The history of Chagas disease control in Peru and many other nations is marked by scattered and poorly documented vector control campaigns. The complexities of human migration and sporadic control campaigns complicate evaluation of the burden of Chagas disease and dynamics of Trypanosoma cruzi transmission.
We conducted a cross-sectional serological and entomological study to evaluate temporal and spatial patterns of T. cruzi transmission in a peri-rural region of La Joya, Peru. We use a multivariate catalytic model and Bayesian methods to estimate incidence of infection over time and thereby elucidate the complex history of transmission in the area. Of 1,333 study participants, 101 (7.6%; 95% CI: 6.2–9.0%) were confirmed T. cruzi seropositive. Spatial clustering of parasitic infection was found in vector insects, but not in human cases. Expanded catalytic models suggest that transmission was interrupted in the study area in 1996 (95% credible interval: 1991–2000), with a resultant decline in the average annual incidence of infection from 0.9% (95% credible interval: 0.6–1.3%) to 0.1% (95% credible interval: 0.005–0.3%). Through a search of archival newspaper reports, we uncovered documentation of a 1995 vector control campaign, and thereby independently validated the model estimates.
High levels of T. cruzi transmission had been ongoing in peri-rural La Joya prior to interruption of parasite transmission through a little-documented vector control campaign in 1995. Despite the efficacy of the 1995 control campaign, T. cruzi was rapidly reemerging in vector populations in La Joya, emphasizing the need for continuing surveillance and control at the rural-urban interface.
The historically rural problem of Chagas disease is increasing in urban areas in Latin America. Peri-rural development may play a critical role in the urbanization of Chagas disease and other parasitic infections. We conducted a cross-sectional study in an urbanizing rural area in southern Peru, and we encountered a complex history of Chagas disease in this peri-rural environment. Specifically, we discovered: (1) long-standing parasite transmission leading to substantial burden of infection; (2) interruption in parasite transmission resulting from an undocumented insecticide application campaign; (3) relatively rapid re-emergence of parasite-infected vector insects resulting from an unsustained control campaign; (4) extensive migration among peri-rural inhabitants. Long-standing parasite infection in peri-rural areas with highly mobile populations provides a plausible mechanism for the expansion of parasite transmission to nearby urban centers. Lack of commitment to control campaigns in peri-rural areas may have unforeseen and undesired consequences for nearby urban centers. Novel methods and perspectives are needed to address the complexities of human migration and erratic interventions.
Citation: Delgado S, Castillo Neyra R, Quispe Machaca VR, Ancca Juárez J, Chou Chu L, Verastegui MR, et al. (2011) A History of Chagas Disease Transmission, Control, and Re-Emergence in Peri-Rural La Joya, Peru. PLoS Negl Trop Dis 5(2): e970. doi:10.1371/journal.pntd.0000970
Editor: Ricardo E. Gürtler, Universidad de Buenos Aires, Argentina
Received: September 20, 2010; Accepted: January 21, 2011; Published: February 22, 2011
This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Funding: This work was supported by the National Institutes of Health (http://nih.gov/; 5P50 AI074285-03 and 04, 3K01AI079162-02S1 to M.Z.L., 5 T37 MD001427 to S.D.), the National Institute of Allergy and Infectious Diseases (http://www.niaid.nih.gov/; 1K01AI079162-03 to M.Z.L.) and the American Society of Tropical Medicine and Hygiene (http://www.astmh.org; Benjamin Kean Fellowship to A.W.T.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The findings and conclusions in this publication are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
Competing interests: The authors have declared that no competing interests exist.
An estimated 8 million people in Latin America are infected by the protozoan parasite Trypanosoma cruzi, the causative agent of Chagas disease . Trypanosoma cruzi is typically transmitted to humans and other mammals through contact with feces of an infected blood-feeding triatomine insect. The primary vector species in southern Peru is Triatoma infestans, which has adapted to live in and around human dwellings. Infection can also occur via congenital transmission, blood transfusion, or organ transplantation . Infection is generally life-long and while most infected individuals remain asymptomatic, 20–30% progress over a period of decades to chronic clinical manifestations, including life-threatening cardiac and/or gastrointestinal disease . As a result, Chagas disease is estimated to be responsible for the loss of 670,000 disability-adjusted life years (DALYs) and 14,000 human lives annually .
