for more formal, widespread governmental vector control activities to be initiated, thus preventing or reducing the chance of the localized outbreak from becoming epidemic 50 . In order to isolate the infected person from mosquito exposure the concept of a Family Vector Control Response Kit was devised. The kits would be supplied by UNICEF or other aid organizations and distributed by health centre workers to all HH in camps or villages where infected persons reside or that have the potential for transmission. The kits would contain commonly available consumer insecticidal products for 1) targeted controlling or repelling adults and treating hiding places, 2) preventing entry of mosquito adults into those houses, 3) stopping emergence of new individuals from water containers within the houses where infected persons lived, and 4) repellent for personal protection. The kit contents would be used by members of the HH and would afford immediate mosquito control without delay that might occur waiting for other actions in the community.
Useful consumer products that could be included in the kits are available in most countries (aerosol insecticides, coils, spatial repellents, hand-trigger sprayers, curtains, larvicides, and repellent lotions).
Typically these products are supported by e cacy data and are registered by national regulatory authorities. However, labelling that gives speci c directions how safely to use and where to apply the products in the house to control vector mosquitoes are complex or unclear for the typical resident in settings where outbreaks might occur. Challenges correctly interpreting and following label use instructions can become even more acute in populations with low levels of formal education and low literacy rates. It is documented that even farm workers who apply pesticides on a regular basis do not have full comprehension of label instructions for use and safety 17,18 . However, the United Nations produced the Globally Harmonized System of Classi cation and Labelling of Chemicals (GHS) 19 that includes a series of pictograms that could be used by manufacturers to provide guidance on safety and handling of their products. With the expectation that use directions needed to be universally understandable, and with prospects that users may be limited in translation or reading ability, consideration was given to providing kits with pictogram directions for use (DFU) for each of the products. The pictograms would reinforce applications that targeted adult mosquito resting and hiding places, coverage, timing, and precautions. The graphics would also reinforce proper treatment of water containers that could not be easily emptied.
The main objective of this study was to determine if participants in a rural setting, that was similar to the environment of a humanitarian calamity or natural disaster, could follow pictogram use directions for each of the products included in Family Vector Control Response Kits. The kits were con gured with different product types commonly found in supermarkets. Preferences and ease of understanding the targeted applications were measured between technologies.

Study Site
A group comparison study was conducted to evaluate HH acceptance and usage of six different combinations (kits) of vector control products.
This study was conducted in Wajir town, which is the capital and largest city in Wajir County in NE Kenya (Fig. 1). It sits at a latitude and longitude of 01°45′00″N 40°03′00″E. Wajir shares its borders with Somalia to the east and Ethiopia to the North. It is part of territory carved out of Somalia during colonial times and as such, the population is 90% ethnic Somali. In 2019 the town had a population of 90,116 20 . Wajir is located in an arid area prone to drought 48 . The county is also prone to seasonal ooding during its two rainy seasons; the 'short' rains experienced between October to December and the 'long' rains from March to May each year. This is a highly seasonal epidemic zone, with low transmission of malaria, rift valley fever, lymphatic lariasis, dengue or other diseases outside of the rainy season 21 . This means that all age groups are susceptible to the diseases transmitted by the bite of infective Aedes, Anopheles, and Culex mosquitoes. The population of interest included all occupants of enrolled HH comprising children and adults of all ages.
Wajir town is made up of 10 distinguishable neighborhoods. From these 10 areas, seven were randomly selected, based on the ecological, economical activities, educational level and reasonably uniform house structures and environmental conditions. These seven areas formed the treatment groups, or arms, for this study. All the villages (manyattas) within each of the seven treatment groups were evaluated for suitability. Out of these, two villages from within each group were randomly selected for the study to make up the targeted number of HH per treatment.
A total of 390 HH were mapped and randomly assigned to seven arms. Mapping was done using tablets for data collection, and GPS coordinates were entered into an electronic data collection tool (supplied by KEMRI). Each treatment arm had 60 HH and the control arm had 30 HH. On average, there were 8 persons per HH for a total of 3,120 persons impacted by the study.
Coordination meetings were held with the County Director of Health and the Director of Public Health in Wajir to plan for the study as well as to select workers for HH mapping, distributing kits, and conducting post distribution observations/interviews. The workers were either public health o cers or graduate students. These study survey team members were trained on interview techniques and methods to conduct observational studies. Observational and interview questionnaires were discussed at length and minor amendments were made. This exercise helped make the questionnaires easy to understand/implement and relevant for the local context. A second day of training was spent in the eld practicing various techniques that were taught during the theoretical part of the training.
After kits were distributed to all households, a team of 2 monitors performed close observations of 10 households in each research arm to better understand if use directions were being followed. They spent a full day observing each household from kit receipt until each product package had been opened and attempts had been made to use every product in the kit. A second team of 2 monitors visited the remaining 50 HH in a treatment arm a few days after the HH had received, opened and used the products. A standardized follow-up questionnaire was used to query if people in the HH could explain and demonstrate the use of the components and give impressions of the kit concept. Data was tabulated by MENTOR staff.

