Surveillance of Trypanosoma cruzi transmission by serological screening of schoolchildren.

Surveillance of Trypanosoma cruzi transmission by serological screening of schoolchildren A.L.S.S. de Andrade,' F. Zicker,2 A.O. Luquetti,1 R.M. Oliveira,' S.A. Silva,' J.M.P. Souza,3 & C.M.T. Martelli1 The seroprevalence of Trypanosoma cruzi infection among children is a sensitive indicator for asses- sing the effectiveness of programmes for control of Chagas disease. In this study we report the result of a cross-sectional serological survey carried out among schoolchildren living in a poor rural area in cen- tral Brazil. Eluates of blood collected on filter-paper were tested for anti-T. cruzi antibodies using immu- nofluorescence, haemagglutination, and enzyme-linked immunosorbent assays. The overall seropreva- lence of T. cruzi infection was 7.9%, which compared with the findings of the national survey carried out in 1975-80 indicates that a twofold-to-threefold reduction in prevalence has occurred over the last 10 years. This is consistent with a reduction of transmission in the area, probably related to vector control efforts. Based on our results, the incidence of new cases was estimated to be 44 per annum in the study region. In rural areas with a scattered population, surveillance of T. cruzi transmission by sero- logical screening of children at school entry is more practical and economical than entomological evalu- ation for assessing both the risk of transmission in the community and the efficacy of vector control measures. A sample size of around 1000 schoolchildren is sufficient to detect prevalences as low as 2%, and such an approach would be practical and applicable to most areas where Chagas disease is endemic. Introduction The WHO Expert Committee on the Control of Cha- gas Disease has stated that Chagas disease control is scientifically and technologically feasible (1). The prevention of Trypanosoma cruzi infection relies on chemical control of the vector with conventional insecticide spraying, improvements in housing, and serological screening of blood donors. Chemical vector control programmes have been successfully implemented in Argentina, Brazil, and Venezuela, resulting in a significant reduction in the domestic population of vectors and interruption of transmis- sion in large parts of these countries (2-4). The environmental changes resulting from de- forestation, mechanization of agriculture, extensive replacement of natural flora by cattle pastures, and 1 Associate Professor, Department of Community Health, Insti- tute of Tropical Pathology and Public Health, Universidade Fed- eral de Goias, Goiania, Brazil. Requests for reprints should be sent to Dr Andrade at the following address: rua 1136 n° 630, Setor Marista, 74180-150 Goiania, Goias, Brazil. 2 Communicable Diseases Epidemiologist, Pan American Health Organization, Maracay, Estado Aragua, Venezuela. 3Associate Professor, Faculty of Public Health, Universidade de Sao Paulo, Sao Paulo, Brazil. Reprint No. 5324 the dislocation of rural workers to urban areas that have occurred in some areas in Brazil have also been cited as factors contributing to the control of triato- mine bugs (5, 6). In Brazil, because of political and economic constraints, as well as frequent shifts of resources to different public health problems, it has not been pos- sible to implement regular chemical control of T. cruzi vectors in all the endemic areas. Also, the replacement of the domestic vector, Triatoma infestans, by secondary species has called attention to the importance of the peridomestic habitat in the transmission of Chagas disease and to the need for environmental management as part of the control effort. Assessment of the impact of programmes for the control of Chagas disease is usually based on ento- mological monitoring of house infestation and on serological evaluation of selected cohorts of the population (7). The seroprevalence of T. cruzi infec- tion among children is a sensitive indicator of house- hold rates of seropositivity, and for the assessment of the risk of transmission in the community is more practical and economical than entomological evalua- tion (8, 9). In sparingly populated rural areas, where house-to-house surveys are operationally difficult, an alternative approach is the serological screening of schoolchildren (10, 11). Bulletin of the World Health Organization: 70 (5): 625-629 (1992) © World Health Organization 1992 625 A.L.S.S. de Andrade et al. Here we report the result of a cross-sectional survey of schoolchildren living in a poor rural area in central Brazil, where very few economic and en- vironmental changes have occurred in the last two decades, and where vector control activities have been kept as