Trypanosoma cruzi transmission by T. infestans has been interrupted in several South American countries through household application of pyrethroid insecticides, but a comprehensive approach to vector control has only recently been instituted in southern Peru , . Throughout Latin America, however, Chagas disease vector control is complicated by the processes of urbanization and migration , . In recent decades in southern Peru, extensive urbanization has occurred at the periphery of cities as well as within previously rural areas . New localities are typically established by rural migrants and share the trait of being situated – geographically as well as socio-culturally – at a rural-urban interface .
To improve understanding of T. cruzi transmission in the peri-rural context, we performed cross-sectional serological and entomological surveys in four contiguous localities located 30 km from the city of Arequipa. We evaluated spatial and temporal patterns of T. cruzi infection, utilizing a multivariate catalytic model  and Bayesian methods to estimate incidence of infection over time.
The ethical review committees of the Johns Hopkins Bloomberg School of Public Health, the Universidad Peruana Cayetano Heredia, and the University of Pennsylvania approved the research protocol. The ethical review committee of the University of Arizona approved the usage of de-identified study data. All individuals ≥1 year old residing within the study area were invited to participate in the serological study. Signed informed consent was obtained prior to participation by adults and parents of participating children. Children also provided signed informed assent prior to participating. All households in the study area were invited to participate in the entomological study. Signed informed consent was obtained prior to participation by an adult resident of each household.
Study area and population
The district of La Joya (population 24,192) is located approximately 30 km southwest of the city of Arequipa (population 864,250) and encompasses a mosaic of rural and peri-rural communities . In this article, peri-rural refers to communities with high-density human habitation within an otherwise rural landscape, whereas peri-urban describes localities with high-density human habitation located at the periphery of an urban center . The La Joya study area (Figure 1) was comprised of four contiguous peri-rural localities, with 2,251 persons living in 678 households within a 41-hectare area. From August through November 2008, we conducted human and entomological surveys. The human survey included a collection of demographic data and a detailed history of where each individual had lived from birth to the present time. The human survey and vector collections were timed to coincide with a household insecticide spray campaign conducted by the Arequipa Regional Office of Health. No documentation of any previous household insecticide campaign in the area was found in the Chagas Control Program records prior to the study.
The high-resolution satellite image (50-cm resolution panchromatic image acquired by WorldView-1 on 04 September 2008, eMap International) shows the peri-rural landscape of the La Joya study area. The inset map illustrates the proximate locations of the La Joya study site and the city of Arequipa.
Serological and socio-demographic data collection
Detailed socio-demographic data, including a comprehensive migration history, were collected from each consenting study participant. Migration data were coded to indicate whether each individual lived inside or outside of the study area during each calendar year over the span of his or her lifetime. Five ml of venous blood was drawn from each study participant by trained medical personnel (3 ml for children younger than 5 years). Blood was maintained at 4°C and separated by centrifugation on the day of collection. Serum and cellular fractions were stored at −20°C until testing. Serum specimens were evaluated for antibodies to T. cruzi using a commercially available enzyme-linked immunosorbent assay (ELISA) (Chagatek, Laboratorio Lemos). The assay is reported to have 100% (95% CI: 82.9–100%) sensitivity; specificity estimates have varied from 87.3% (80.9–92.2%) to 98.1% (93.8–99.7%) . All specimens with positive ELISA results and a 20% random sample of specimens with negative results were tested by the indirect immunofluorescent antibody test (IFA) using a titer of 1:32 as the positive cutoff . Specimens with a positive result for both ELISA and IFA were considered confirmed positive for T. cruzi infection.