Test Design
The treatment arms consisted of 6 different kit con gurations based upon the inclusion or exclusion of long-lasting insecticidal curtains (LLIC). The purpose of the curtains was to prevent entry of new adults after adult mosquito control treatments were applied to eliminate resident vectors. The mosquito adult control tools included mosquito coils, aerosol insecticide sprays, and slow-release passive repellent emanators (referred to as spatial repellent in this study) 22 . Kits for all arms included a trigger hand pump bottle of insecticide for targeted application to hiding places, a bottle of larvicidal lm, a y swatter for mechanical control of stray mosquitoes that might be ying around in the house, and a bottle of personal repellent. Kit products were assembled and packaged into paper bags in six different combinations, together with pictograms for each of the kit components (for example see Fig. 2).
Kit con gurations for each arm are shown below: Arm 1: Spatial repellent + LLIC + trigger sprayer, larvicidal lm, y swatter, & personal repellent Arm 2: Spatial repellent + trigger sprayer, larvicidal lm, y swatter, & personal repellent While not yet registered in Kenya, the spatial repellent product included in this study was shown to be effective against both vectors and disease incidence in other testing. A 2017 laboratory and semi-eld test involving female Aedes mosquitos saw a 96% reduction in blood feeding success. When hung near entry points, the emanator reduced mosquito entry by 88% 23 . A 2018 study found that eave ribbons treated with the spatial repellent trans uthrin effectively protected against indoor-biting and outdoor-biting Anopheles mosquitoes 24 . A 2019 cluster randomized placebo control trial conducted in Sumba, Indonesia with the spatial repellent showed a signi cant impact on malaria infection (about 60% protective e cacy) among the highest-risk clusters 25 . More recently another randomized control trial provided conclusive evidence of the protective e cacy of trans uthrin based spatial repellents against Aedes vectors of arboviral diseases 51 .

