Control of Aedes Breeding Sites With a Dual-action Insecticidal Coating is Effective in Reducing Dengue Transmission: Cluster-randomised Controlled Trial Study

1. Introduction. Arboviral diseases such as dengue, Zika and chikungunya (DZC) - transmitted by Aedes mosquitoes - are spreading rapidly around the world. Without mass availability of vaccines or specific treatments, vector control strategies are crucial to reduce transmission and prevent such diseases. Unfortunately, insecticide resistance, the high cost of control alternatives using modern technology and low community participation are major challenges for health programs. There is an urgent need to develop new effective, acceptable and sustainable interventions to reduce the incidence of DZC. In alliance with the Center for Medicine and Society in Freiburg, Germany ("Zentrum für Medizin und Gesellschaft" -ZMG) and Institute of Public Health, Gothenburg University, researchers from the National Institute of Health together with the Vector Control Group of the State Institute of Health (IDS) and GIGA group of UFPS have applied and validated in water containers, a novel vector control method called "insecticide coating (IC)" developed by researchers from the Spanish company INESFLY, under the framework of a research project that seeks to combat Aedes mosquitoes in the metropolitan area of Cúcuta, Colombia. The IC is an aqueous solution containing polymeric microcapsules of insecticides and insect growth regulators (pyriproxyfen-PPF (0.063%) and alphacypermertrin-ACM (0.7%) in suspension, without interaction between them. Its technology allows a gradual release, thus achieving less harmful products, with a high residual effect and high persistence, which makes it effective for the control of larvae and adult mosquitoes.
2. Place of study. The study was conducted in Metropolitan Cucuta in the North-East of Colombia, including the city of Cucuta (629,414), and two adjacent municipalities Villa del Rosario (93,735) and Los Patios (81,411 inhabitants) located in the Norte de Santander state.
3. This study consists of five phases. 3.1. The first phase consisted of preliminary activities, which included socializations with Colombian health managers, project referents and community leaders. A safety evaluation was carried out at one workstation by studying the leaching of alphacypermethrin and pyriproxyfen in tanks of various surfaces treated with the IC product (plastic, metal, cement-tile). In this evaluation, the dose additivity model was applied (samples were processed at the Liverpool School of Tropical Medicine, UK by Dr. Mark Paine's laboratory).

3.2. The second phase was the experimental evaluation of the insecticidal effect of the IC on Aedes, under laboratory and semi-field conditions developed by the Entomology Group of the National Health Institute and the Entomology Unit of the State Health Institute (IDS) of Norte de Santander. Bioassays were designed to determine the following effects of IC on Aedes states: Lethal effect and sterilising effect on gravid females; Effect on egg hatching and adult emergence inhibition; Residual effect for 12 months.

3.3. The third phase was the baseline study to collect information on the characteristics of the population, perceptions of these diseases, willingness to apply the insecticide coating in their water containers, the epidemiological profile and entomological indices (larvae and pupae/person). The sample size was calculated for detection of a 50% reduction in the House Index with > 99% power at 5% significance level. Given a baseline HI of 30%, an intra-cluster correlation coefficient (ICC) of 0.01 and a cluster size of 2,000 households it was found that a minimum of 12 clusters per study arm was needed. The same result but with a lower power (68%) was found for a baseline DZC incidence of 3% and a 50% reduction in the intervention arm. A large number of households per cluster was needed as Ae. aegypti vectors are day-biters and it was assumed that many household members stay within their neighbourhoods during the day (mainly housewives, small children and school children when the school is close to their house) where they are at risk to get infected. All estimations were done through The Shiny CRT Calculator, a web-based app to determine sample size and power for cluster trials.

3.3.1. Sampling procedure for household survey and entomological inspections. By this way, 24 high and medium endemic areas (clusters) were selected in Cucuta (16 clusters), Los Patios (4 clusters) and Villa del Rosario (4 clusters). For the baseline study they were not yet paired and allocated to the intervention and control arm as this was planned to do with the information resulting from the baseline study enabling us to find pairs of clusters with similar characteristics (endemicity, socio-economic conditions and characteristics of public spaces) For the entomological/socio-demographic household survey, a proportion of 10% of 2000 houses was sampled in this study, resulting in 200 houses in each cluster. This number was rounded to 250 by convenience. Therefore, the study targeted 6000 households (250 households in each cluster) using a systematic random sampling method. For this we first estimated the sampling interval by dividing the total number of households in each cluster (around 2000) by the number of sample households (250), resulting in a sampling interval of 8. All households located within each cluster were mapped and a randomly defined starting point was selected and from there every 8th household was visited. 3.3.2. Data collection. The baseline surveys included a household questionnaire and entomological survey. The instruments were discussed in a participatory workshop with experts and pre-tested among the members of vector control staff in the three municipalities leading to some modifications. An extensive training was conducted by the research team before starting the field work.

3.3.3. Household questionnaire. Demographic and health data was collected through face-to-face interviews using a standard questionnaire that included both structured and semi- structured questions. Household questionnaire was adapted from published research in Colombia. The questions referred to socio-demographic parameters (number of people living in the house, age, sex, and educational level), self-reported DZC disease acquired during the last months, population mobility during the day, usage of water in tanks and willingness to accept the intervention trial.