Entomological and parasitological data collection
Immediately after application of the residual pyrethroid insecticide deltamethrin 5% Wettable Powder (K-Othrine, Bayer) at 25 mg a.i. per m2, two trained triatomine collectors systematically searched all domestic and peri-domestic areas, which included human and animal living spaces, for a total of one person-hour for each participating household . Captured T. infestans were transported the same day to the Arequipa-based Universidad Peruana Cayetano Heredia laboratory. Insect life stage, sex (adults), and count were documented by site of collection. Gut contents of all insects (except first instars) were evaluated microscopically to determine T. cruzi infection status .
Geographic data collection
Household geographic coordinates were collected using a global positioning system (GPS) unit (Juno ST, Trimble Navigation Limited). Geographic coordinates; household and individual participant codes; and diagnostic, socio-demographic, and entomological-parasitological data were maintained in a relational database management system and imported into a geographic information system (GIS) (ArcGIS 9.3.1, ESRI) for subsequent analysis and visualization.
Univariate statistical analysis.
Univariate statistical analyses were used to assess associations between participant T. cruzi serostatus and potential risk factors, which included socio-demographic characteristics and household entomological-parasitological conditions. The χ2 test was used for discrete covariates and the Wilcoxon rank-sum test for continuous covariates. Statistical tests were conducted using Stata 10 (StataCorp).
Using a GIS, household locations were mapped to visualize the spatial distributions of T. cruzi-seropositive, T. infestans-positive, and T. infestans T. cruzi-positive households. To preserve study participant confidentiality while maintaining actual and relative distance and orientation relationships, locations of all households were masked by displacing each point from its original location by a constant increment .
We employed K-function difference analysis to assess spatial clustering of households by serological, entomological, and entomological-parasitological status. The K-function, K(d), measures the total number of observations occurring within a given distance of any specified observation and describes the degree of spatial dependence of a set of observations . Evaluation of the difference in K-functions for positive versus negative observations provides an estimation of the relative degree of clustering across specified spatial scales . Specifically, we evaluated K-function differences for T. cruzi-seropositive versus seronegative, T. infestans-positive versus negative households, and T. infestans T. cruzi-positive versus negative households at spatial scales from 10 to 500 m in 10-m increments. Statistical significance of results was assessed using a 9999-iteration Monte Carlo random labeling simulation . Spatial statistical analyses were conducted in R 2.9.2 .
Bayesian Markov Chain Monte Carlo regression analysis.
We expanded upon the classic catalytic model  and utilized Bayesian Markov Chain Monte Carlo (MCMC) methods to estimate temporal and geographical parameters of individual T. cruzi infection risk. We evaluated three alternative independent models:
1. Age-only model: A basic catalytic model in which infection probability is modeled as a function of an individual's age and an intercept term:where represents the average annual risk of infection per year of age; represents an individual's age in years.
2. Migration-only model: An expanded catalytic model in which infection probability is modeled as a function of an individual's years of residence inside and outside the study area and an intercept term:where and represent the average annual risk of infection per year of residence inside and outside the study area, respectively; and represent an individual's years of residence inside and outside the study area, respectively.
3. Migration-transmission interruption model: An expanded catalytic model in which infection probability is modeled as a function of an individual's years of residence inside and outside the study area, where residence in the study area is divided into time before and after a postulated insecticide application campaign, and an intercept term:where , , and represent the average annual risk of infection per year of residence inside the study area before a postulated intervention campaign, after a postulated intervention campaign, and outside the study area, respectively; , , and represent an individual's years of residence inside the study area before a postulated intervention campaign, after a postulated intervention campaign, and outside the study area, respectively. For all models, the intercept term, , represents the probability of T. cruzi infection independent of time of exposure, including the probability of congenital transmission of T. cruzi infection. All β and α coefficients were assumed to be constant and independent of each other.