Mosquito Coil
The mosquito coil used in this study had d-allethrin (0.2%) as active ingredient (Kenya Reg. No. PCPB(CR)0972), brand name KAPI FLOWER® Mosquito Coil, manufactured by KAPI Ltd., Kenya. A systematic review of mosquito coils and passive emanators concluded they provide a deterrence effect (prevent entry of mosquitoes) of 40-80%, increase proportion of mosquitoes that exit earlier from huts burning coils compared to huts without coils, and confer protection against mosquito bites 26 . A 2013 study evaluated the bio-e cacy of ve commonly used d-allethrin (0.1-0.3%) based coils on non-blood fed female adult An gambiae mosquitos. It found that mortality ranged from 36% − 72% with KT50 and 34.92% with KT90 27 .
Aerosol Spray Can SUPAKILL® aerosol spray (Kenya Reg. No. PCPB(CR)0800) active ingredients include: pyrethrins (2.0%) + cypermethrin (1.56%) + tetramethrin (4.0%) + piperonyl butoxide (5.0%), manufactured by Pyrethrum Board of Kenya and registered by Sumitomo Corp. A simulated eld study on the e cacy of commercial household aerosol insecticides was conducted in 2009 and found that the eld e cacy of commercial household aerosols applied at their recommended dosages against Ae aegypti mosquitoes induced total mortality 28 . Another study of 13 commercially available aerosol insecticides showed that 3 of the products provided 100% control after 24 hours exposure but there was considerable variability in e cacy provided by other products 29 . The authors also concluded that location of mosquito cages in a room affected results. In another study comparing consumer aerosols for control of An gambiae in Kenya it was demonstrated that some products had rapid knockdown and mortality but products with the same brand name coming from another country did not provide similar speed of action 30 . It was also determined that indoor resting densities were not greatly reduced a few days after spraying. The authors concluded the spray did not suppress entry of mosquitoes from outside.
Piperonyl-Butoxide (PBO) is a synergist that increases the insecticidal activity of pyrethroids in areas with pyrethroid-resistant mosquitoes. While there was awareness of the increasing pyrethroid resistance in Kenya, the range of product choices for use in trigger sprayers was limited to those which are registered by the Pest Control Products Board (PSPB) and which are easily available in Kenya.
Long-Lasting Insecticidal Curtains (LLIC) DawaPlus® 2.0 (Kenya Reg. No. PCPB(CR)1321) active ingredient includes: deltamethrin 80 mg/sqm, manufactured by Tana Netting, a subsidiary of NRS Moon Netting FZE as an insecticidal net that can also be manufactured as curtains. The Thirteenth WHOPES working group paper measuring the e cacy of LLIC to kill and or inhibit blood-feeding of An gambiae found that mortality and knockdown was 100%. Even after washing the curtains 20 times they remained effective with mortality at 96.6% and knockdown at 98.3%. Furthermore, when tested in huts, the blood feeding rate in huts with LLIC was 4.8% as compared to control huts of 37.5% 31 . Tests in Zaire against Ae aegypti mosquitos also demonstrated the e cacy of deltamethrin-impregnated curtains with a 100% reduction in indoor biting 32 . In a study in Mexico LLIC were shown to reduce the indoor abundance of Ae aegypti in HH in rural but not in urban/suburban study sites 33 . In Cuba there was strong acceptance of LLIC initially but that enthusiasm waned over time because of lowered perceived e cacy. The HH expected to observe dead mosquitoes beneath the curtains in that trial 34 .

Larvicidal lm (LF)
A solution of 78% polydimethylsiloxane (silicone), brand Aquatain® AMF, was prequali ed in 2018 by the World Health Organization (WHO) for public health use 35 , is exempt from EU Biocide regulations, and is registered in over 60 countries for use by HH or professionals. The manufacturer, Aquatain Products Pty Ltd, Australia, has received permits for import and testing by the Kenya Pest Control Products Board. This LF spreads across the surface of standing water, forming a very thin coating which lasts for at least 4 weeks. As the silicone polymer has a low surface tension, mosquito larvae and pupae cannot attach to the surface to breathe -causing them to drown 36 . A trial was conducted to evaluate the potential of LF as a mosquito control agent in a Kenyan rice irrigation scheme. An application at a dose of 2 ml/m2 on rice paddies showed early stage anopheline larvae numbers being reduced by 36%, and late stage anopheline larvae by 16%. However, even at lower doses of 1 ml/m2 there was a 93.2% reduction in the emergence of anopheline adults and 69.5% reduction in the emergence of culicine adults. Furthermore, no pupation was observed in treated buckets that were part of a eld bio-assay carried out parallel to the trial. The LF had no negative effect on rice plants or on a variety of non-target organisms, except backswimmers 37 . In another test, a laboratory assessment was conducted to determine LF e cacy against Ae aegypti mosquitoes. At the recommended dosage (1 mL/m2 of water surface), mortality of pupae was 99%. Gravid females also avoided laying eggs in LF-treated oviposition cups. There was no in uence of physical factors on LF's e cacy and no toxic effects on sh and plants 38 .