regular as possible. The study was designed to provide baseline information on T. cruzi transmission and to select individuals for partici- pation in further studies on risk factors for the infec- tion. Materials and methods Study area The survey was carried out between March and September 1991 in rural communities in north-east Goia's State, central Brazil. A population census and housing plan of the study areas were provided by the Ministry of Health. The counties of Posse (25 295 inhabitants), Simolandia (6242 inhabitants) and Guarani de Goias (4766 inhabitants) were selected because about 50% of their population was rural and they had a high seroprevalence of T. cruzi infection in the national rural survey conducted in 1975-80 (12). Over the last 5 years, house spraying with insecticide has been greatly reduced in this area, as well as in Brazil as a whole. The rural population in the study area is fairly stable and depends on subsistence farming, while the dwellings are mainly made of mud and wattle or unplastered adobe bricks. Study population A total of 60 village schools with 20-40 students each, i.e., a total of 1990 schoolchildren aged 7-12 years, corresponding to 82% of the estimated number of 7-12-year-olds in the three counties, were includ- ed in the study. The schools were randomly selected from a list provided by the local education depart- ment. Data collection and laboratory tests A blood sample was collected from all participants by finger-prick onto filter-paper (Whatman No. 1), filling two circles of 1.6-cm diameter (area, 2 cm2) (13). After being dried at room temperature, the filter-papers were stored at 4 °C according to standard procedures (14) and sent weekly for exami- nation to the Reference Laboratory for Chagas Disease, Goiania (15) (600 km from the study area). Eluates from the filter-paper were tested simulta- neously for anti-T. cruzi antibodies using indirect immunofluorescence (IIF) (16), indirect haemagglu- tination (IHA) (17) and enzyme-linked immunosor- bent assays (ELISA) (18), as described previously (19). A sample was considered positive when at least two serological tests were positive, i.e., under the following conditions: IIF titre >1:20; ELISA absorbance .1.2 times the cut-off value; and IHA .1:10. Borderline results were retested using the sample on the stored filter-paper and taken to be negative if the results were persistently borderline. Data processing and analysis The data were analysed using the Statistical package for the social sciences (SPSS/PC, 1987).a The significance of differences between proportions was determined using X2 tests. P-values <0.05 were considered to be statistically significant. The asso- ciation between seropositivity, age, and sex was assessed by calculating X2 tests for trend, estimates of relative risks (odds ratios), and 95% confidence limits (95% CL). Tests for interaction were perform- ed using the Epidemiological graphic, estimation and testing package (EGRET) software.b Results Of the 1990 schoolchildren aged 7-12 years who were tested, 52.6% were males and 53.9% were aged 7-9 years. The overall seroprevalence of T. cruzi infection was 7.9% (95% CL: 6.8-9.1%) and there were no statistically significant differences between the prevalences estimated for each county. For a given age group, males had a higher prevalence of infection than females, the overall prevalence for males being 9.3% versus 6.5% for females (P = 0.02) (Table 1). The seroprevalence increased with age, ranging from 3.9% for children aged 7 years to 11.5% among 12-year-olds (test for trend, P<0.01). The same pattem was observed in the three counties. Table 2 shows the odds ratios for anti- bodies against T. cruzi among 10-12-year-olds com- pared with those aged 7-9 years, after controlling for sex. For 10-12-year-olds, males had a significantly higher risk of infection than females (3.3 versus 2.2). No significant difference in the risk of infection was observed between younger males and females, and for positive serological tests there was no statistically significant interaction with age and sex. a SPSS Inc., Chicago, IL, USA. b SERC, Seattle, WA, USA. 626 WHO Bulletin OMS. Vol 70 1992 Trypanosoma cruzi infection among schoolchildren Table 1: Seroprevalence of Trypanosoma cruzi infec- tion, by age and sex, among schoolchildren in central Brazila Age Males Females Totalb (years) Pos/n c Pos/n Pos/n 7 10/229 (4.4) d 7/210 (3.3) 17/439 (3.9) 8 11/164 (6.7) 8/160 (5.0) 19/324 (5.9) 9 11/150 (7.3) 8/159 (5.0) 19/309 (6.1) 10 24/169 (14.2) 10/150 (6.7) 34/319 (10.6) 1 1 20/154 (13.0) 15/150 (10.0) 35/304 (11.5) 12 21/180 (11.7) 13/115 (11.3) 34/295 (11.5) Total 97/1046 (9.3) e 61/944 (6.5) e 158/1990 (7.9) a Based on the results of indirect