For each model, uniform prior distributions were specified for β coefficients (0–0.05), the term (0–0.05), and the number of years since a postulated intervention campaign (0–50). Models were fit using a 200,000-iteration Metropolis-Hastings MCMC algorithm , with a 100,000-iteration burn-in and 100-fold thinning factor. Regression coefficients and 95% credible intervals were derived from posterior distributions. Model convergence was evaluated using the Gelman-Rubin statistic . Goodness-of-fit of the three models was compared using the deviance information criterion (DIC) .
Univariate statistical analysis results
Of 2,251 study area residents, 1,333 (59.2%) participated in the serological survey. Compared with study participants, non-participants were younger (mean age = 22.6 versus 26.6 years; p<0.001) and were more likely to be male (57.1% versus 42.2%; p<0.001). One hundred forty-four participants had positive results by T. cruzi ELISA (10.8%; 95% CI: 9.1–12.5%), and 101 (7.6%; 95% CI: 6.2–9.0%) had confirmed positive results by IFA. Forty-three individuals had ELISA-positive/IFA-negative diagnostic results and were considered seronegative in subsequent analyses. Of 238 randomly selected ELISA-negative specimens, 226 were IFA-negative and 12 were IFA-positive. All were considered seronegative in subsequent analyses. ELISA-IFA inter-diagnostic concordance was evaluated using the kappa statistic: kappa = 0.687 (95% CI: 0.610–0.764). All T.cruzi seropositive individuals were evaluated by Arequipa Regional Office of Health medical personnel and were offered treatment in accordance with the Ministry of Health of Peru guidelines.
Study participants contributed a total of 35,501 person-years at risk for T. cruzi infection, 19,514 (55.0%) of which were experienced inside the study area and 15,987 (45.0%) of which were experienced outside the study area. The prevalence of T. cruzi infection increased with increasing age, years of residence inside the study area, and years of residence outside the study area (Table 1). Of 513 study participants aged 18 years or younger, only five (1.0%; 95% CI: 0.1–1.8%) had confirmed positive serological results, and three of these were children of T. cruzi seropositive mothers. There were no significant associations between individuals' serostatus and the presence or density of T. infestans or T. cruzi-infected T. infestans in their households (Table 2).
Of 678 study area households, 405 (59.7%) had one or more residents who participated in the serological survey. Eighty-five (21.0%; 95% CI: 17.0–25.0%) households surveyed contained at least one person with T. cruzi infection. Six hundred twenty-three (91.9%) households participated in the entomological survey, and T. infestans were collected in 171 (27.4%; 95% CI: 23.9–31.0%) households surveyed. Of the 170 infested households with vectors examined, 46 (27.1%; 95% CI: 20.3–33.8%) contained T. cruzi-positive vectors (Figure 2).
Comparisons of spatial distributions of households with T. infestans versus households with T. cruzi-infected insects (top panel); and households with T. cruzi-seropositive humans versus households with T. cruzi-infected insects (bottom panel). To preserve study participant confidentiality while maintaining actual and relative distance and orientation relationships, locations of all households were masked by displacing each point from its original location by a constant increment.
K-function difference analysis demonstrated statistically significant spatial clustering of vector-infested households at spatial scales from 10 to 500 m. Similar analysis found significant spatial clustering of households with T. cruzi-infected vectors at spatial scales from 10 to 130 m and 390 to 410 m. No significant clustering was found at any spatial scale evaluated for households containing T. cruzi-infected humans (Figure 3).
K-function difference analyses, where observed values above 99% tolerance limits indicate statistically significant spatial clustering at a given spatial scale. Clustering of households with T. infestans was statistically significant at all spatial scales from 10 to 500 m (top panel), while clustering of households with T. cruzi-infected vectors was statistically significant only at spatial scales from 10 to 130 m and 390 to 410 m (middle panel). Clustering of households with T. cruzi-seropositive humans was not statistically significant at any spatial scale from 10 to 500 m (bottom panel).
