Acceptability and effectiveness of a monofilament, polyethylene insecticide-treated wall lining for malaria control after six months in dwellings in Vhembe District, Limpopo Province, South Africa

The ITWL mesh was produced on a commercial line at the Huhthamaki factory in Springs, Gauteng Province, South Africa. Four monofilament, polyethylene mesh formulations/treatments containing deltamethrin and alpha-cypermethrin actives were extruded using polyethylene as base, and the pyrethroids were introduced through masterbatches [17]. Different insecticidal treatments were tested in order to determine the most suitable formulation. Both deltamethrin and alpha-cypermethrin are listed as WHO Class III pesticides. They are deemed within the normal safety range for use during vector control and are WHO Pesticide Evaluation Schemes (WHOPES)-recommended insecticides for IRS and for mosquito nets [18]. These two pyrethroids appear to be less excito-repellent than permethrin, which is also used in LLINs. A resting mosquito would potentially be in contact with the nets for longer thus permitting a lethal dose uptake from the net surface [19]. Plain, untreated mesh was also produced for use as a negative control. The different formulations are presented in Table 1. The final mesh products were ca. 420 mm wide and 700 m in length with a mass of ca. 80–120 g/m². After production all five mesh types (treated and untreated) were put through bioassays to determine efficacy. The different mesh types were stored in large polyethylene bags for about 10 months at room conditions at the Department of Chemical Engineering, University of Pretoria, before installation in community dwellings. Linings were stored after production whilst study logistics (including study site selection, negotiations with local traditional leaders and ethical approval) were finalized and the start of the malaria spray season was anticipated.

Malaria transmission is seasonal in South Africa from October to May and spraying for vector control occurs annually prior to the start of the malaria season. Transmission by Anopheles funestus and Anopheles arabiensis peaks during the rainy period from December to February. Rural villages in the northern and eastern parts of Vhembe District are located in a medium- to high-risk malaria area [13]. A total of 1398 malaria cases were reported in Limpopo Province from October 2012 to end March 2013 and 754 (53.93 %) of these cases were from the Vhembe District. However, the malaria incidence in neighbouring countries is a major concern.

The study was carried out from October 2012 to April 2013 in the community village of Tshilivho, which is only partially sprayed during the annual spray programme. The study focused on huts and homes in the unsprayed section of the village in order to avoid interfering with the annual IRS programme. The housing in the village (Fig. 1) consists of traditional Venda homesteads, which consist of a number of round thatch-roof huts, 3–5 m in diameter, built in a compact circular arrangement. The walls are a mixture of mud and cement, while the floors consist of mud and/or cow manure or cement [13]. The number of traditional homesteads is declining with more modern (western style) homesteads emerging. The majority of these modern homesteads are brick and cement houses built by the South African Government’s Reconstruction and Development Programme (RDP) redress policy initiative. These houses feature corrugated iron roofs. Privately built modern houses also feature corrugated iron and, in some cases, tiled roofs. More often one or two traditional mud huts still form part of the homestead, but these are mainly used as storage units or for cooking purposes, often over open fires, when it rains. Water for drinking, cooking, bathing, and washing, collected from community taps (borehole water), is usually stored on the property in large plastic containers before use. The plastic containers are often kept inside mud huts and are potential breeding habitats for mosquitoes. All participants had electricity in their homes, with the exception of one participant living in a hut.

Participation in the study was voluntary. The necessary permission was obtained from the Department of Health and Social Development of the Limpopo Provincial Government and local tribal authorities to conduct the study. The study was introduced to the community during two civic meetings. Written informed consent was sought from participants and this included consent to publish. Participants aged 18 years or younger (two in this study) had to sign an assent form and their legal guardians had to sign the consent form. The consent forms were translated into Tshivenda, the local language, to ensure that participants understood all aspects of the study and their rights to privacy, refusal to participate or to withdraw. Participants received an incentive in the form of pre-paid electricity vouchers to the value of R80.00 (~10USD in 2012) to compensate for the use of electricity during the installation of the linings and for the participants’ time when lining samples were collected during interviews in month 1, 2, 4, and 6. Ethics approval was obtained from the Faculty of Health Sciences Research Ethics Committee at the University of Pretoria.