immunofluorescence, indirect haemagglutination, and enzyme-linked immunosorbent assays. b %2 for trend adjusted for sex = 24.2; P<0.01 c Pos/n = No. of positive samples/total number of samples. d Figures in parentheses are percentages. e X2 = 5.3; P= 0.02. Table 2: Odds ratios for antibodies against Trypano- soma cruzi among schoolchildren, according to sex and age group Odds ratio Age group (years) Males Females Alla 10-12 3.3 (1.9-5.5) b 2.2 (1.2-3.9) 2.3 (1.6-3.3) 7-9 1.4 (0.8-2.5) 1.0 1.0 a Odds ratio for 10-12-year-olds versus 7-9-year-olds, adjusting for sex. b Figures in parentheses are the 95% confidence limits. Discussion Serological screening is a useful tool for monitoring T. cruzi transmission and for evaluating the effec- tiveness of vector control programmes for Chagas disease (10, 20). The development of simplified methods for blood collection and sensitive techniques to detect anti-T. cruzi antibodies have made the diagnosis of T. cruzi infection easy and affordable in many large-scale field studies in urban and rural areas (12, 21). An overall T. cruzi sero- prevalence of 13.1% has been reported among 5425 unskilled urban workers in central Brazil, reflecting the level of infection among rural migrants (21). Serological screening for T. cruzi infection among blood donors in this region indicated a decreasing trend of infection for the period of 1980-90 (22, 23). Few serological studies of T. cruzi infection have been conducted on representative samples of rural populations. In Brazil, the only such national serological survey, carried out between 1975 and 1980, indicated an overall seroprevalence of infec- tion of 4.2%, although a wide regional variation was observed (12); an estimated 2 million people were infected in rural areas (24). For the rural counties of Goias State, a mean prevalence of 7.4% was reported in the national survey (range, 0.2-26.3%). The prevalence of T. cruzi-positive serology among 7-14-year-olds was around 23% for Posse and Simolandia and 12% for Guarani de Goias (Ministry of Health, unpublished data, 1980). In the current study, which was carried out approximately a decade later, the 7.9% positive serology to T. cruzi was esti- mated for a representative sample of 7-12-year-olds, indicating that transmission was still occurring. Comparison of the results of the national survey with those from the present study shows that the prevalence of T. cruzi infection decreased by 2-3 times over the last 10 years. However, it should be borne in mind that for the national survey no specific age-sex prevalences and denominators are available to permit adequate comparison. The increasing sero- prevalence with age found in our study is similar to that described in other cross-sectional studies in rural areas (25, 26), and arose through either a cumulative opportunity for infection or a cohort effect. Children born between 1979 and 1981 (10-12-year-olds) possibly had a higher risk of infection than those born in the period 1982-84 (7-9-year-olds) (odds ratio = 2.3; 95% CL = 1.6-3.3). This suggests that the reduction of transmission in the area was prob- ably related to vector control. The cohort of schoolchildren studied represents a sample of the population born before the attack phase of the vector control programme (1984-85) in the study area. When the study was carried out, the three counties were under entomological surveil- lance. Although serological surveys among children below the age of 6 years would have provided more accurate information on the impact of the program- me, since only incidence cases of infection would have been detected, such surveys are much more difficult to conduct. The incidence of T. cruzi infection in the study area is difficult to estimate. Based on the approach suggested by Hayes & Schofield (27) for an estima- ted rural population of 21 563, an overall prevalence of infection of 7.9% and a crude birth rate of 25.6% (national census data, 1990), 44 new cases of Chagas disease would be expected to occur annually in the study area; however, this is probably an underesti- mate, since the prevalence of T. cruzi infection in children is generally lower than that in adults. Vector control operations using insecticidal paints and fumigant canisters have resulted in favour- WHO Bulletin OMS. Vol 70 1992 627 A.L.S.S. de Andrade et al. able outcomes. These novel approaches, together with community-based bug monitoring using sensor boxes, form the basis of altematives for affordable vector control activities (15). Evaluation of the impact of altemative approaches on the control of Chagas disease has as a prerequisite the development of indicators of efficiency to facilitate comparison of different strategies and to perform trend