Bayesian Markov Chain Monte Carlo regression analysis results
The migration-transmission interruption model provided the best fit to the seroprevalence data (Table 3, Figure 4). This model indicated that transmission was interrupted around 1996, approximately 12 years prior to our study. Based on model estimates, the mean annual incidence of T. cruzi infection in the study area was approximately 1% prior to transmission interruption and fell to 0.1% from then onward. The mean incidence per year of residence outside the study area was approximately 0.1%. The intercept value was estimated to equal 0.5%. This value represents the probability of T. cruzi infection independent of time of exposure, including the probability of congenital infection.
Curves show the predicted prevalence of T. cruzi infection versus time at risk for each of the three regression models evaluated. A histogram of seroprevalence by five-year age category is provided for comparison of model and empirical results. The value inside each bar represents the number of T. cruzi-seropositive individuals in each 5-year age category.
Based on model results, we interviewed Ministry of Health personnel who had been working since at least the early 1990s. Several officials recounted that then-President Alberto Fujimori visited La Joya in 1995 as part of his re-election campaign and authorized financing for an insecticide spray campaign in response to local political demands. Ministry field personnel confirmed that a single, district-wide application of the pyrethroid insecticide 10% lambda-cyhalothrin at 25 mg a.i. per m2 was implemented in May to October 1995. We subsequently searched local newspaper archives and encountered an article documenting the control campaign . Data obtained from personal interviews and newspaper records strongly support model estimation of transmission interruption (Figure 5).
An expanded catalytic model and a Bayesian MCMC algorithm were used to estimate the year of an insecticide intervention in La Joya. The frequency histogram shows the distribution of the 1,000 estimates generated by the model. Model prediction of the year of the intervention was confirmed by historical documentation of a campaign authorized by then-President Fujimori as part of his reelection campaign.
The transmission of Trypanosoma cruzi by Triatoma infestans can be interrupted through coordinated and sustained vector control efforts, as has been demonstrated in Brazil, Chile, and Uruguay . The history of Chagas disease control in Peru and many other nations, however, is marked by sporadic and poorly documented vector control campaigns . These campaigns, although ephemeral, can affect the patterns of infection in a population. Through a catalytic model we were able to reconstruct the history of T. cruzi transmission in the peri-rural community of La Joya, Peru, and in the process describe the effect of a nearly forgotten vector control campaign.
The classic catalytic model has been utilized widely to retrospectively estimate incidence from age-prevalence data for a variety of lifelong infections –, including T. cruzi infection –. The basic model assumes that incidence of infection is constant with time and age; we show here that the model can be easily expanded to accurately describe a complex history of transmission and control, even in a peri-rural community with a highly mobile population. We found that incidence of T. cruzi infection was high prior to a vector control campaign in 1995. This control campaign successfully disrupted transmission of T. cruzi. Our estimates from the expanded catalytic model show prevalence in the study population increasing from 5% in twenty year olds, up to 10% among thirty year olds and 20% among those over sixty. Little of the prevalence could be attributed to infection outside of the study area.
Peri-rural and peri-urban places play a crucial role in contemporary dynamics of migration and urbanization across Latin America . In turn, migration and urbanization are changing the geography and epidemiology of many parasitic infections, including Chagas disease , . The incursion of T. infestans and T. cruzi into nearby urban and peri-urban Arequipa ,  may have its source in peri-rural communities like La Joya. Inhabitants of peri-urban Arequipa typically migrate to nearby rural and peri-rural localities in search of seasonal agricultural employment . The cycle of seasonal migration provides a plausible means for the rural-to-urban transportation of vector and parasite. The timing, at least, is suggestive. Triatoma infestans emerged in peri-urban communities of Arequipa in the 1980s and 1990s . Our analysis here shows that transmission was very common in La Joya in the 80s and midway through the 90s. Molecular studies  of the vector and parasite could provide a definitive answer to the question of provenance of Chagas disease in and around the city of Arequipa.
