SciELO - Scientific Electronic Library Online

vol.8 issue6Improving survival among Brazilian children with perinatally-acquired AIDSThe influence of occult infection with hepatitis B virus on liver histology and response to interferon treatment in chronic hepatitis C patients author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Brazilian Journal of Infectious Diseases

Print version ISSN 1413-8670On-line version ISSN 1678-4391

Braz J Infect Dis vol.8 no.6 Salvador Dec. 2004 



Risk of tuberculosis among household contacts in Salvador, Bahia



Antonio Carlos LemosI, II; Eliana D MatosII; Diana B Pedral-SampaioI; Eduardo M. NettoI

IProf. Edgard Santos Universitary Hospital, Federal University of Bahia
Octávio Mangabeira Hospital, Secretary of the State of Bahia, Salvador/BA, Brazil





Tuberculosis is one of the most important infectious diseases in the world. Only 68% of the estimated new tuberculosis (TB) cases in Brazil are diagnosed. Our aim was to determine the risk of infection among household contacts. Study design. Cohort of tuberculin-negative household contacts followed for 12 Months. Methods. Household contacts of randomly selected index acid-fast bacilli (AFB)-positive TB cases were evaluated through clinical examination, thorax X-ray, tuberculin, AFB smear and culture. Contacts with a negative response to the tuberculin test (less than 10 mm diameter) were retested after 90 days. Tuberculin reversal (used as a parameter of infection risk) was defined as an increase of at least 10 mm from the last measurement. Results. 269 household contacts were followed. The prevalence of disease in this population was 3.7%. The prevalence of infection after the 12-month follow-up period was 63.9%. The risk of infection was 31.1% within 120 ± 48 days. Conclusion. Household contacts of AFB positive tuberculosis patients have a very high prevalence and risk of tuberculosis infection. TB preventive or therapeutic measures directed towards this group should be implemented in Brazil.

Key Words: Tuberculosis, infection, epidemiology, prevalence.



Tuberculosis is presently one of the main causes of morbi-mortality in the developing world. It is estimated that 8.3 million new tuberculosis (TB) cases occurred in 2000, including 3.5 million sputum acid-fast bacilli (AFB) positive diagnoses and 1.8 million deaths. One third of the world population is already infected [1]. In 1993, the World Health Organization (WHO) declared the tuberculosis problem as a "global emergency" [2]. Along with this crisis is the problem of the M. tuberculosis being resistant to several drugs. The purported causes for the permanence and even the increase of TB in the world are: (a) migration of populations from highly prevalent areas; (b) the HIV epidemic; (c) impoverishment of several populations, mainly in the underdeveloped world; and finally (d) increased transmission in closed environments (including hospitals, prisons, host houses and poorly ventilated houses hosting index cases) [3-5].

In Brazil, the prevalence of infection due to tuberculosis in the overall population is quite high; it is presumed to be over 50%. It is estimated that only 68% of the new cases of patients with symptomatic TB are officially notified [1]. The investigation of household contacts is a recommendation of the National Program for the Control of Tuberculosis [6]. However, this investigation has not been a priority of Brazilian public services. In 1997, only 10.1% of the estimated contacts were investigated [7]. With the objective of reexamining the role of household contacts, we investigated the risk of transmission in the area of Salvador, the coastal capital city of Bahia, with 2.5 million inhabitants, having an HIV-prevalence rate of 133 cases/100,000 inhabitants [8].


Material and Methods

From March/1997 to April/1999, 69 patients, diagnosed by AFB positive sputum smear or by culture for Mycobacterium tuberculosis, were randomly selected to participate in the cohort. They were selected from the outpatient clinic of the state reference hospital for tuberculosis treatment in Bahia - Octavio Mangabeira Hospital. The inclusion criteria were: (a) being the first TB case in the household (called "index case"); (b) being HIV-negative; (c) living in Salvador; and (d) never having been treated for TB. To improve treatment adherence, free food stamps were offered.