Forty households (20 traditional mud huts and 20 western-style houses), as determined by a biostatistician, participated in the study. The lining was installed in one bedroom in houses or in the sleeping hut (Fig. 2) of the household (henceforth termed study areas), by students from the University of Pretoria. Some of these sleeping areas served as cooking and living areas too. Each lining treatment was randomly installed in eight study sites (four huts and four houses). Installation occurred along the inside wall of the study area right below where the roof and wall met. Anopheles arabiensis tend to rest more often on the upper section of walls after a blood meal. Flexible wooden strips were mounted to the wall with screws at the top and about 40 cm lower down the wall to form a frame. The lining was attached to this frame by Velcro^® tape. The installation process was explained in detail to participants beforehand, and they were present during installation. The term participant was assigned to the person who normally slept in the study area and who was to be interviewed during scheduled visits over the 6 months. If he/she was not available then a suitable representative was identified. No instructions were given to participants on cleaning or removing wall linings. The people had to go about their daily routines without any influence from the researchers on how to ‘live’ with the new addition to their homes.

All data collection was conducted through interviews with questionnaires translated into TshiVenda. One trained VhaVenda field assistant conducted all interviews from informed consent of participants to the final summary interview to ensure comparable data. The assistant translated comments from participants into English. The principal investigator moderated all questionnaires to ensure they were completed in full and to query potential errors.

Prior to installation the participants completed an enrolment questionnaire that served as a baseline for future interviews. The questionnaire focused on important features of the study area: number of people (including children) that normally sleep in the study area, other purposes of the area, time spent in the area, animals and insects that enter the area, pesticide usage, methods to repel mosquitoes, spray status of the homestead, and malaria incidence in the respective households.

The acceptability, durability and the perceived effectiveness of the linings were assessed monthly. The monthly questionnaire focused on the use of pesticides for vector control, observations on the numbers of mosquitoes and insects in the study area, any damage, colour change or loosening of the linings over the month, and any observed effects on the people and animals that may have been related to the linings. The condition of the wall linings was physically evaluated during each interview.

A final, summary questionnaire was completed a week after the last monthly questionnaire: the participant’s overall perception towards the linings, attachment methods and materials, positioning of linings, the efficiency in controlling irritating insects and mosquitoes specifically; colour preferences for future products and any suggestions or complaints were assessed. Participants were encouraged to express their opinions towards the linings during all interviews.

Lining durability was not only physically assessed but also subjected to entomological residual efficacy testing in the laboratory. Sample strips (100 mm wide) of each lining were collected on month 1, 2, 4, and 6 post-installation. Samples were placed in individual A4-sized, zip-lock type, plastic bags, marked and sealed. Analysis was done at the National Institute for Communicable Diseases (NICD) in Johannesburg, South Africa. Residual efficacy was determined by analysing knockdown (KD) and mortality rates of mosquitoes through WHO-recommended laboratory-scale contact or cylinder test for IRS [17, 20]. One cylinder bioassay replicate was done per lining (40 linings), with eight cylinder bioassay replicates per treatment (five treatments × eight = 40 linings). Each bioassay was carried out using 25 susceptible (to all insecticides) non-blood fed, 2–5 day old laboratory-reared female An. arabiensis mosquitoes KGB colony (standard strain used in testing) housed at the NICD in Johannesburg. Bioassays included blank controls where no mesh was placed in the test cylinders. Mosquitoes were exposed to the samples for 30 min. KD was recorded 30 min after the end of initial exposure and mortality was recorded 24 h after initial exposure. After exposure the mosquitoes were given sugar solution as nourishment. Efficacy was measured against the WHO efficacy criterion for IRS of a minimum of 95 % KD and a minimum of 80 % mortality post exposure.

The sample size (n = 40 study areas) can be regarded as sufficient since it meets with the agreement to have at least 30 degrees of freedom for the error term in the analysis of variance (ANOVA) with response to the variable ‘slow release pyrethroid’ concentration. The study was conducted as a recurring measures factorial experiment. Quantitative data from questionnaires were entered onto an Excel spreadsheet. Acceptability criteria data (colour preference, position of the linings, etc.) and perceived efficiency were exploratory and descriptive statistics were used as required using proportions and means. Qualitative data analysis was conducted following a partially grounded theory evaluation, which allows repetitive theoretical generalizations to emerge from the data [5]. No codes were used but patterns in data were grouped together and interpreted.

