Ana Luiza Bierrenbach

Effect of BCG revaccination on incidence of tuberculosis in school-aged children in Brazil: the BCG-REVAC cluster-randomised trial.

BACKGROUND Many countries offer a second BCG vaccination to prevent tuberculosis, although there is little evidence of whether this confers additional protection. BCG vaccination is routine in Brazil but BCG revaccination procedures vary by state. We studied revaccination efficacy in two Brazilian cities with tuberculosis prevalence representative of Brazil. METHODS We did a cluster-randomised trial of the protection against tuberculosis from BCG revaccination in school-aged children who had had one BCG vaccination as infants. 767 schools in the cities of Salvador and Manaus, Brazil, participated; schools were the unit of randomisation. The study was open label with no placebo. Cases of tuberculosis were identified through record linkage to the Tuberculosis Control Programme. Revaccination status was masked during linkage and validation of cases. The incidence of tuberculosis was the primary outcome. Analysis was by intention to treat. FINDINGS 386 schools (176,846 children) were assigned BCG revaccination and 365 (171,293 children) no revaccination. 42,053 children in the vaccine group and 47,006 in the control group were absent from school on the day of the visit and were excluded. 31,163 and 27,146, respectively were also excluded because they had no BCG scar, two or more scars, or a doubtful scar on assessment. The crude incidence of tuberculosis in the intervention group was 29.3 per 100,000 person years and in the control group 30.2 per 100,000 person-years (crude-rate ratio 0.97; 95% CI 0.76-1.28). The efficacy of BCG revaccination was 9% (-16 to 29%). INTERPRETATION Revaccination given to children aged 7-14 years in this setting does not provide substantial additional protection and should not be recommended. Follow-up is ongoing and needed to assess the effect of other factors on revaccination efficacy: time since vaccination, age at vaccination, and high or low prevalence of environmental mycobacteria.

Incidência de tuberculose e taxa de cura, Brasil, 2000 a 2004

OBJETIVO: Descrever a distribuicao geografica da incidencia de tuberculose, a partir de um conjunto de indicadores epidemiologicos e operacionais de dados de notificacao oficial. METODOS: Dados sobre incidencia de tuberculose foram coletados no Sistema de Informacao de Agravos de Notificacao, apos processo de pareamento e depuracao de registros repetidos. As taxas de incidencia de tuberculose foram calculadas segundo unidade geografica, grupo etario, sexo, forma clinica e regime de tratamento, e padronizadas para a distribuicao etaria da populacao com base no Censo de 2000. RESULTADOS: Em 2004, o Brasil apresentou taxa de incidencia de 41/100.000 habitantes, com 74.540 casos novos notificados. Desses, 52,8% eram casos pulmonares com baciloscopia positiva, 24,1% estavam em tratamento supervisionado, 63,5% eram provenientes de capitais ou das regioes metropolitanas e 54,9% eram casos curados. Excluindo-se os registros sem preenchimento da variavel de desfecho, a proporcao de cura alcancou 72,4% para casos novos, 47% para casos novos HIV positivos, 64,9% para recidivas, 64,5% transferencias e 40% para reingressos apos abandono. A taxa de cura para os casos novos em tratamento supervisionado foi de 77,1%. A proporcao de registros sem informacao sobre desfecho foi maior em anos mais recentes. CONCLUSOES: Houve extensas diferencas estaduais em relacao a incidencia e as categorias de desfecho. Para alcancar a meta de 85% de cura para casos novos e aumentar a cura dos casos HIV positivos e reingressos sao necessarios esforcos adicionais por parte do Programa Nacional de Controle da Tuberculose, incluindo a expansao da estrategia de tratamento diretamente supervisionado.

