Study design
This study investigated the efficacy of the FTPE and BGS in a push-pull system to reduce human-landing rates compared to the control (no intervention). We also determined if the combination of FTPE and BGS was better than either FTPE or BGS alone whereby the following treatment arms were compared: 1) two FTPE versus negative control 2) BGS trap versus negative control 3) the combination of FTPE and BGS versus negative control. The study design was a randomised block design over 16 days per treatment arm. Each intervention and its control were assigned to one of two separate compartments in the semi-field system (SFS) for a block of four days, after which the treatment and its control were switched between compartments. Preliminary experiments showed that removing FTPE immediately after experiment and airing the compartment for 20 hours was enough to prevent any residual effect of the transfluthrin. In each block of four days, four volunteers rotated daily between chambers.
Study Site
The experiment was conducted in the SFS located in Bagamoyo-Tanzania, between January 2018 and December 2018. The SFS measures 21 x 29 x 4.5 m with two compartments (21 x 9 m), separated by a corridor. A heavy-duty polyethene wall separates these compartments preventing air movement between the chambers and reducing any chance of cross-contamination when working with spatial repellents or other aerosols. The SFS allows for controlled experiments with set densities of disease-free mosquitoes to be conducted under field-like climatic conditions throughout the year [20].
Mosquitoes
Laboratory reared, pyrethroid susceptible Ae. aegypti (Bagamoyo strain) were used. Susceptibility bioassays performed prior the implementation of the experiment following the World Health Organization (WHO) guidelines [21] showed that mortality of these mosquitoes was > 99% after exposure to all the pyrethroid insecticides tested (Deltamethrin (0.03%), Permethrin (0.25%) and Alpha-cypermethrin (0.03%) . These mosquitoes were colonized from Bagamoyo in December 2015. Larvae were fed on Tetramin® fish food and adult mosquitoes on 10% sucrose ad libitum and cow blood meals (heparinized) were given to adult females for egg production using a membrane-feeding assay. The colony is maintained at 27±5°C and 50-99% relative humidity.
For this experiment, 3–8 days old female mosquitoes, previously unfed with blood were used. These mosquitoes were sugar-starved for 12 hours before the start of experiments. Female mosquitoes that were responsive to human odour were selected from three different rearing cages and transferred to small releasing cages. Selection from cages was done by placing a hand close to the cage and choosing only aggressive host-seeking mosquitoes.
Preparation of the freestanding transfluthrin passive emanator (FTPE)
We designed a device that can easily be placed anywhere in the peri-domestic space (Figure 1 A-E). The emanator passively releases transfluthrin vapors into the surrounding area through evaporation. The device is a stool-like structure that supports hessian strips (made from plants of the genus Corchorus olitorius or Corchorus capsularis also called jute, burlap or gunny sacks), treated with the emulsifiable concentrate (EC) transfluthrin active ingredient (Bayothrin EC, Bayer AG Monheim am Rhein, Germany) as the push. The hessian fabrics were chosen as they have been shown to retain transfluthrin active ingredient for up to six months due to their high cellulose content [11, 15]. The hessian fabric were locally bought, washed with OMO® detergent powder (Unilever Kenya Limited, Kenya) and dried under direct sunlight. The fabrics were cut into strips with a surface area of 0.5m2 (10cm x 5m) and treated with 5.25g of emulsified concentrate transfluthrin and left to dry under the shade in the SFS (Figure 1B-C) to prevent photolysis of transfluthrin. The strips were then wound around a pole into a spiral and sealed with outer wire mesh to prevent access to the treated hessian ribbon by children or animals (Figure 1 D). Two FTPE with a total of 10.50 g (5.25g each) of transfluthrin were used per experiment.
BG sentinel trap
The BGS (Biogents AG, Regensburg, Germany) has been widely used as the standard trap for collection of adult Aedes mosquitoes [22, 23]. The BGS was used with a Biogents-Lure (BGL) and carbon dioxide as a pull. The BGL is a synthetic lure consisting of lactic acid, caproic acid and ammonium bicarbonate dispensed via granules [23]. It is effective for five months however, for the purposes of this experiment a new lure was used for each experimental round of sixteen days. Carbon dioxide was released from a pressurized cylinder at the rate of 500 ml/min, using acrylic gas flow meter (Hangzhou Darhor, Technology Co., Limited, China).
Procedure to determine the protective efficacy of the FTPE and odor baited trap
To simulate the peridomestic setting, human landing catches (HLC) were performed with a volunteer sitting 2 m from an experimental hut inside the SFS (Figure 2 A-C). For the “push” alone evaluation, two FTPE were positioned six meters apart with human volunteer sitting in between them conducting HLC (Figure 2A). During the “pull” alone evaluation, the BG-sentinel was placed 10 m away from the HLC (Figure 2B) as this exceeds the maximum distance over which mosquitoes select between hosts so that the action of the trap could be measured independently of the effect of the human collector [24]. For the “push-pull” evaluation, both FTPE and the BGS were used and positioned as described above in the “push” and “pull” set ups (Figure 2C). In the control, untreated emanators and HLC were used.