The intervention that nearly eliminated T. cruzi transmission in the study communities consisted of a single application of pyrethroid insecticide. The intervention effectively stopped T. cruzi transmission for many years. However, as is often the case, over time T. infestans and T. cruzi reemerged in the end . The spatial patterns we observed in La Joya are precisely those to be expected following a reasonably effective intervention campaign. A large number of older individuals were infected, and they presumably were infected prior to the intervention campaign by vector populations that were spatially diverse. The vectors we observed during our study were clustered, suggesting that the vector population is so young that it was still in the process of re-dispersing through the community. Households with vectors carrying the parasite were even more spatially clustered, suggesting that the re-dispersal of T. infestans has outpaced that of T. cruzi. We found very low prevalence of infection in children, and some of the few children diagnosed may well have been infected congenitally. The resurgent T. cruzi transmission clearly had not yet caused many infections in humans. However, had the parasite been allowed to continue to spread there is little doubt that it would have resulted in significant disease in the human population.
The contrasting presence of T. cruzi in the vector population and absence of new human infections suggest that parasite reemergence was limited to animal reservoirs. There are diverse and dense domestic animal populations in peri-rural communities of Peru, and La Joya is no exception. The guinea pig is particularly common, and is a highly competent reservoir for T. cruzi. We speculate that the presence of high densities of guinea pigs has allowed for reemergence of T. cruzi.
At least three important study limitations merit mention. First, T. cruzi diagnostics lack a gold standard, and are especially difficult to interpret in Arequipa . We considered only confirmed (ELISA-positive/IFA-positive) cases of infection; spatial analyses elsewhere in Arequipa suggests that ELISA-positive/IFA-negative results likely represent true positives . Accordingly, our estimates of seroprevalence and incidence in the study population may be lower than the true values. Second, non-participants were on average slightly younger than participants, which might lead to a slight underestimate of seroprevalence in the population, though this would not affect the incidence estimates from the catalytic model. Third, we were unable to consider a more detailed model of geographic variability in risk of T. cruzi infection. This limitation resulted from the vast variety of study participants' migration histories, which included many small towns that we could not locate.
We attempted to gather information on the previous insecticide control activities in the study site prior to conducting our study. In-depth interviews conducted with community members failed to elicit recall of the 1995 insecticide application campaign . Insecticide application was simply not a memorable event in the lives of community members. We may have lacked some due diligence in examining files in the health post; these were not made readily available to us. We note that had we included exact information on the timing of the intervention campaign in the catalytic model, our point estimates of incidence of infection before and after the campaign would not likely have changed, although the credible intervals around those point estimates may have been narrower. We present our method for making such estimates without perfect knowledge of the history of control activities because we believe they may prove useful for others conducting similar serologic studies on Chagas, or other diseases, when historical information on control activities is incomplete.
The geography and epidemiology of Chagas disease – like that of many parasitic diseases – is changing. Decreased funding and insecticide resistance are endangering gains achieved by the insecticide-based interventions such as the Southern Cone Initiative against T. infestans . Waning political interest further complicates implementation of sustained vector control, and economically and politically marginalized populations may suffer disproportionately. Transmission cycles of T. cruzi are emerging in peri-urban communities , , and re-emerging in peri-rural communities like La Joya. Integrated epidemiological, entomological, environmental and historical data are needed to better elucidate past processes driving the changing geography of Chagas disease, and to facilitate control of new cycles of transmission, especially at the complex rural-urban interface.
(0.09 MB DOC)
We gratefully acknowledge the invaluable contributions of the following institutions and individuals. The Tropical Medicine Research Center (TMRC) Peru working group conducted field-based collection of seroprevalence and entomological data as well laboratory-based diagnostic, entomological, and parasitological analyses. The Ministerio de Salud del Perú (MINSA), Dirección General de Salud de las Personas (DGSP), Estrategia Sanitaria Nacional de Prevención y Control de Enfermedades Metaxénicas y Otras Transmitidas por Vectores (ESNPCEMOTVS), Dirección General de Salud Ambiental (DIGESA); Gobierno Regional de Arequipa; Gerencia Regional de Salud de Arequipa (GRSA); Pan American Health Organization (PAHO); and Canadian International Development Agency (CIDA) facilitated the implementation of this study. We thank the communities of Las Dunas y Primavera, Los Laureles, Rio de Janeiro, and Villa La Joya for their participation in this study. We are extremely grateful to the three reviewers who provided thoughtful and insightful commentary on the manuscript, which immensely improved the quality of this work. Finally, we thank Katelyn Levy for her outstanding graphic design work and administrative assistance.