Out of the 40 participants, 82.5 % (33/40) were female and 17.5 % (7/40) were male. Participant ages ranged from between 16 and 78 years of age at the start of the study. Fifteen per cent (6/40) of the participants, all living in huts, indicated that they had no knowledge of malaria. Only two people, both staying in huts, had ever been diagnosed with malaria: one participant’s wife (in the year 1967) and another participant’s husband (in the year 1989). One trial house (1/20) had ever been sprayed by DDT spray workers for malaria, whilst 35 % (7/20) of the study huts had been sprayed before, but all less than 6 years ago. However, not all participants had spray cards available and the information is based on what the participant could remember if they had knowledge of spray status. Two participants (2/40) possessed one untreated bed net each that had been used before. The bed nets were not used during the study.

Some important features of the housing structures that can impact on mosquito access or the effectiveness of IRS insecticides are presented in Table 2. Roof material can impact on the number of mosquitoes drawn to a dwelling. The presence of windows with glass and presence or absence of a ceiling can impact on mosquito access to dwellings. Gaps between the roofs and tops of the walls (eave gaps) that were larger than 2 cm were present in 95 % (19/20) of the traditional huts and 15 % (3/20) of the houses. The gaps allow for easy access routes in and out of houses and could potentially impact on the number of mosquitoes in the homes. Coverage of the inner walls is important due to the re-smearing of hut walls with fresh mud and cement every year or every second year.

On average, 2.6 (range 1–8) people slept in a study area and 75 % (15/20) of huts and 50 % (10/20) houses had children under the age of 13 that slept there. Nineteen of 20 participants in houses indicated that they only slept in the room where the lining was to be installed versus 85 % (17/20) of participants in huts. The remaining 5 and 15 %, respectively, cooked in the study area, and this percentage changed over the study period especially in the huts over the rainy season.

Participants were asked for the time (on average) that they normally went to bed and the time that they got up in the mornings. This was done to determine the range of time when the participants would be inside the dwelling under the protection of the linings (Table 3). Majority of the people went to bed after 21:00 h and majority woke up between 06:00 and 07:00 h. The time does not indicate when they entered the sleeping areas and closed the door for the night or when they exited the area in the morning to start the day.

Mounting of the frames and lining installation as described, took longer than anticipated. Challenges included hard walls that affected hole drilling for the screws to mount the frame. The reason for using flexible wooden strips was to compensate for the circular shape of the traditional huts and the ease to remove or change the lining if this was required during the study. Two to three people were needed to install the wall linings because the mesh had to be cut to fit the area’s dimensions. Furniture and containers filled with water had to be moved away from the wall, but no permanent re-arrangement of furniture occurred. All participants indicated that they were pleased with the installation process and they had no problem with the people who did the installation. Three of the 40 participants commented that the process took longer than was explained to them.

One month post-installation, all ITWLs were still in position and all the participants expressed their approval of the linings. During the 1 month post-installation interview one participant, from a house, indicated that there was a bad smell for a couple of days after installation, but no negative effects (headache, nausea, etc.) were experienced. The source of the smell was not identified. No further bad smells were noted as the study continued.

Participants noted that the number, bites or irritation of mosquitoes had decreased since lining installation.

A decrease in other nuisance insects, including cockroaches, ants, flies, in some instances, termites and summer flyers (type of month), and also spiders were noted.

Observations were based on the number of dead mosquitoes and insects on the floor, dead on the furniture, dead when cleaning in general, bites, and irritation. The scale of decrease (a little, more than a little or a lot) is presented in Table 4. A diagram consisting of three squares with different numbers of mosquito images per square was used to standardize responses. Similar responses to the first month were noted at the second month post-installation interview.

During the third month, high rainfall occurred for more than 2 weeks causing floods in parts of Mozambique and in the Kruger National Park, about 55 km from the study village. All linings were still in position and still effective according to the participants. Since the previous interviews, one participant had moved out of the hut with the lining and started using it as a cooking area (over an open wood fire).