The FTPE were positioned in the experimental chambers forty-five minutes before the experiment started to allow release of active ingredients into the experimental space, and mosquitoes transferred to the buffer chamber (corridor) of the SFS one hour before the experiment began to allow for acclimatization. During the acclimatization process, mosquitoes remained free from transfluthrin exposure. After the acclimatization period, cages with approximately 25 mosquitoes each were positioned in the four corners of both compartments (approximately 100 mosquitoes per compartment/treatment). Mosquitoes were released at 07:00hrs by a gentle pull of the strings connecting the releasing cages and the chair where volunteers were sitting. The experiment was conducted from 07:00hrs to 10:00hrs to reflect natural biting time for Aedes mosquitoes [25]. As preliminary experiment showed that after 10 hrs > 90 % of the mosquitoes were already collected.
The volunteers conducted HLC, collecting mosquitoes that landed on the area between the ankle and the knee for three consecutive hours. All volunteers were males aged between 25-40. They were non-drinkers, non-smoker and were asked not to apply perfume, bathe using perfumed soap, before the experiments. During the experiment, volunteers wore shorts, covered shoes, and bug jackets to standardize the area available for mosquito landings. Mosquitoes were recaptured continuously for 50 minutes using a mouth aspirator. After 50 minutes the volunteers would take a break for 10 minutes, after which a new paper cup labeled with time and date were used. Collected mosquitoes were transferred to the insectary for sorting. After the experiment, mosquitoes that were not collected during the HLC were recaptured using Prokopack aspirators and killed by freezing to prepare the SFS for the next day’s experiment. A Tinytag® view 2 data logger (model TV- 4500, Gemini data logger, United Kingdom) was placed inside the SFS throughout the experiment to record temperature and relative humidity.
Experiment to assess the longevity of the FTPE
To assess the longevity of FTPE protection, the devices were evaluated at zero-, three- and six-months post impregnation. The same set up as described previously (Figure 2A) was followed. Between the evaluations, the emanators were “field aged” by storing the FTPE in an outdoor environment under a tree in the shade to simulate aging on a verandah of a house, i.e. placed outdoors under ambient conditions, protected from direct sunlight and rain (Figure 1E).
Sample size
Sample size calculations were performed using simulation-based power analysis [26] in R statistical software version 3.02 http://www.r-project.org with significance level of 0.05 for rejecting the null hypothesis. Data analysis for experimental data was conducted using generalized linear mixed models (GLMMs). Therefore, one thousand simulations of generalized linear mixed models approximating those that will be used to analyze project data were run using the same experimental design. The power to predict the difference in mosquito landings between control and treatment was estimated as the proportion of the 1000 simulated data sets in which the null hypothesis was rejected when the generalized linear mixed model was run. Parameters were set at 10% estimated variability between chambers, 10% variability between mosquito releases and 10% variability between volunteers. Simulations indicated that with an estimated 100 mosquitoes released per night and 60% recapture of released mosquitoes in the control, there was 94% power (95% CI: 92 – 96%) to detect a 50% reduction in mosquito landings in the treatment arm after 16 nights of experimentation. Furthermore, there was 70% power (68% CI: 74 – 72%) to detect a 15% difference between the treatments.
Data analyses
Data were entered in Microsoft Excel 2010 and analyzed in Stata 13 (Stata Corp). The data were analyzed to determine the efficacy of each intervention (push alone, pull alone and push-pull) to reduce the human landing rate compared to the control. The mean percentage recapture and confidence intervals were calculated for each intervention and control. From this daily protective efficacy was measured by comparing the human landing rate on a volunteer with the intervention to the negative control using the following formula and the overall arithmetic mean PE and 95% confidence interval for the experiment was calculated;
Protective efficacy= [(C-T)/C] x100%.
Where C stands for the number of mosquitoes landing in the control and T is the number of mosquitoes landing in the treatment.
The effect of each intervention was determined by fitting a generalized linear mixed model (GLMM) with a binomial distribution and logit function. The binomial distribution was chosen, as the dependent variable was the proportion of recaptured mosquitoes out of those released. Independent fixed effect categorical variables were treatment (intervention or control), compartment, volunteer, block (push, pull, push-pull) and an interaction term between treatment and block. Day was included as a random effect.
To determine the longevity of FTPE across six months after impregnation; a GLMM with a binomial distribution and logit function was also used. For this model, the dependent variable was again the proportion of recaptured mosquitoes. Independent fixed effect categorical variables were treatment (intervention or control), compartment, volunteer, month of testing (month 0, month 3, month 6) and an interaction term between treatment and months. This interaction was used to determine if the protective efficacy of FTPE changed between months. Day was included as a random effect.
During the evaluation, there were no significant association between humidity and temperature on the proportion of recaptured mosquitoes p=0.705 and p=0.203 for the humidity and temperature respectively. Therefore, these variables were not included in the GLMMs.