Tendência da mortalidade por tuberculose no Brasil, 1980 a 2004

OBJETIVO: Analisar o perfil atual e as tendencias da mortalidade por tuberculose no Brasil, de 1980 a 2004. METODOS: Estudo descritivo de tuberculose como causa basica ou associada de obito do Brasil. Foram utilizados dados secundarios do Sistema de Informacao sobre Mortalidade. RESULTADOS: Houve reducao inconstante do numero e da taxa de mortalidade por tuberculose ao longo do periodo estudado, de 5,8 em 1980 para 2,8 por 100 mil habitantes, em 2004. Foi observada reducao acentuada da mortalidade ate 1985, mais evidente nos estados do Rio de Janeiro e Sao Paulo. Houve tendencia de aumento da mortalidade por tuberculose em idades mais avancadas. Em 2004, foram notificados no Brasil 4.981 obitos por tuberculose como causa basica, valor que aumentaria para 50% se fossem incluidos os obitos por tuberculose como causa associada e por sequela de tuberculose como causa basica. Em 2004, as maiores taxas padronizadas foram dos estados de Pernambuco (5,4) e Rio de Janeiro (5,0), e das capitais Recife (7,7) e Belem (5,8). CONCLUSOES: Dada a conhecida associacao entre tuberculose e Aids, a influencia da epidemia de Aids reflete indiretamente nas tendencias de mortalidade de tuberculose.

Efeito da remoção de notificações repetidas sobre a incidência da tuberculose no Brasil

OBJETIVO: Avaliar o impacto nas taxas de incidencia de tuberculose com a exclusao de registros indevidamente repetidos no sistema de notificacao. METODOS: Foram analisados dados do Sistema de Informacao de Agravos de Notificacao do Ministerio da Saude, referentes ao periodo de 2000 a 2004. Os registros repetidos foram identificados por pareamento probabilistico e classificados em seis categorias excludentes que determinaram suas remocoes, vinculacoes ou permanencias na base. RESULTADOS: Verificou-se que 73,7% das notificacoes eram unicas, 18,9% formavam duplas, 4,7% triplas e 2,7% grupos de quatro ou mais registros. Dentre os registros repetidos, 47,3% foram classificados como transferencia entre unidades de saude, 23,6% reingresso, 16,4% duplicidade verdadeira, 10% recidiva, 2,5% foram inconclusivos e 0,2% tinham dados incompletos. Essas percentagens variaram entre estados. A exclusao de registros indevidamente repetidos resultou em reducao na taxa de incidencia por 100.000 habitantes de 6,1% em 2000 (de 44 para 41,3), 8,3% em 2001 (de 44,5 para 40,8), 9,4% em 2002 (de 45,8 para 41,5), 9,2% em 2003 (de 46,9 para 42,6) e 8,4% em 2004 (de 45,4 para 41,6). CONCLUSOES: Os resultados sugerem que as taxas observadas de incidencia de tuberculose representem estimativas mais proximas do que seriam os valores reais do que as obtidas com a base em seu estado bruto, tanto em nivel nacional como estadual. A pratica de pareamento de registros de notificacao de tuberculose deve ser estimulada e mantida para melhoria da qualidade dos dados de notificacao.

Tuberculin reactivity in a population of schoolchildren with high BCG vaccination coverage

OBJECTIVE To investigate the influence of BCG vaccination or revaccination on tuberculin skin test reactivity, in order to guide the correct interpretation of this test in a setting of high neonatal BCG vaccination coverage and an increasing BCG revaccination coverage at school age. METHODS We conducted tuberculin skin testing and BCG scar reading in 1 148 children aged 7-14 years old in the city of Salvador, Bahia, Brazil. We measured the positive effect of the presence of one or two BCG scars on the proportion of tuberculin skin test results above different cut-off levels (induration sizes of > or = 5 mm, > or = 10 mm, and > or = 15 mm) and also using several ranges of induration size (0, 1-4, 5-9, 10-14, and > or = 15 mm). We also measured the effects that age, gender, and the school where the child was enrolled had on these proportions. RESULTS The proportion of tuberculin results > or = 10 mm was 14.2% (95% confidence interval (CI) = 8.0%-20.3%) for children with no BCG scar, 21.3% (95% CI = 18.5%-24.1%) for children with one BCG scar, and 45.0% (95% CI = 32.0%-58.0%) for children with two BCG scars. There was evidence for an increasing positive effect of the presence of one and two BCG scars on the proportion of results > or = 5 mm and > or = 10 mm. Similarly, there was evidence for an increasing positive effect of the presence of one and two scars on the proportion of tuberculin skin test results in the ranges of 5-9 mm and of 10-14 mm. The BCG scar effect on the proportion of results > or = 5 mm and > or = 10 mm did not vary with age. There was no evidence for BCG effect on the results > or = 15 mm. CONCLUSIONS In Brazilian schoolchildren, BCG-induced tuberculin reactivity is indistinguishable, for results under 15 mm, from reactivity induced by Mycobacterium tuberculosis infection. BCG revaccination at school age increases the degree of BCG-induced tuberculin reactivity found among schoolchildren. This information should be taken into account in tuberculin skin test surveys intended to estimate M. tuberculosis prevalence or to assess transmission patterns as well as in tuberculin skin testing of individuals used as an auxiliary tool in diagnosing tuberculosis. Taking this information into consideration is especially important when there is increasing BCG revaccination coverage.