Personal Repellent
Tabard® Lotion active ingredient is DEET (19.5%), manufactured by Acorn Products Pty Ltd, South Africa. There are numerous studies over many years demonstrating the effectiveness of the active ingredient DEET. This formulated product was shown to be effective under laboratory and eld conditions 39 . A 2002 study showed that topical application of insect repellent can inhibit mosquitos from biting 40 . A study on An arabiensis females showed that treating ankles and feet with a consumer brand of DEET repellent led to a greater than threefold reduction in biting 41 . Similarly, a 2004 paper evaluating the sensitivity of Ae aegypti and An gambiae mosquitoes to DEET found that Ae aegypti were highly sensitive to DEET based repellents 42 .

Fly Swatter
Mechanical device consisting of a piece of semi-rigid mesh connected to a handle. While there are no scienti c studies on the e cacy of y swatters as a vector control method, this inexpensive tool was included in the kit as an additional level of personal protection. The intent was to give users greater involvement in the control of vectors and hopefully increase the desirability of the kits.

Household demographics
Of the 360 HH enrolled for the study, kits were effectively distributed to 350 HH (97%). The number of HH successfully used for data collection totalled 324, comprising 60 HH (19%) for direct observation to document behaviors as the kits were opened and contents were being used, and 264 HH (81%) for interviews after using the kit contents. Most of the responders were uneducated adult females who communicated mainly in the local Somali language. The interviews were conducted in either that dialect or in Swahili as possible. The participants indicated they understood the purpose of the kits they were given and the purpose of most of the components.
The remaining HH were not observed or interviewed for the following reasons: either HH was locked or the household head was absent (n = 11); the HH refused the product (n = 10); the kit was received but not used (n = 4), or the household head declined to be interviewed (n = 1).
The majority of the HH respondents were female (71%, n = 188), while 29% (n = 76) were male. Most of the participants/responders were between ages 20-49 years 62% (n = 163) while 37% (n = 97) were above 40 years of age. Regarding education, 61% of the respondents reported that they had never attended school, 17% had attended primary school, and 9% had attended post-secondary school or college. Overall, 97% of the response could speak Somali language, while 44% could speak Swahili and only 33% could speak the English language. Of the 264 HH interviewed, 83% (220) had another adult present in the HH. Of the 220 adults, 38% had never attended school, 18% had attended primary school, 28% had attended secondary school, and 15% had attended post-secondary school.

Participant understandings
When asked if the purpose of the kits had been adequately explained and understood 97% of respondents con rmed they de nitely or at least somewhat understood. When asked if they understood the purpose of each mosquito control product included in the kit there were also positive responses (Fig. 3).
Each kit contained a booklet of pictogram Directions for Use (DFU) that described how the product was to be put to practical use, safety precautions, and speci c directions where the products should be applied for best performance against Aedes mosquitoes. Most interview respondents indicated positively they understood the pictograms for each product (Fig. 4). This was an important outcome given the educational level of the responders.
The interview responses were con rmed by direct observation of a subset of HH as they handled products in the kits (Table 1). Over 80% of those observed were able to follow the pictogram instructions for adult control products. The larvicidal lm presented a challenge for some users (21%). Fly swatter 89 11 0 When asked why some of the products in the kits were not used the most common reason was they did not know how to use the product (36%) or for what is was to be used (28%). Other rationale for not using a product included they did not think it would be effective (14%), did not know when it should be used (11%), or thought it was unsafe (7%).
Participants were asked to rank each product by di culty to follow the DFU (Fig. 5). Coil pictograms were the easiest to follow and the larvicidal lm was the most di cult. Further, when asked if the DFU pictograms explained how to use each product in a way that was easy to understand, most of the respondents stated that the DFUs were easy to understand.
When participants were asked about their knowledge of the time of the day each product was supposed to be used, as indicated in some of the pictograms, the average response was during morning hours and all were thought to be used before evening hours (Fig. 6). The majority of the respondents used the products in the bedroom. The exceptions were the larvicidal lm, for which 83% of the respondents used the product outside the house, and the personal repellent that was applied to individuals rather than surfaces (Fig. 7).
When participants were asked to list the products from the most useful (1) to the least useful (6), the ndings were analyzed in terms of the proportion of participants that listed the product within the rst three rankings (1-3) as most useful, and within the last three rankings (4-6) as least useful. The trigger sprayer had the highest usefulness rating (90% of responders) followed by aerosol insecticide (84%), larvicidal lm (84%), coils (75%), and personal repellent (70%). Responders rated less useful the interventions with which they were less familiar: insecticidal curtains (49%), spatial repellent (34%), and y swatters (30%).
When asked about types already in the HH, 7% (n = 18) of respondents stated that they had products similar to the trigger sprayer (Dudu Spray) in their HH. Another 25% (n = 22) of respondents stated that they had products similar to the aerosol spray (SUPAkill) in their HH. Mosquito coils and personal repellents were also identi ed as product types that could be found in HH (9% of respondents each). However, participants indicated they did not think those product types came with clear DFU (Fig. 8).
The participants were given the opportunity to suggest other control products that should be included in the kits. The list included insecticidal mosquito nets (74% response), more aerosol insecticide (12%), coils, candles, and emitters (6%), treated mattresses and bedsheets (4%), water treatment tablets or chlorine (3%), and lemongrass oil (1%). When asked to rate the overall value of the kit the response was favourable with 85% rating the Family Vector Control Response Kit as good to excellent (Fig. 9).