Conceived and designed the experiments: CN RHG CB MZL. Performed the experiments: RCN VRQM JAJ LCC MRV GMMA CDB. Analyzed the data: SD AWT MZL. Contributed reagents/materials/analysis tools: SD RCN VRQM JAJ LCC MRV GMMA CDB AWT CRS ACC JGCdC MZL. Wrote the paper: SD CB MZL.
- 1. World Health Organization (2007) Reporte sobre la enfermedad de Chagas. Geneva: World Health Organization.
- 2. World Health Organization (2002) Control of Chagas disease: Second report of the WHO expert committee. Geneva: World Health Organization.
- 3. Heymann DL, editor. (2004) Control of communicable diseases manual (18th edition). Washington, DC: American Public Health Association.
- 4. World Health Organization (2004) Revised global burden of disease (GBD) 2002 estimates. Geneva: World Health Organization.
- 5. Dias JCP, Silveira AC, Schofield CJ (2002) The impact of Chagas disease control in Latin America - A review. Mem Inst Oswaldo Cruz 97: 603–12.
- 6. Mott KE, Desjeux P, Moncayo A, Ranque P, de Raadt P (1990) Parasitic diseases and urban development. Bull WHO 68: 691–98.
- 7. Molyneux DH (1997) Patterns of change in vector-borne diseases. Ann Trop Med Parasitol 91: 827–39.
- 8. Instituto Nacional de Estadística e Informática del Perú (2009) Censos Nacionales 2007: XI de Población y VI de Vivienda. Lima: Instituto Nacional de Estadística e Informática del Perú.
- 9. Iaquinta DL, Drescher AW (2000) Defining the peri-urban: Rural-urban linkages and institutional connections. Land Reform, Land Settlement and Cooperatives 2: 8–26.
- 10. Muench H (1959) Catalytic models in epidemiology. Cambridge, MA: Harvard University Press.
- 11. Caballero ZC, Sousa OE, Marques WP, Saez-Alquezar A, Umezawa ES (2007) Evaluation of serological tests to identify Trypanosoma cruzi infection in humans and determine cross-reactivity with Trypanosoma rangeli and Leishmania spp. Clin Vaccine Immunol 14: 1045–49.
- 12. Williams JE (1995) Leishmania and Trypanosoma. In: Gillespie SH, Hawkey PM, editors. Medical parasitology: A practical approach (1st edition). Oxford, UK: Oxford University Press. pp. 151–76.
- 13. Levy MZ, Bowman NM, Kawai V, Waller LA, Cornejo del Carpio JG, et al. (2006) Periurban Trypanosoma cruzi-infected Triatoma infestans, Arequipa, Peru. Emerg Infect Dis 12: 1345–52.
- 14. Gurtler RE, Cohen JE, Cecere MC, Lauricella MA, Chuit R, et al. (1998) Influence of humans and domestic animals on the household prevalence of Trypanosoma cruzi in Triatoma infestans populations in northwest Argentina. Am J Trop Med Hyg 58: 748–58.
- 15. Armstrong MP, Rushton G, Zimmerman DL (1999) Geographically masking health data to preserve confidentiality. Stat Med 18: 497–525.
- 16. Ripley B (1976) The second order analysis of stationary point processes. J Appl Prob 13: 255–66.
- 17. Kelsall JE, Diggle PJ (1995) Non-parametric estimation of spatial variation in relative risk. Stat Med 14: 2335–42.
- 18. Diggle PJ, Chetwynd AG (1991) Second-order analysis of spatial clustering for inhomogeneous populations. Biometrics 47: 1155–63.
- 19. The R Project for Statistical Computing (2009) R.