Subnotificação da comorbidade tuberculose e aids: uma aplicação do método de linkage

OBJECTIVE To analyze the underreporting of the tuberculosis (TB) and AIDS comorbidity. METHODS Surveillance study using records from the Notifiable Diseases Information System - Tuberculosis and AIDS in Brazil from 2000 to 2005. Records of TB without information on the presence of Aids were considered to be underreporting of the comorbidity when paired off with AIDS records in which the year of diagnosis of AIDS was the same or previous to the year of reporting of TB, as well as records from the same patient whose previous records had this information. An indicator was created: recognized TB-AIDS comorbidity, based on the TB records that had information on the presence of AIDS. RESULTS The underreporting of TB-AIDS was 17.7%. This percentage varied between states. The incorporation of the underreported records into the previously recognized ones increased the proportion of TB-AIDS in Brazil from 6.9% to 8.4%. The highest proportions of underreporting were noted in Acre (Northern), Alagoas, Maranhão and Piauí (Northeastern) (more than 35% each) and the lowest in São Paulo (Southeastern) and Goiás (Central-western) (around 10% each). CONCLUSIONS The underreporting of the TB-AIDS comorbidity found in Brazil will probably trigger modifications in the surveillance system in order to provide information for the national programs.OBJETIVO: Analisar a subnotificacao da comorbidade tuberculose (TB) e aids. METODOS: Estudo de vigilância utilizando os registros do Sistema de Informacao de Agravos de Notificacao de Tuberculose e de aids no Brasil de 2000 a 2005. Registros de TB sem informacao da presenca de aids foram considerados subnotificacoes da comorbidade quando pareados a registros de aids que apresentassem ano de diagnostico de aids igual ou anterior ao ano de notificacao da TB, assim como os registros de um mesmo paciente cujos registros anteriores apresentavam essa informacao. Criou-se um indicador: comorbidade TB-aids reconhecida, a partir dos registros de TB com a informacao de presenca de aids. RESULTADOS: A subnotificacao de TB-aids foi de 17,7%. Esse percentual variou entre estados. A incorporacao dos registros subnotificados aos previamente reconhecidos elevou a proporcao de TB-aids no Brasil de 6,9% para 8,4%. As maiores proporcoes de subnotificacao foram observadas no Acre, Alagoas, Maranhao e Piaui (mais de 35% cada) e as menores em Sao Paulo e Goias (cerca de 10% cada). CONCLUSOES: A subnotificacao da comorbidade TB-aids encontrada no Brasil deve deflagrar modificacoes no sistema de vigilância para prover informacoes aos programas nacionais.