At 4 months post-installation, participants reported that there was an increase in mosquito irritation and biting even in the study areas, possibly due to an increase in mosquito populations after the heavy rains.

One participant’s cooking hut collapsed due to the rain and she started cooking over an open wood fire in her sleeping hut. Water damage was observed in her sleep hut, leaving large sections of the wall damp and parts of the lining came undone. Brown water marks on the white Velcro^® were visible (Fig. 3). Holes were noted in the lining and the participant mentioned that she saw rats climb on the lining (Fig. 3). Rats entered the huts of other participants (Table 4) but no other linings were damaged.

At 5 months post-installation mosquito irritation and biting was still noticed, but was lower than the previous month. One participant decided to start sleeping in a different room in her house but mosquito irritation was too high and she moved back into the study area a week later.

After 6 months, 98 % (39/40) of the ITWLs were still in position. The lining that had come undone due to water damage did so again and the Velcro^® was replaced.

Over the study period, from enrolment to 6 months post-installation, a pattern was noticed in the mosquito irritation whilst sleeping (Fig. 4) compared to the use of insecticides and the burning of mosquito coils and other materials as repellents (Fig. 5). At the start of the malaria season, prior to lining installation, irritation was higher in huts where 70 % (14/20) of the participants were irritated whilst sleeping versus 40 % (8/20) of participants in houses.

Thirty per cent (6/20) of participants in houses used insecticides to kill mosquitoes and other annoying insects compared to 45 % (9/20) in huts. The burning of mosquito coils to repel mosquitoes was done in 40 % (8/20) of the houses and 60 % (12/20) of the huts, respectively. Circumstantial evidence exists that smoke is an effective insect repellent [21] and the burning of items such as dung, toilet paper, egg cartons, or plant material to produce smoke to repel mosquitoes is often practised by the VhaVenda people. Overall, the participants in huts burnt more of these other materials to repel mosquitoes pre-installation.

Within the first month post-installation a drop in mosquito irritation (based on perception) was noted by 25 % (5/20) of participants in houses and 15 % (3/20) in huts (Fig. 4). At the same time, decline in the use of insecticides and mosquito coils was noted (Fig. 5). An increase in mosquitoes as the season moved on is noted in the slight increase of irritation reported by the participants at 2 months post-installation. A similar trend was noted at 3 months post-installation. A drastic increase in mosquito irritation was noted at the fourth monthly interview where 65 % (13/20) of participants in houses and 95 % (19/20) in huts were irritated whilst sleeping. An increase in insecticide and mosquito coil usage was noted at this point including the burning of other materials. Irritation was still high in the fifth month but already lower than the previous month. However, a drastic decrease in irritation was noted over the last month of the study, with 10 % (2/20) of the participants in houses and 15 % (3/20) in huts indicating some irritation by mosquitoes whilst sleeping. Insecticide and coil usage was almost non-existent during the final month.

None of the participants indicated a dislike of the ITWL over the study period. The final questionnaire offered participants the chance to give their overall feelings regarding the linings. All participants responded in the positive to all lining features and almost no suggestions to alter these were presented.

There was no colour change noticed in any of the linings over the trial period except for dust coverage that made white (untreated) linings appear brown. All the participants were pleased with the colour of the lining that was installed in their homes. Reasons for the positive response towards lining colour varied (Table 5) but not all participants supplied reasons. Eight of the 40 participants indicated that the linings were decorative.

Participants were in favour of the original positioning of the ITWLs (Table 6). Two participants appreciated that the lining was placed out of the reach of children.

Two participants indicated a preference for the linings to be lower down the wall, whilst one wanted the lining to cover his bed.

These two participants had no children under the age of 13 that slept in the study area. Therefore, lining damage or children’s health was not a concern. Surprising though is that 65 % (13/20) participants in the traditional huts indicated that the lining could cover the entire wall, in spite of 75 % (15/20) indicating that young children stayed with them.

All participants indicated that they were pleased with the attachment method of the ITWL. A list of alternative attachment methods were presented to the participants and each was allowed to select more than one preference. The preferred alternative method was capped nails, and screws were a close second option (Table 7). Glue and staples, both more permanent methods, ranked high with participants in huts at 50 % (10/20) and 55 % (11/20), respectively. This was less in houses at 30 % (6/20) and 35 % (7/20), respectively. No suggestions for other methods were presented, but comments were offered on the method utilized in the study. Four participants (4/40) indicated that the frame covered holes after installation, and 8 % (3/40) of participants said the attachment method was decorative.