analyses. Serological screening is a valuable tool in all phases of vector control since it can provide baseline information and age-prevalence data for monitoring the impact of control programmes (9). The challenge is to design an affordable surveillance system within vector control programmes that includes both sero- logical and entomological evaluations. In scattered populations the screening of school- children is more feasible than efforts to screen infants-which would, however, reveal more accu- rately the incidence of T. cruzi infections over the last 1-2 years. Also, it is cheaper and easier to deter- mine the risk of transmission in the community than to detect triatomine bugs in dwellings. Serological screening of children at school-entry age in a random sample of rural schools, every 4-5 years, would permit reliable assessment of the impact of Chagas disease control programmes. The number of schools and children to be examined would depend on the prevalence of infection in the area under surveillance. A sample size of around 1000 schoolchildren would be sufficient to estimate prevalences as low as 2% (95% CL, 1.1-2.8), which should be practical and applicable to most endemic areas. A network of laboratories with appropriate quality control standards is required for a serological evaluation programme. Surveillance by serologically screening school- children for T. cruzi could also be useful for mapping morbidity pattems and for detecting any changes in the time trend, thereby permitting com- parisons between regions. The decrease in the incidence of T. cruzi infections estimated from prevalence data may reflect the efficacy of the vector control measures; however, changes in environ- mental factors should also be taken into account when assessing the impact of such programmes. Acknowledgements We are grateful to the staff of the Superintendencia de Campanhas (SUCAM) for their cooperation in the field- work and to Miss Ana Maria de Castro for technical assistance in performing the laboratory tests. The UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases is thanked for providing financial support. Resume Surveillance de la transmission de Trypanosoma cruzi par des tests de depistage s6rologique chez les ecoliers Le pr6sent article analyse les resultats d'une enquete transversale menee aupres de 1990 6co- liers ages de 7-12 ans, soit 82% du nombre esti- m6 des enfants d'age scolaire de 3 districts ruraux du centre du Br6sil. L'etude a porte sur un total de 60 6coles de village. Une goutte de sang 6tait recueillie sur papier filtre aprbs piqOre au doigt et la recherche des anticorps anti-Trypano- soma cruzi dans l'6luat etait faite simultan6ment par immunofluorescence indirecte (IIF), hemagglu- tination indirecte (IHA) et titrage immuno-enzyma- tique (ELISA). Un 6chantillon 6tait consider6 comme positif lorsque deux epreuves serolo- giques au moins donnaient un r6sultat positif. Globalement, la seroprevalence de l'infection a T. cruzi a ete 6valu6e a 7,9%, les garcons etant plus atteints que les filles (9,3% contre 6,5%, P = 0,02). D'autre part, dans le groupe des 10-12 ans, le risque d'infection etait significativement plus 6lev6 chez les garcons que chez les filles (3,3 contre 2,2). La comparaison des r6sultats de cette etude avec ceux de l'enquete nationale effectuee dans la meme r6gion en 1984-1985 montre que la pr6valence de la maladie de Cha- gas a 6te divisee par deux ou trois au cours des dix dernieres annees, ce qui semble indiquer une reduction de la transmission, probablement lide aux efforts accomplis en matibre de lutte antivec- torielle. On peut s'attendre a ce qu'au moins 44 nou- veaux cas de maladie de Chagas se declarent annuellement dans la r6gion a l'6tude. Afin d'6va- luer l'efficacit6 des programmes de lutte, nous estimons qu'il serait possible de d6tecter une pre- valence de l'ordre de 2% avec une limite de confiance de 95% en pratiquant des tests de d6pistage serologique tout les 4 a 5 ans sur envi- ron 1000 enfants au d6but de leur scolarisation dans des 6coles rurales choisies au hasard. Une telle approche serait applicable en pratique dans la plupart des r6gions ou la maladie de Chagas est end6mique. References 1. WHO Technical Report Series No. 811, 1991 (Control of Chagas disease: report of a WHO Expert Committee). 2. Dias, J.C.P. Control of Chagas disease in Brazil. Parasitology today, 3: 336-341 (1987). 628 WHO Bulletin OMS. Vol 70 1992 Trypanosoma cruzi infection among schoolchildren 3. Garcia-Zapata, M.T. et al. Epidemiological vigilance with community participation in the control of vectors of Chagas' disease in Goias, Central Brasil. Revista Argentina de Microbiologia, 20: 106-117 (1988). 