- 20. Hastings WK (1970) Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57: 97–109.
- 21. Gelman A, Rubin DB (1992) Inference from iterative simulation using multiple sequences. Stat Sci 7: 457–72.
- 22. Spiegelhalter DJ, Best NG, Carlin BP, van der Linde A (2002) Bayesian measures of model complexity and fit. J R Stat Soc Ser B Stat Methodol 64: 583–639.
- 23. Anonymous (April 7, 1995) Guerra total contra chirimachas. Correo 11.
- 24. Schofield CJ, Jannin J, Salvatella R (2006) The future of Chagas disease control. Trends Parasitol 22: 583–88.
- 25. Gürtler RE (2009) Sustainability of vector control strategies in the Gran Chaco Region: current challenges and possible approaches. Mem Inst Oswaldo Cruz 104: 52–59.
- 26. Fontanet AL, Messele T, Dejene A, Enquselassie F, Abebe A, et al. (1998) Age- and sex-specific HIV-1 prevalence in the urban community setting of Addis Ababa, Ethiopia. AIDS 12: 315–22.
- 27. Wasley A, Alter MJ (2000) Epidemiology of hepatitis C: Geographic differences and temporal trend. Semin Liver Dis. pp. 1–16.
- 28. Armstrong GL, Schillinger J, Markowitz L, Nahmias AJ, Johnson RE, et al. (2001) Incidence of herpes simplex virus type 2 infection in the United States. Am J Epidemiol 153: 912–20.
- 29. Paz-Bailey G, Monroy C, Rodas A, Rosales R, Tabaru R, et al. (2002) Incidence of Trypanosoma cruzi infection in two Guatemalan communities. Trans R Soc Trop Med Hyg 96: 48–52.
- 30. Feliciangeli MD, Campbell-Lendrum D, Martinez C, Gonzalez D, Coleman P, et al. (2003) Chagas disease control in Venezuela: Lessons for the Andean region and beyond. Trends Parasitol 19: 44–49.
- 31. Gürtler RE, Cecere MC, Lauricella MA, Petersen RM, Chuit R, et al. (2005) Incidence of Trypanosoma cruzi infection among children following domestic reinfestation after insecticide spraying in rural northwestern Argentina. Am J Trop Med Hyg 73: 95–103.
- 32. Levy MZ, Kawai V, Bowman NM, Waller LA, Cabrera L, et al. (2007) Targeted screening strategies to detect Trypanosoma cruzi infection in children. PLoS Negl Trop Dis 1: e103.
- 33. Bowman NM, Kawai V, Levy MZ, Cornejo del Carpio JG, Cabrera L, et al. (2008) Chagas disease transmission in periurban communities of Arequipa, Peru. Clin Infect Dis 46: 1822–28.
- 34. Bayer AM, Hunter GC, Gilman RH, Cornejo del Carpio JG, Naquira C, et al. (2009) Chagas disease, migration and community settlement patterns in Arequipa, Peru. PLoS Negl Trop Dis 3: e567.
- 35. Marcet PL, Mora MS, Cutrera AP, Jones L, Gürtler RE, et al. (2008) Genetic structure of Triatoma infestans populations in rural communities of Santiago del Estero, northern Argentina. Infect Genet Evol 8: 835–46.
- 36. Verani JR, Seitz A, Gilman RH, LaFuente C, Galdos-Cardenas G, et al. (2009) Geographic variation in the sensitivity of recombinant antigen-based rapid tests for chronic Trypanosoma cruzi infection. Am J Trop Med Hyg 80: 410415.
- 37. Levy MZ, Bowman NM, Kawai V, Plotkin JB, Waller LA, et al. (2009) Spatial patterns in discordant diagnostic test results for Chagas disease: Links to transmission hotspots. Clin Infect Dis 48: 1104–06.
- 38. Tarleton RL, Reithinger R, Urbina JA, Kitron U, Gurtler RE (2007) The challenges of Chagas disease - Grim outlook or glimmer of hope? PLoS Med 4: 1852–57.