In the final questionnaire participants were asked if they thought the decrease in mosquitoes and other insects was due to the ITWL. The answer was a unanimous yes. A follow-on question enquired as to what happened to the number of mosquitoes and insects over time. Fifteen out of 20 (75 %) participants from huts and 85 % (17/20) from houses responded that the numbers kept on decreasing. None of the participants responded that numbers kept increasing or stayed the same. Eight out of 40 participants, 25 % (5/20) from huts and 15 % (3/20) from houses, responded that they noticed a decrease in mosquito numbers at first, then an increase for a short period, followed by a decrease again. The eight participants were requested to select options from a list to explain why they thought this had occurred. All selected that there were too many mosquitoes due to rain, and some also selected that there were usually more mosquitoes later on in the season (Table 8). No participant indicated that they thought the linings stopped working. After the trial of 6 months was completed, participants were asked if they preferred to retain the ITWL for longer.

Efficacy was determined as part of lining durability. The WHO effectiveness criterion is mortality exceeding 80 % after 24 h following a 30-min exposure of An. arabiensis mosquitoes to test samples. KD is used to indicate the bio-availability of the insecticide. All ITWL samples achieved 100 % KD and 100 % mortality right through to 6 months post-installation in study homes (Fig. 6). Samples of the water-damaged lining, soot and dirt-covered linings (especially after 6 months) still resulted in 100 % KD and 100 % mortality.

Some of the participants with untreated ITWL indicated that mosquito numbers, biting and irritation had decreased as was reported by participants with treated ITWL over the 6-month period. At first this was suspicious: participants were experiencing a placebo effect or reporting what they thought the researchers wanted to hear. However, results from the bioassays coincided with what participants perceived as a decrease in mosquitoes most likely due to the linings. KD and mortality was recorded in some of the untreated samples that should not have occurred, but this was observed at levels below WHO criteria in all the untreated samples. There exists a possibility that the untreated roll of mesh was contaminated whilst it was stored with the four treated mesh types. The initial bioassays after production resulted in no KD or mortality in the untreated mesh. The blank controls (no mesh in cylinders) resulted in no mosquito mortality during any of the post-installation efficacy analyses.

These types of preliminary studies are important to determine user preferences towards a new product’s development and design [5]. There are a couple of issues and recommendations to consider in future studies. The ITWLs were installed in a section of the village that is not readily affected by malaria. The linings demonstrated their efficiency against mosquitoes in general along with other annoying insects, and they were well received by participants. However, in spite of this, the overall purpose of such an intervention is to effectively prevent malaria transmission and not look decorative in a user’s home. Before WHOPES evaluation is considered, further testing should be done with the functionality of the lining in mind. Experimental hut trials should occur to determine efficacy with regard to preventing mosquitoes from entering huts and having access to a blood meal. This will present a better understanding of ITWL transmission prevention capabilities. Epidemiological and entomological surveys should be done.

Alternative methods for faster and easier installation should be researched. This will enable community members to remove and re-install ITWL themselves as needed, especially when re-smearing of huts, painting of walls or cleaning of linings occurs. An economic feasibility study should be completed to determine the production and implementation costs and to compare the annual cost against that of IRS as vector control programmes.

The lining acceptability and durability should be looked at over a longer period. Factors such as dirt build-up over time will alter the aesthetics that in turn may affect lining acceptance and user compliance. The effectiveness over a longer period should be analysed to determine how long ITWL remains functional in the homes of users. At the end of the study participants were presented with the option to retain the linings in their homes for longer and all participants opted to do so. Permission was obtained to collect samples annually for at least 3 years for further efficacy testing (until end 2015). The effect of washing on ITWL efficacy should also be determined, especially over the long-term period.

Pyrethroids are currently the only insecticides approved by the WHO for use on/in ITNs or LLINs and also ITWL. However, vector pyrethroid resistance is of major concern and alternative insecticides should be tested for use in ITWL. Researchers at the University of Pretoria are already investigating alternative insecticides to use in ITWL.