4. Tonn, R.J. Review of recent publications on the ecology, biology and control of vectors of Chagas' disease. Revista Argentina de Microbiologia, 20: 4-24 (1988). 5. Dias, J.C.P. Rural resource development and its potential to introduce domestic vectors into new epidemiological situations. Revista Argentina de Microbiologia, 20: 81-85 (1988). 6. Litvoc, J. et al. [Determinants of domestic infesta- tion by Panstrongylus megistus: the role of type of dwelling and of deforestation]. Revista do Instituto de Medicina Tropical de Sao Paulo, 32: 443-449 (1990) (In Portuguese). 7. Souza, A.G. et al. Consolidation of the control of Chagas' disease vectors in the state of Sao Paulo. Mem6rias do Instituto Oswaldo Cruz, Rio de Janeiro, 79: 125-131 (1984). 8. Mott, K.E. et al. The epidemiology and household distribution of seroreactivity to Trypanosoma cruzi in a rural community in north-east Brazil. American journal of tropical medicine and hygiene, 25: 551-562 (1976). 9. Hoff, R. et al. Serologic surveillance of Chagas' disease. Annales de la Soci6t6 belge de M6dicine tropicale, 65: 187-196 (1985). 10. Segura, E.L. et al. Decrease in the prevalence of infection by Trypanosoma cruzi (Chagas' disease) in young men of Argentina. Bulletin of the Pan Ameri- can Health Organization, 19: 252-264 (1985). 11. Salvatella, R. et al. [Seroprevalence of antibodies against Trypanosoma cruzi in 13 departments in Uruguay]. Boletin de la Oficina Sanitaria Panameri- cana, 107: 108-117 (1989) (in Spanish). 12. Camargo, M.E. et al. [Serological study of the prevalence of Chagas disease in Brazil, 1975-1980]. Revista do Instituto de Medicina Tropical de Sao Paulo, 26: 192-204 (1984) (in Portuguese). 13. Souza, S.L. & Camargo, M.E. The use of filter- blood smears in a practical fluorescent test for American trypanosomiasis serodiagnosis. Revista do Instituto de Medicina Tropical de Sao Paulo, 8: 255-258 (1966). 14. Guimaries, M.C.S. et al. Comparison of IgG and IgM contents in serum and filter-paper blood eluates. American journal of tropical medicine and hygiene, 27: 350-353 (1978). 15. Tropical diseases: progress in research, 1989-1990: tenth programme report of the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases. Geneva, World Health Organization, 1991, pp. 69-77. 16. Camargo, M.E. Fluorescent antibody test for the serodiagnosis of American trypanosomiasis. Technical modification employing preserved culture forms of Trypanosoma cruzi in a slide test. Revista do Instituto de Medicina Tropical de Sao Paulo, 8: 227-234 (1966). 17. Camargo, M.E. et al. Hemagglutination with preser- ved, sensitized cells; a practical test for routine serologic diagnosis of American trypanosomiasis. Revista do Instituto de Medicina Tropical de Sao Paulo, 15: 81-85 (1973). 18. Voller, A. et al. Microplate enzyme-linked immuno- sorbent assay for Chagas' disease. Lancet, 1: 426-427 (1975). 19. Zicker, F. et al. Mass screening for Trypanosoma cruzi infections using immunofluorescence, ELISA and haemagglutination tests on serum samples and on blood eluates from filter-paper. Bulletin of the World Health Organization, 68: 465-471 (1990). 20. Chuit, R. et al. Usefulness of serology for the evaluation of Trypanosoma cruzi transmission in endemic areas of Chagas' disease. Revista da Sociedade Brasileira de Medicina Tropical, 22: 119-124 (1989). 21. Zicker, F. et al. Seroprevalence of Trypanosoma cruzi infection among unskilled urban workers in central Brazil. Transactions of the Royal Society of Tropical Medicine and Hygiene, 83: 511-513 (1989). 22. Andrade, A.L.S.S. et al. [Serological screening for infectious diseases in blood banks as an indicator of illness in the population]. Revista de Saude Publica, Sao Paulo, 23: 20-25 (1989) (in Portuguese). 23. Zicker, F. et al. Trends of T. cruzi infection based on data from blood bank screening in central Brazil. Revista do Instituto de Medicina Tropical de Sao Paulo, 32: 132-137 (1990). 24. Silveira, A.C. & Sakamoto, T. [Medico-social importance of Chagas disease in Brazil and its control]. Revista brasileira de Malariologia, 35: 127-134 (1983) (in Portuguese). 25. Pereira, J.B. & Coura, J.R. [Chagas disease mor- bidity. A cross-sectional study in the endemic area of Virgem da Lapa, Minas Gerais]. Revista da Sociedade Brasileira de Medicina Tropical, 19: 139-148 (1986) (in Portuguese). 26. Coura, J.R. et al. [Chagas disease morbidity, Part IV: longitudinal study over 10 years in Pains and Iguatama, Minas Gerais, Brazil]. Mem6rias do Instituto Oswaldo Cruz, 80: 73-80 (1985) (in Portu- guese). 27. Hayes, R. & Schofield, C.J. [Estimating incidence rates of infections and chronic parasitoses from prevalence data: Chagas disease in Latin America]. Boletin de la Oficina Sanitaria Panamericana, 108: 308-316 (1990) (in Spanish). WHO Bulletin OMS. Vol 70 1992 629