The contact was followed for at least 12 months. The visits were set to be every four months, when a clinical examination was made, an X-ray was taken and a tuberculin skin test (TST) was done to evaluate the infection rate. The TST was done on all contacts at the first meeting. The patient with an initial negative skin test performed a second test (in the contra lateral forearm) after at least 90 days following the first test, and repeated it again during the third visit if it was still negative. A TST was considered positive when an induration equal to or greater than 10 mm diameter was detected. A patient was considered newly reactive when there was an increase in the reaction of at least 10 mm compared to the previous reaction measurement. Since children in Brazil are routinely vaccinated against tuberculosis with BCG within 30 days after birth, the patients were divided into vaccinated (presence of a scar in the outface of the right arm) or not.

If there were any respiratory symptoms, a sputum smear was taken and cultured. All sputum samples were cultivated in Lowestein-Jaensen media. The individual was classified as "TB case" if he (she) had a positive AFB smear or TB culture and/or a clinical picture of TB, a suggestive thorax X-ray plus clinical improvement with TB treatment within 30 days. All patients weighing 45 Kg or more were treated with Isoniazid (400 mg/day) plus Rifampim (600 mg/day) for six months, plus Pyrazinamide (2g/day) for the initial two months. Patients weighing less than 45 Kg had their dosages adjusted based on their weight, according to the Brazilian Ministry of Health guidelines for TB treatment.



Of the selected 69 index cases treated, 68 patients completed the TB treatment and were diagnosed as cured. One abandoned the treatment after two months of therapy. Among the 69 index cases, 282 household contacts were found; 269 (95.4%) contacts completed the follow-up. The mean follow-up time was 346 ± 154 days (median: 314 days). Table 1 describes the demographic characteristics of the index cases and their contacts. Contacts with the BCG vaccination scar were younger than those without a scar (17.1 ± 10.0 versus 38.0 ± 17.8 years-old; p<0.0001). An increased disease or infection risk was not found among the contacts when they were stratified according to the amount (months) of exposure time to the infectious index case (less than 1 month, between 1 and 3 months and more than that) (p>0.05).



Table 2 describes the prevalence and incidence among the cohort of contacts; seven cases were found with TB disease at the first evaluation and three more during the follow-up (Figure 1). Infection was much more prevalent and 136 out of 269 contacts were found to be infected. During the follow-up, 36 more contacts were infected, with a cumulative prevalence of 63.9%. The risk of TB disease among the contacts was 1.1% and the risk of infection was 31.1%. There were no differences in the prevalence, the risk of infection or the development of disease among the contacts who were vaccinated or not.





The mean time between the first and second evaluation was 120 ± 48 days (median: 130 days). Thirty-six (31.1%) contacts converted during follow-up of the 117 negative TST contacts followed. During this time, 33 contacts converted into positive; this represented 91.7% (33/36) of those who would convert into positive at any time. Only three more contacts converted the skin test between the second and third evaluation. After the third evaluation, there were no more skin test conversions.