Sensibilidade e especificidade da leitura da cicatriz vacinal do BCG

OBJECTIVE To validate the BCG scar as a marker of BCG vaccination status. METHODS A cross-sectional survey was carried out among 53,348 schoolchildren aged 6-14 years who underwent BCG scar examination as part of a large BCG vaccine trial taking place in the city of Manaus, Brazil. Results of BCG scar reading were compared with information on vaccine status of their vaccination cards or provided by parents or guardians. Double-reading was performed in a sub-sample. Data analysis was conducted using Stata 7 and Kappa coefficient. RESULTS Of 52,348 schoolchildren studied, vaccine status information from parents/guardian letters was available for 29,254 and from vaccination cards for 4,947. There was found a high agreement between the double-readings of the scars (Kappa=0.81). When the agreement between letter and card information was the gold standard, the sensitivity of BCG scar readings was 96.6% (95%CI 96.0-97.1) and the specificity was 71.1% (95%CI 55.7-83.7). The sensitivity was 96.1%, 97.3% and 95.3% for children vaccinated up to one month of age, four months and one year, respectively. CONCLUSIONS Sensitivity and specificity did not show an association with the childs age at the scar reading. BCG scar was a good marker of BCG vaccination status regardless of age - from the first years of life up to 14 years old.OBJETIVO: Validar a utilizacao da cicatriz vacinal de BCG como um indicador de vacinacao. METODOS: Foi realizado um estudo transversal em 52.348 escolares, entre 6 e 14 anos de idade, que possuiam exame de cicatriz vacinal do BCG e que participaram de um ensaio clinico randomizado e controlado na cidade de Manaus, Brasil. Os dados da leitura da cicatriz vacinal foram comparados com a informacao sobre a vacinacao passada fornecida pelos cartoes vacinais ou informacao dos responsaveis. Em uma subamostra foi realizada leitura dupla com calculo do coeficiente Kappa. Para analise dos dados utilizou-se o Stata 7. RESULTADOS: Do total de 52.348 escolares estudados, 29.254 possuiam informacao sobre cicatriz vacinal coletada por meio de carta aos pais, e 4.947 possuiam historia de vacinacao coletada pelo cartao de vacinas. Observou-se elevada concordância entre a dupla leitura de cicatriz vacinal (Kappa =0,81). A sensibilidade da leitura de cicatriz vacinal foi 96,6% (95% IC 96,0-97,1) e a especificidade foi 71,1% (95% IC 55,7-83,7) quando o padrao ouro utilizado foi a concordância entre a carta aos pais e a informacao do cartao de vacinas. A sensibilidade foi de 96,1%, 97,3% e 95,3% para criancas vacinadas ate um mes de idade, ate 4 meses e ate um ano de idade, respectivamente. CONCLUSOES: Os valores encontrados para sensibilidade e especificidade foram independentes da idade da realizacao da leitura de cicatriz vacinal. O exame da cicatriz vacinal mostrou ser um bom indicador para avaliar a situacao vacinal referente ao BCG.

Underreporting of the tuberculosis and AIDS comorbidity: an application of the linkage method

OBJECTIVE To analyze the underreporting of the tuberculosis (TB) and AIDS comorbidity. METHODS Surveillance study using records from the Notifiable Diseases Information System - Tuberculosis and AIDS in Brazil from 2000 to 2005. Records of TB without information on the presence of Aids were considered to be underreporting of the comorbidity when paired off with AIDS records in which the year of diagnosis of AIDS was the same or previous to the year of reporting of TB, as well as records from the same patient whose previous records had this information. An indicator was created: recognized TB-AIDS comorbidity, based on the TB records that had information on the presence of AIDS. RESULTS The underreporting of TB-AIDS was 17.7%. This percentage varied between states. The incorporation of the underreported records into the previously recognized ones increased the proportion of TB-AIDS in Brazil from 6.9% to 8.4%. The highest proportions of underreporting were noted in Acre (Northern), Alagoas, Maranhão and Piauí (Northeastern) (more than 35% each) and the lowest in São Paulo (Southeastern) and Goiás (Central-western) (around 10% each). CONCLUSIONS The underreporting of the TB-AIDS comorbidity found in Brazil will probably trigger modifications in the surveillance system in order to provide information for the national programs.OBJETIVO: Analisar a subnotificacao da comorbidade tuberculose (TB) e aids. METODOS: Estudo de vigilância utilizando os registros do Sistema de Informacao de Agravos de Notificacao de Tuberculose e de aids no Brasil de 2000 a 2005. Registros de TB sem informacao da presenca de aids foram considerados subnotificacoes da comorbidade quando pareados a registros de aids que apresentassem ano de diagnostico de aids igual ou anterior ao ano de notificacao da TB, assim como os registros de um mesmo paciente cujos registros anteriores apresentavam essa informacao. Criou-se um indicador: comorbidade TB-aids reconhecida, a partir dos registros de TB com a informacao de presenca de aids. RESULTADOS: A subnotificacao de TB-aids foi de 17,7%. Esse percentual variou entre estados. A incorporacao dos registros subnotificados aos previamente reconhecidos elevou a proporcao de TB-aids no Brasil de 6,9% para 8,4%. As maiores proporcoes de subnotificacao foram observadas no Acre, Alagoas, Maranhao e Piaui (mais de 35% cada) e as menores em Sao Paulo e Goias (cerca de 10% cada). CONCLUSOES: A subnotificacao da comorbidade TB-aids encontrada no Brasil deve deflagrar modificacoes no sistema de vigilância para prover informacoes aos programas nacionais.

