Background: Screening of blood donors for Chagas' disease by using currently available serologic tests is complicated by the lack of adequate sensitivity, discordant results between tests, and the absence of a gold standard.
Study design and methods: The study was designed to evaluate the serologic tests by using epidemiologic data relating to the risk of exposure to Trypanosoma cruzi in the urban centers of Brazil. The serologic results obtained from screening 411,617 voluntary blood donations in São Paulo during 1993 and 1994 were reviewed, as well as follow-up results on 1,267 donors who initially were repeatably reactive in at least one of three screening tests. Epidemiologic data were obtained from 321 individuals who on follow-up remained reactive in at least one test and who returned for medical counseling. Controls included 119 screen-negative blood donors and 45 blood donors who were repeatably reactive in at least one screening test but were negative on follow-up.
Results: Of the individuals who reacted in three screening tests, 94.6 percent remained reactive on follow-up. Of the individuals who were repeatably reactive in only one screening test, 70.8 percent were negative in all three tests on follow-up. Most individuals who reacted in two or three tests on follow-up had epidemiologic evidence of a risk of exposure to Chagas' disease. A significant proportion (29.1%) of those who were reactive in only one test on follow-up had epidemiologic evidence of exposure to the Chagas' disease vector as compared to 14.6 percent of controls (p = 0.007). This suggests that some of these individuals truly were infected.
Conclusion: No single test for Chagas' disease is sufficiently sensitive to prevent transfusion transmission of the disease in the urban centers of Brazil.