Ten cases of active tuberculosis disease were found during the follow-up of one year of the 282 contacts of 69 AFB-positive index cases. This represents a cumulative prevalence of 3.7% or 3,745 cases per 100,000 habitants, which is 50 times the annual incidence rate in Brazil of 75/100,000. This finding is similar to that reported by Picon et al. [9] (3.3%), similar to that found by Lapa et al. [10] (4.2%), greater than the 0.96% found by Carvalho et al. [11], and less than that found by Menzies et al. [12] 11%, Lemos and Matos [13] in a similar population (8.4%) and Vidal et al. [14] (7.6%). Sundaram and Driver, in a review of a series of 21 studies from 1940-66, found a prevalence of 3.1 to 24.2% [15]. These rate variations could be due to methodology differences; in our previous report that demonstrated a prevalence of 8.4%, higher than that found in this study, the patients were invited to participate in the evaluation in a non-sequential and non-random way and there was no incentive for adherence [13]. Much lower adherence to the follow-up was reported by Askew (87% in children, 13% in adults) [16], Garcia Ordonez (26.9%) [17], Picon (57.0%) [9] and Drobniewski (85.0%) [18] This may have lead to a selection of those index cases that had symptomatic individuals at home. In our study, only 4.6% of the contacts did not adhere to the proposed follow-up as we supported them with specialized social and psychological care besides offering food stamps to the most needy ones. The micro epidemics in Menzies's report [12] increased the prevalence of TB. Contact definition differences, such as the nuclear family, which is more prevalent in economically advanced countries versus enlarged families that are prevalent in Africa and Brazil can affect the results; in our cohort only 25.3% of the contacts were not related to the index cases. The inclusion of other than AFB-positive index cases significantly diminished the transmission in some studies [9,12,14]. BCG vaccination among the contacts could also have an effect. Studies on TB vaccination are still controversial; they have demonstrated from no [19] to high protection [20], or protection against only severe forms of TB [21], mostly in infants [22]. We found no additional protection in vaccinated individuals, similar to other reports [11,15,23,24].

Longer exposure may increase the prevalence among contacts, as detected by studies with multidrug-resistant TB cases, whose natural course is infectious for longer periods [25 - 27]. On the other hand, the infectiousness diminishes greatly after the beginning of treatment [28-30]. Similar prevalence rates were found by Picon, Lapa & Silva and Carvalho [9,10,11]. We found no correlation between time (in months) of exposure and increased disease risk.

A cut-off of 10 mm to consider a PPD positive test was used for the diagnosis of prevalence and risk of infection, and an additional 10 millimeters over the initial test result was considered a positive test for the second and following ones. These values are greater than those used in other places where BCG vaccination is not given and the prevalence of non-tuberculosis mycobacterium is low [31,32], different from Brazil [33,34]. This strategy decreased the detection of false TB infections.

We found a point-prevalence infection of 52.3% and a cumulative prevalence of 63.9%. These findings are 2.1 and 2.5 times the estimated prevalence in the Brazilian population [1], but are similar to other authors' findings [11,13,15,23,34]. However our prevalence rates were significantly lower the than those of contacts of multidrug-resistant patients (p=0.0001) (87%) [25]. Regarding infectiousness, only AFB-positive patients are important among contacts [27]. The index AFB positive patient has a greater importance for contact sickness, however, if the index patient is AFB negative, but culture positive, the infection rate is similar to that provoked by AFB positive patients.

The incidence of infectious TB can be calculated from the Annual Risk of Infection (ARI). An increase of 1 point percentage in the ARI has been associated with an increase of 49 smear-positive cases per 100,000 population [1]. The risk of infection during the 346 days of follow-up was 31.1%; using this rule-of-thumb, we project 1524 cases/100,000 population. The ARI was similar to that found by Carvalho (26.3%)[18]. The estimated ARI for Brazil is 0.8% [34], so our calculated ARI among the contacts was significantly higher.

Given similar risks of infection, younger contacts have less chance of being infected by M. tuberculosis; however this is the age at which most of the contacts were vaccinated in Brazil. We found a mean age of vaccinated subjects of only 17.1 years, versus 38 years for non-vaccinated individuals (p = 0.0005). On the other hand, we found that the chance of being recently infected (risk of infection), in this contact cohort, was the same and independent of age.



The contacts evaluated in this study were subjected to a high risk for developing active tuberculosis. The prevalence of infection was 2.5 times higher than for the general population, and the ARI was almost 40 times than that estimated for Brazil. This warrants a much better contact tracing and treatment program than what is currently implemented.



This paper was supported by NUPEP - "Núcleo de Pesquisa em Pneumologia" Tuberculosis Reference Center - State of Bahia Department of Health.