Accuracy of probabilistic and deterministic record linkage: the case of tuberculosis

ABSTRACT OBJECTIVE To analyze the accuracy of deterministic and probabilistic record linkage to identify TB duplicate records, as well as the characteristics of discordant pairs. METHODS The study analyzed all TB records from 2009 to 2011 in the state of Rio de Janeiro. A deterministic record linkage algorithm was developed using a set of 70 rules, based on the combination of fragments of the key variables with or without modification (Soundex or substring). Each rule was formed by three or more fragments. The probabilistic approach required a cutoff point for the score, above which the links would be automatically classified as belonging to the same individual. The cutoff point was obtained by linkage of the Notifiable Diseases Information System – Tuberculosis database with itself, subsequent manual review and ROC curves and precision-recall. Sensitivity and specificity for accurate analysis were calculated. RESULTS Accuracy ranged from 87.2% to 95.2% for sensitivity and 99.8% to 99.9% for specificity for probabilistic and deterministic record linkage, respectively. The occurrence of missing values for the key variables and the low percentage of similarity measure for name and date of birth were mainly responsible for the failure to identify records of the same individual with the techniques used. CONCLUSIONS The two techniques showed a high level of correlation for pair classification. Although deterministic linkage identified more duplicate records than probabilistic linkage, the latter retrieved records not identified by the former. User need and experience should be considered when choosing the best technique to be used.

Sensilidade e especificidade da leitura da cicatriz vacinol do BCG [Sensitivity and specificity of the BCG scar reading]

OBJECTIVE To validate the BCG scar as a marker of BCG vaccination status. METHODS A cross-sectional survey was carried out among 53,348 schoolchildren aged 6-14 years who underwent BCG scar examination as part of a large BCG vaccine trial taking place in the city of Manaus, Brazil. Results of BCG scar reading were compared with information on vaccine status of their vaccination cards or provided by parents or guardians. Double-reading was performed in a sub-sample. Data analysis was conducted using Stata 7 and Kappa coefficient. RESULTS Of 52,348 schoolchildren studied, vaccine status information from parents/guardian letters was available for 29,254 and from vaccination cards for 4,947. There was found a high agreement between the double-readings of the scars (Kappa=0.81). When the agreement between letter and card information was the gold standard, the sensitivity of BCG scar readings was 96.6% (95%CI 96.0-97.1) and the specificity was 71.1% (95%CI 55.7-83.7). The sensitivity was 96.1%, 97.3% and 95.3% for children vaccinated up to one month of age, four months and one year, respectively. CONCLUSIONS Sensitivity and specificity did not show an association with the childs age at the scar reading. BCG scar was a good marker of BCG vaccination status regardless of age - from the first years of life up to 14 years old.OBJETIVO: Validar a utilizacao da cicatriz vacinal de BCG como um indicador de vacinacao. METODOS: Foi realizado um estudo transversal em 52.348 escolares, entre 6 e 14 anos de idade, que possuiam exame de cicatriz vacinal do BCG e que participaram de um ensaio clinico randomizado e controlado na cidade de Manaus, Brasil. Os dados da leitura da cicatriz vacinal foram comparados com a informacao sobre a vacinacao passada fornecida pelos cartoes vacinais ou informacao dos responsaveis. Em uma subamostra foi realizada leitura dupla com calculo do coeficiente Kappa. Para analise dos dados utilizou-se o Stata 7. RESULTADOS: Do total de 52.348 escolares estudados, 29.254 possuiam informacao sobre cicatriz vacinal coletada por meio de carta aos pais, e 4.947 possuiam historia de vacinacao coletada pelo cartao de vacinas. Observou-se elevada concordância entre a dupla leitura de cicatriz vacinal (Kappa =0,81). A sensibilidade da leitura de cicatriz vacinal foi 96,6% (95% IC 96,0-97,1) e a especificidade foi 71,1% (95% IC 55,7-83,7) quando o padrao ouro utilizado foi a concordância entre a carta aos pais e a informacao do cartao de vacinas. A sensibilidade foi de 96,1%, 97,3% e 95,3% para criancas vacinadas ate um mes de idade, ate 4 meses e ate um ano de idade, respectivamente. CONCLUSOES: Os valores encontrados para sensibilidade e especificidade foram independentes da idade da realizacao da leitura de cicatriz vacinal. O exame da cicatriz vacinal mostrou ser um bom indicador para avaliar a situacao vacinal referente ao BCG.