Factors affecting infestation by Triatoma infestans in a rural area of the humid Chaco in Argentina: a multi-model inference approach

Abstract

Background: Transmission of Trypanosoma cruzi by Triatoma infestans remains a major public health problem in the Gran Chaco ecoregion, where understanding of the determinants of house infestation is limited. We conducted a cross-sectional study to model factors affecting bug presence and abundance at sites within house compounds in a well-defined rural area in the humid Argentine Chaco.

Methodology/principal findings: Triatoma infestans bugs were found in 45.9% of 327 inhabited house compounds but only in 7.4% of the 2,584 sites inspected systematically on these compounds, even though the last insecticide spraying campaign was conducted 12 years before. Infested sites were significantly aggregated at distances of 0.8-2.5 km. The most frequently infested ecotopes were domiciles, kitchens, storerooms, chicken coops and nests; corrals were rarely infested. Domiciles with mud walls and roofs of thatch or corrugated tarred cardboard were more often infested (32.2%) than domiciles with brick-and-cement walls and corrugated metal-sheet roofs (15.1%). A multi-model inference approach using Akaike's information criterion was applied to assess the relative importance of each variable by running all possible (17,406) models resulting from all combinations of variables. Availability of refuges for bugs, construction with tarred cardboard, and host abundance (humans, dogs, cats, and poultry) per site were positively associated with infestation and abundance, whereas reported insecticide use showed a negative association. Ethnic background (Creole or Toba) adjusted for other factors showed little or no association.

Conclusions/significance: Promotion and effective implementation of housing improvement (including key peridomestic structures) combined with appropriate insecticide use and host management practices are needed to eliminate infestations. Fewer refuges are likely to result in fewer residual foci after insecticide spraying, and will facilitate community-based vector surveillance. A more integrated perspective that considers simultaneously social, economic and biological processes at local and regional scales is needed to attain effective, sustainable vector and disease control.

Figure 1. Distribution of house-specific infestation by T. infestans according to timed-manual collections in Pampa del Indio.

A: Map of the study area with inhabited houses, villages and infestation status. B: domestic and peridomestic prevalence of infestation according to village; diamond symbols denote the percentage of houses inhabited by Toba families in each village; numbers above bars indicate the number of inhabited houses in each village, whiskers indicate ±1 standard deviation. Village acronyms: 10 M, 10 de Mayo; CT, Campo Los Toros; CO, Colonia Ombú; SV-CV, El Salvaje- Los Ciervos; FB, Fortín Brown; H, La Herradura; LL, La Loma; BV, Las Bravas; CHU, Las Chuñas; RI, Santa Rita; LUG, Santos Lugares, and 3L, Tres Lagunas.

Figure 2. Typical structures of rural Pampa del Indio.

A: Domicile and part of peridomicile exhibiting a variety of roof materials (corrugated metal and tarred-cardboard sheets, and thatch). B: A domicile, partly with mud-stick walls, partly with walls made of brick laid on mud, and a corrugated metal roof. C: A mud ‘nidero’ where chickens and other poultry nest.

Figure 3. Prevalence of infestation and median abundance of T. infestans according to ecotope and ethnic group.

Data corresponds to timed-manual collections. Numbers above bars indicate the number of sites examined for infestation within each category. Whiskers indicate ±1 standard deviation.

Figure 4. Abundance of T. infestans per site in infested sites according to their ecotope.

Bug abundance determined by timed-manual collections. Light gray symbols indicate median abundance; boxes, interquartile range; whiskers, values adjacent to interquartile range; dark symbols, outside values.

Figure 5. Prevalence of infestation with T. infestans according to refuge availability among the main infested ecotopes.

Numbers above points indicate the number of sites inspected in each category. Whiskers indicate ±1 standard deviation.

Figure 6. Spatial analysis of site-specific infestation by T. infestans among the main infested ecotopes.

Only domiciles, kitchens, storerooms, chicken coops, and ‘nideros’ are considered. C.E.: confidence envelope according to the null model (random labeling) under a homogeneous Poisson process.