1. Dye C., Scheele S., Dolin P., et al. Global Burden of Tuberculosis. Estimated incidence, prevalence, and mortality by country. JAMA 1999;282:677-86.         [ Links ]

2. WHO. Tuberculosis. WHO Fact sheet, 1996. Report, 1996. World Health Organization. Found at: Consulted in 1999.         [ Links ]

3. Corbett E.L., Watt C.J., Walker N., et al. The growing burden of tuberculosis: Global trends and interactions with the HIV epidemic. Arch Intern Med. 2003;163:1009-21.         [ Links ]

4. Raviglione M.C., Sudre P., Rieder H.L., et al. Secular trends of tuberculosis in western Europe. Bull WHO, 1993;71:297-306.        [ Links ]

5. Raviglione M.C., Snider Junior D.E., Kochi A. Global epidemiology of tuberculosis. Morbidity and mortality of a worldwide epidemic. JAMA 1995;273:220-6.         [ Links ]

6. Manual de Normas para o Controle da Tuberculose. 4. ed. Brasília. Centro Nacional de epidemiologia e Coordenação de Pneumologia Sanitária, Ministério da Saúde - Brasil, 1995.         [ Links ]

7. Notificação da tuberculose no Estado da Bahia, Salvador-Ba, Secretaria Estadual da Saúde, 1998.        [ Links ]

8. Boletim epidemiológico AIDS/Ministério da Saúde. Coordenação Nacional de DST e Aids. - Brasília: Ministério da Saúde, 1987, Ano XVI, n. 1 14a a 52a semanas epidemiológicas abril a dezembro de 2002.         [ Links ]

9. Picon P.D., Della-Guistina M.L., Moraeo de Melo M.D. Avaliação de contatos domiciliares em condições de rotina na unidade 3 de TB do Rio Grande do Sul. Secretaria Estadual de Saúde e do Meio Ambiente do Rio Grande do Sul. 6p, 1995 [not printed].         [ Links ]

10. Lapa E., Silva J.R., Cardoso A.P., et al. Estudo de comunicantes em tuberculose: características dos comunicantes que adoeceram. J Pneumol 1984;10:253.         [ Links ]

11. Carvalho A.C.C. Infecciosidade do paciente com tuberculose pulmonar infectado pelo HIV. Tese de doutorado. Universidade Federal do Rio de Janeiro, Rio de Janeiro, 122p, 1999.         [ Links ]

12. Menzies D. Issues in the management of contacts of patients with active pulmonary tuberculosis. Can J Pub Health 1997;88:197-201.         [ Links ]

13. Lemos A.C.M., Matos E.D. Rastreamento dos comunicantes de 47 casos índice de tuberculose pulmonar. In: Anais do XXIV Congresso Brasileiro de Pneumologia e Tsiologia. Recife. Jornal de Pneumologia, 1988;14 (suplemento número 1):151-2.         [ Links ]

14. Vidal R., Miravitlles M., Cayla J.A., et al. Increased risk of tuberculosis transmission in families with microepidemics. European Respiratory Journal 1997;10:1327-31.         [ Links ]

15. Sundaram & Driver. Investigation of contacts to tuberculosis cases. New York: Bureau of Tuberculosis Control, 1996.        [ Links ]

16. Askew G.L., Finelli L., Hutton M., et al. Mycobacterium tuberculosis transmission from a pediatrician to patients. Pediatrics 1997;100:19-23.         [ Links ]

17. Garcia Ordonez A.L., Buznego Alvarez B., Alonso Fernandez M., et al. Analysis of a study of the contacts of pulmonary tuberculosis patients in a health area during 1991 and 1992. Atenção Primaria, 1994;14:733-6.         [ Links ]

18. Drobniewski F.A., Ferguson J., Barritt K., et al. Follow up of an immunocompromised contact group of a case of open pulmonary tuberculosis on a renal unit. Thorax 1995;50:863-8.         [ Links ]