Household (HH) knowledge and perception of vector control
When respondents were asked if anyone in the household had ever been infected with an Aedestransmitted disease there was on average an 8% positive response. When pressed if it was dengue fever, chikungunya fever, yellow fever or Rift Valley fever; 12% ( n = 31) of the respondents con rmed that at least one person in their HH had been infected with either Rift Valley fever (3%), yellow fever (9%), dengue fever (6%), or chikungunya fever (5%).
On average, 89% of respondents did not know how diseases were spread by Aedes mosquitoes. Regarding respondents' knowledge of how dengue fever, chikungunya fever and yellow fever are transmitted, only 12% (31/264) of those respondents had correct knowledge of how dengue fever is transmitted, followed by correct knowledge about the transmission of chikungunya fever (9%); yellow fever (9%); and Rift Valley fever (15%).
Regarding respondent's knowledge of mosquito breeding sites around the household, more than half (56%) were not able to identify breeding sites. The other 44% (115/264) of respondents had correct knowledge of mosquito breeding sites around the house; stating that stagnant water, latrines, wells, water containers and bathrooms can act as breeding sites for mosquitoes.
When asked how often mosquito control products were used, 51% of the respondents (134/263) indicated daily or at least more than once per week. Others indicated they use only when mosquitoes are present (35%) or only during the rainy season (14%).
Most respondents (87%) listed mosquito nets as the most common method of mosquito control (229/264). Other control methods included spraying (5% of respondents), coils (3%), or smoke, candle, and brush clearing (5%). When asked about speci c types of control over half the respondents had used smoke or candles (51%) and a few (4%) had experience with plug-in emitters.
More than half of the respondents indicated they purchase mosquito control products from either a kiosk (52%) or a supermarket (34%). Other places where such products were purchased included pharmacies, small stores, or from the street vendors. There is a clear indication from this brief survey of current practice that respondents were familiar with available mosquito control products.

