Intensified surveillance and insecticide-based control of the Chagas disease vector Triatoma infestans in the Argentinean Chaco

Abstract

Background: The elimination of Triatoma infestans, the main Chagas disease vector in the Gran Chaco region, remains elusive. We implemented an intensified control strategy based on full-coverage pyrethroid spraying, followed by frequent vector surveillance and immediate selective insecticide treatment of detected foci in a well-defined rural area in northeastern Argentina with moderate pyrethroid resistance. We assessed long-term impacts, and identified factors and procedures affecting spray effectiveness.

Methods and findings: After initial control interventions, timed-manual searches were performed by skilled personnel in 4,053 sites of 353-411 houses inspected every 4-7 months over a 35-month period. Residual insecticide spraying was less effective than expected throughout the three-year period, mainly because of the occurrence of moderate pyrethroid resistance and the limited effectiveness of selective treatment of infested sites only. After initial interventions, peridomestic infestation prevalence always exceeded domestic infestation, and timed-manual searches consistently outperformed householders' bug detection, except in domiciles. Most of the infestations occurred in houses infested at baseline, and were restricted to four main ecotopes. Houses with an early persistent infestation were spatially aggregated up to a distance of 2.5 km. An Akaike-based multi-model inference approach showed that new site-level infestations increased substantially with the local availability of appropriate refugia for triatomine bugs, and with proximity to the nearest site found infested at one or two preceding surveys.

Conclusions and significance: Current vector control procedures have limited effectiveness in the Gran Chaco. Selective insecticide sprays must include all sites within the infested house compound. The suppression of T. infestans in rural areas with moderate pyrethroid resistance requires increased efforts and appropriate management actions. In addition to careful, systematic insecticide applications, housing improvement and development policies that improve material conditions of rural villagers and reduce habitat suitability for bugs will contribute substantially to sustainable vector and disease control in the Gran Chaco.

Figure 1. Infestation and colonization with T. infestans and interventions performed during the 35-month follow-up.

Pampa del Indio, 2007–2010. Numbers above bars indicate number of houses inspected for infestation. The bold arrow indicates the initial community-wide insecticide spraying; thin full arrows indicate selective sprays with pyrethroid insecticides (either infested sites or infested house compounds), and thin dashed arrows indicate selective sprays with malathion of infested house compounds.

Figure 2. Persistent infestations with T. infestans after each selective spray round.

Pampa del Indio, 2007–2010. A: Site-level persistence of infestation. B: House-level persistence of infestation. Whiskers represent ±1 standard error; numbers above bars are number of observation units (sites or houses) infested, sprayed, still existing and re-inspected at the subsequent survey after the selective spray.

Figure 3. Spatio-temporal distribution of the abundance of T. infestans per site as determined by timed-manual collections.

Pampa del Indio, 2007–2010. MPS: months postspraying. Dotted line at 4 MPS divides the western and eastern sections of the study area.

Figure 4. Distribution of houses infested with T. infestans before and 4 months after initial interventions.

Pampa del Indio, 2007–2008. A: Map showing the location of persistently infested houses. B: Spatial analysis of persistently infested houses with respect to prespraying infested houses. C.E.: confidence envelope according to null model.

Figure 5. Prevalence and abundance of site-specific infestation with T. infestans according to main type of ecotope.

Pampa del Indio, 2007–2010. Infestations assessed by timed-manual collections (TMC). Only site-level data from the main ecotopes (domiciles, kitchens or storerooms, fowl coops and ‘nideros’) are included. Domiciliary sites numbered 401 because a given house compound may have more than one domiciliary site, in an analog fashion to peridomestic sites. Numbers above bars indicate number of sites inspected. Symbols indicate median bug abundance by TMC in infested sites; whiskers represent the range between the first and third quartiles.

Figure 6. Prevalence of site-level infestation with T. infestans at 4 MPS according to prespraying bug abundance.

Pampa del Indio, 2007–2008. Infestation assessed by timed-manual collections (TMC). Only data from the main ecotopes (domiciles, kitchens or storerooms, fowl coops and ‘nideros’) are included. Numbers above bars indicate the number of sites within each category.

Figure 7. Predicted infestation under different criteria of selective sprays with insecticides.

Each panel presents the expected outcome at the survey subsequent to the selective spray. A: 12 MPS. B: 17 MPS. C: 22MPS. D: 28 MPS. Bold line represents the predicted infestation as a function of the radius of insecticide spray coverage around each of the detected foci. Thin lines represent the 95% confidence envelope (C.E.) of predicted values. Circles indicate the predicted infestation if the entire house compound had been sprayed. Dotted lines represent the number of hypothetically sprayed sites as a function of spray coverage radius.