19. WHO. Trial of BCG vaccines in south India for tuberculosis prevention: first report. Bull WHO 1979;57:819-27.         [ Links ]

20. Quarterly data on tuberculin testing and BCG vaccination. World Health Stat Rep 1976;29:21-5.         [ Links ]

21. Divisão Nacional de Pneumologia Sanitária. Segundo informe técnico sobre vacinação/revacinação com BCG. Ministério da Saúde do Brasil Brasília, pp 11-7, 1994.         [ Links ]

22. Secretaria Nacional de Programas Especiais. A vacina BCG. Controle da tuberculose, uma proposta de integração ensino e serviço. 1. ed., Ministério da Saúde do Brasil, Brasília, pp 71-5, 1987.         [ Links ]

23. Godoy P., Diaz J.M., Alvarez P., et al. Brote de tuberculosis: importancia del tiempo de exposición frente a la proximidad a la fuente de infección. Med Clin (Barc) 1997;108:414-8.         [ Links ]

24. Lemos A.C.M., Matos E.D. Avaliação dos comunicantes de TP vacinados com BCG intradérmico. In: Anais do XXIV Congresso Brasileiro de Pneumologia e Tisiologia. Recife. Jornal de Pneumologia 1988;14(Suplemento número 1):152.         [ Links ]

25. Kritski A.L., Marques M.J., Rabahi M.F., et al. Transmission of tuberculosis to close contacts of patients with multidrug-resistant tuberculosis. Am J Resp Crit Care Med 1996;153:331-5.         [ Links ]

26. Fiuza de Melo F.A., Seiscentos M., Ide Neto J., et al. Estudo de comunicantes de pacientes com tuberculose resistente, no Estado de São Paulo, São Paulo, Instituto Clemente Ferreira, 1994 [Not printed].         [ Links ]

27. Rose Junior C.E., Zerbe G.O., Lantz S.O., Bailey W.C. Establishing priority during investigation of tuberculosis contacts. Am Rev Respir Dis 1979;119: 603-9.         [ Links ]

28. Hopewell P.C. Factors influencing the transmission and infectivity of Mycobacterium tuberculosis: implications for clinical and public health management. In Sande MA, Hudson LD, Root RK (eds), Respiratory Infections. New York: Churchill Livingstone, pp 191-216, 1986.         [ Links ]

29. Sultan I., Nyka W., O'grady F., et al. Tuberculosis disseminators. Am Rev Resp Dis 1960;82:358-69.         [ Links ]

30. Kamat S.R., Dawson J.J., Devadatta S., et al. A controlled study of the influence of segregation of tuberculous patients for one year on the attack rate of tuberculosis in a 5-year period in close family contacts in South India. Bull WHO 1966;34:517-32.         [ Links ]

31. CDC. Prevention and control of tuberculosis in correctional facilities: recommendations of Advisory Council for the Elimination of Tuberculosis (ACET). MMWR 1996;45:1-27.         [ Links ]

32. CDC. Use of BCG vaccines in the control of tuberculosis: a joint statement by the ACIP and the Advisory ACIP Committee for elimination of tuberculosis. Centers for Disease Control (CDC). JAMA 1998;260:2983-91.         [ Links ]

33. Serviço de Vigilância Epidemiológica. Dados epidemiológicos da tuberculose. Boletim Epidemiológico, Ministério da Saúde, v. 9: pp 12, 1995.         [ Links ]

34. Feitosa J.V.P. Aspectos epidemiológicos da tuberculose numa população favelada, favela do Escondidinho, Rio de Janeiro. Dissertação de Mestrado, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 1985.         [ Links ]



Correspondence to
Dr. Eduardo M Netto, MD, MPH
Hospital Universitário Prof. Edgard Santos/Universidade Federal da Bahia
Rua João das Botas, S/N 6o andar, Sala 1124 - Canela
Zip code: 40110-160 - Salvador - Bahia - Brazil

Received on 16 July 2004; revised 13 December 2004.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License