Chagas disease (American trypanosomiasis) has emerged as an important neglected tropical disease in the United States [Reference Cantey1]. The disease is caused by the parasite Trypanosoma cruzi, which is most commonly transmitted to humans following a blood meal by an infected triatomine insect. About one-third of those who become chronically infected can develop cardiomyopathy, other cardiac sequelae, and death. While the epidemiology of this disease has been extensively studied in Latin America, little is known about the burden of disease in the southern part of the United States where there is increasing evidence of autochthonous transmission [Reference Cantey1, Reference Custer2]. Previous reports from individual blood centres in Texas have suggested a higher than expected burden of disease; however, these reports only covered isolated regions of the state [Reference Custer2]. This study aimed to evaluate associated demographic risk factors and estimate the prevalence of Chagas blood donors who tested positive for T. cruzi antibodies across multiple, large blood donation centres throughout Texas.

Data were consolidated from all blood donors tested for T. cruzi antibodies between 2008 and 2012 from the five collaborating blood centres and the AABB Chagas Biovigilance Network [3]. Data from each individual donor were only included once in the analysis. Any blood donor who screened repeatedly reactive for antibodies to T. cruzi by either FDA licensed screening test, the Ortho T. cruzi ELISA system (Ortho-Clinical Diagnostics, USA) or Abbott PRISM Chagas chemiluminescent immunoassay (Abbott Laboratories, USA), had a sample submitted for additional confirmation testing by radio-immunoprecipitation assay (RIPA; Quest Diagnostics, USA). Only those cases found positive by RIPA were considered antibody-confirmed positive. The study objective was to describe the demographic and geographical variables of confirmed blood donors in Texas. All statistics were analysed using Stata v. 12 software (StataCorp., USA). This study was reviewed by the Institutional Review Board at Baylor College of Medicine and was determined to be exempt since all data were de-identified.

Over the 5-year study period, 907 398 individual blood donors were tested for T. cruzi antibodies. The blood donor population was equally distributed by gender (50% female, 50% male). The mean age of donors was 37 years (range 15–99 years). Blood donors were primarily Caucasian (52%) or Hispanic (38%). African American (6%), Asian (2%), and other minorities were minimally represented in this population.

Of those tested, 1/6500 (n = 140) persons were confirmed positive for T. cruzi antibody by RIPA testing. Figure 1 shows the geographical distribution of positive donors in the state, with T. cruzi-infected donors predominately located around the main urban donation centres. Prevalence of cases increased with age in a linear trend: ages 16–20 years (18/100 000), ages 21–30 years (24/100 000), ages 31–40 years (26/100 000), ages 41–50 years (33/100 000), and ages >51 years (39/100 000). Prevalence of cases was highest in Hispanics (34/100 000) and mixed races (33/100 000), and lowest in Caucasians (5/100 000) and Asians (6/100 000). No African Americans were confirmed T. cruzi antibody positive during the study period.

Fig. 1. Distribution of Trypanosoma cruzi-infected blood donors in Texas.

Blood donors originated from 75% (1938/2599) of zip codes in Texas, and positive cases came from 109 different zip codes. Zip code-level data were collected from the US Census Bureau's American Fact Finder database [4]. Wilcoxon rank sum analysis was used to measure associations between zip codes of confirmed donors vs. zip codes of negative donors. We found that confirmed T. cruzi antibody-positive blood donors were statistically more likely (P < 0·002) to reside in a zip code with a higher percentage of residents living in poverty (median 16·5%) compared to zip codes of negative donors (median 14·4%). Figure 2 geographically displays the percentage of persons living below the poverty line for each zip code. The US Census Bureau defines the poverty line as the minimum level of resources that are adequate to meet basic needs [4]. For example, the poverty line is US$23 283 for a family of four [4]. We also found that confirmed T. cruzi antibody-positive donors were statistically more likely (P < 0·001) to reside in zip codes that had a higher percentage of foreign-born residents (median 17·2%) compared to zip codes of negative donors (median 6·5%).

Fig. 2. Potential high-risk areas for disease transmission: percentage of homes living below the poverty line by zip code.

Finally, we estimated societal costs for chronic Chagas disease in blood donors in Texas using the RIPA-confirmed donors (n = 140). Cost estimates were based on the assumption that 30% would develop chronic Chagas disease (42 of RIPA confirmed). The societal cost for healthcare and lost wages of chronic infection for these cases would equal about US$3.8 million (US$91 531/case) [Reference Lee5]. These chronic Chagas cases would accrue 150 disability-adjusted life years lost (3·57 DALYs/case). The cost per case had been adjusted to US standards by the originating author [Reference Lee5].

One limitation of this study is that our study population might not be representative of the entire state. There were few blood donors screened in the more south central part of the state (South Texas Plains and Western Hill Country) where T. cruzi transmission risk is predicted to be highest based on vector abundance, a well-documented domestic transmission cycle, and other environmental factors [Reference Sarkar6–Reference Bern8]. However, this high-risk area has a smaller population compared to the rest of state, which is most likely the contributing factor of low blood donor numbers. We found poverty was associated with testing positive in our donor population, and as seen in Figure 2, those in the south central part of the state had the highest percent of impoverished residents. By not including this area, and only considering blood donors for this study, we are likely underestimating the true burden of Chagas disease.

Blood centre screening is an important component in understanding the epidemiology of T. cruzi in Texas. By combining data from multiple blood centres, we were able to more accurately describe the prevalence of T. cruzi infection in the state. About 1/6500 blood donors were confirmed positive for T. cruzi antibodies in Texas. This rate is consistent with other studies in the southern United States, indicating a substantial disease burden [Reference Leiby9]. People of advanced age, Hispanics, and mixed races were found to be at highest risk for infection. While the number of Caucasians and Asians infected were low, there is a possibility that these populations represent locally acquired infections. Infection also disproportionately occurred in people living in impoverished zip codes. Future studies should aim to understand the histories and specific risk factors for acquiring Chagas infection to better understand the transmission risk within the United States. Additionally, blood donors testing positive for T. cruzi infection should be evaluated for cardiac manifestations of disease to determine health outcomes in this population.