Discussion
The complex setting in Wajir is representative of factors that affected disease risks and operational disease control challenges in many complex humanitarian emergency settings today. Innovative solutions are urgently needed to help ll the gaps in VBD protection for families that are otherwise put at great risk. Furthermore, effectiveness of the core existing vector control tools and strategies for VBD prevention, based on killing mosquitoes when they seek to bite people sleeping in their homes at night, such as long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS), is increasingly challenged by insecticide resistance. These tools are also unsuited to prevent Aedes-borne diseases because this mosquito bites during daytime, inside or outdoors, and often rests outside 43,44 .
This study represents the rst of its kind to consider putting tools for vector control immediately in the hands of HH at initial alert of an outbreak. Evidence-based Family Vector Control Response Kits that can quickly be deployed and distributed, that are easy to use by HH, and that are effective against mosquitos could potentially ll the gap in disease prevention when families are affected by con ict, natural disasters, or early stage epidemics, whilst awaiting larger scale and longer lasting disease control interventions. Despite the modest scope of the study it produced important results on the feasibility of Family Vector Control Response Kits being acceptable and usable by householders. The potential needs to be explored for reducing the risk of vector-borne disease transmission during gaps in response to natural disasters or humanitarian emergencies that include outbreaks of mosquito-transmitted diseases.
The main objective of the Family Vector Control Response Kit user acceptance study was to determine if different combinations of vector control tools in the kits would be used appropriately and accepted by HH when given with only visual instructions. The products were those commonly found in retail markets. The graphic use directions (pictograms) were customized by MENTOR Initiative to demonstrate application techniques speci c for control of Aedes mosquitoes, as opposed to the general instructions commonly found on labels, often in text that cannot be translated by users.
Participants did not have much awareness of key diseases spread by Aedes mosquitoes nor basic facts about biology and means by which transmission occurred. However, they did regularly engage in mosquito control in their houses. It is assumed this was mainly to control nuisance mosquitoes such as Culex spp. HH did rely upon mosquito nets as the most common method of control but also used consumer sprays, coils, and smoke. The products were purchased from kiosks, supermarkets, pharmacies, and small vendors 52 . This behaviour demonstrated a clear awareness of the risk or irritation caused by biting insects and the HH desire to reduce biting insects in their homes. It explained why the acceptance and the generally correct usage of the vector control household kits, and its different tools, was so high across all arms.
A principal nding of the study was the con rmation that the pictogram DFU for most of the product types were understood and followed by most participants. This is an important validation since these directions did not require translation or education to be understood by the users. The pictograms can also be used across a range of products for each product type thus expanding utility in most countries even as product availability varies.
Participants indicated the usefulness of different products in the kits varied in ranking. Some trend related to di culty of following the respective DFU may have in uenced the ranking. If that was the case, improvement in perceived usefulness may be possible with revised pictograms. Alternatively, kit contents should be revised and some components dropped if the value cannot be con rmed. However, when asked to rate the kits the majority of participants gave it a good to excellent value rating.
This study was not designed speci cally to evaluate mosquito control by this integrated approach. That undertaking would require larger scale and attention to key environmental factors that affect increases in population densities 45  None of those studies combined adult knockdown aerosol with targeted surface spraying, prevention of re-entry with insecticidal curtains, and larval control applied by HH members with the intention of creating a mosquito-free environment combined with personal protection. This study represents the rst attempt to provide a kit of vector control tools to HH for them to use prior to more formal action by the community or government. The focus of the trial was to document the ability to follow pictorial application instructions that are not usually included with products obtained from local markets. The theory is when products are appropriately applied the HH population of mosquitoes can be controlled or eliminated before further transmission occurs.

Conclusions
The present research validates that pictogram use directions can be understood and followed by users with low rates of literacy. There is no need for translations as carefully designed pictures allow for high rates of adoption for the intended use without need for a comprehensive educational campaign. This is a rst step toward the goal of providing Family Vector Control Response Kits that can be rapidly deployed at rst signs of outbreak and for use after natural disasters when normal supply chains and operations are disrupted. Further steps include con rming the e cacy of reducing indoor mosquito densities with positive impact on disease transmission, gaining global buy-in to using HH vector control kits for rapid large scale emergency response, evaluating kits in more countries, improving the picture diagrams, and re ning key contents to most essential products.
As consumers grow more familiar with bene ts of the kits, their perceptions of usefulness and ability to interpret the instructions are expected to increase. A comprehensive education campaign, conducted without the pressure of an immediate outbreak, could also improve adoption over time. This would t with the premise that supplies will arrive before a more formal, comprehensive campaign gets moving.
This is important because innovation in the space of rapid, household vector control should lead to userfriendly, less expensive interventions.   Interviewed participants (%) understanding of the pictogram Directions for Use (DFU) for each product in their kits.
Page 26/28 Participants interviewed (%) who ranked product Directions for Use (DFU) easier or harder to follow.

Figure 6
Time of day when respondents, on average, thought mosquito control products should be used.

Figure 7
Household locations where products were used, as identi ed by interviewed participants (%).

Figure 8
Participants (%) who thought products they used before the study came with clear Directions for Use (DFU).