Tuberculosis is a major opportunistic infection in persons infected with HIV . Following treatment with short-course chemotherapy, patients have a small risk for developing recurrent tuberculosis as a result of a relapse or exogenous reinfection [2,3].
Recurrent tuberculosis is a significant problem for tuberculosis control programs, as treatment for recurrent episodes is often associated with drug resistance and low cure rates [4–6]. Recent evidence suggests that rates of recurrent disease can be significantly greater than rates of new tuberculosis in high incidence areas, even among patients who successfully complete therapy [7–9].
Risk factors for recurrence due to relapse include poor tuberculosis treatment adherence at initial diagnosis , residual cavitation on chest radiograph [2,3] and positive sputum culture at 2 months of treatment [11,12]. A primary risk factor for both relapse and reinfection is HIV-infection [2,3,10,13]. Among HIV-infected populations, low initial CD4 cell count  and receiving less than 37 weeks of antituberculosis therapy [10,13] have been observed to increase recurrence.
The frequency of and risk factors for recurrent tuberculosis in HIV-infected populations with access to antiretroviral therapy (ART) have not been determined. Although ART reduces the risk of initial episodes of tuberculosis  and decreases rates of recurrence of other HIV-related opportunistic infections, its impact on subsequent tuberculosis is not known. The aim of this analysis was to measure the magnitude of and determine risk factors associated with tuberculosis recurrence in a cohort of HIV-infected patients in Rio de Janeiro, Brazil.
The study was conducted within the Tuberculosis and HIV in Rio (THRio) study, as previously described [14,15]. In brief, data were abstracted from medical records of HIV-infected patients attending 29 HIV clinics in Rio de Janeiro who had at least one clinic visit between September 2003 and September 2005. Chart abstractions are conducted semiannually at all participating clinics. Information routinely collected includes demographic data, dates of HIV diagnosis, opportunistic infections including tuberculosis, treatment history for HIV, and tuberculosis, and results of diagnostic tests including CD4 cell counts, HIV viral loads, and tuberculin skin tests. Although patients were required to have had a clinic visit in the 2003–2005 time period, all clinical information from the date of the first clinic visit, which may have occurred years earlier, was collected. Analysis of the baseline data collected as part of the THRio study has been reported . The data for this analysis were updated through February 2008.
Eligible patients were those who were diagnosed with a first episode of tuberculosis after their initial clinic visit. According to Brazilian surveillance definitions, tuberculosis is diagnosed in patients presenting with signs and symptoms compatible with the disease on the basis of chest radiographs, sputum acid-fast bacilli smears, and response to antituberculosis therapy . Sputum smear results were not regularly recorded and have not been included in analyses.
Recurrent tuberculosis was defined according to the Brazilian national guidelines as a subsequent diagnosis of tuberculosis at least 9 months after the initial diagnosis . Patients who had less than 270 days of follow-up after their initial tuberculosis diagnosis were excluded because they did not have enough follow-up time to record a recurrent tuberculosis event. Beginning with the latter of 1 January 1998 or their initial visit to the HIV clinic, patients with a recorded tuberculosis diagnosis were followed to the earlier of date of a second tuberculosis diagnosis or last recorded clinic visit date. Because individuals are not at risk for a recurrent tuberculosis diagnosis until at least 270 days following their initial tuberculosis diagnosis and to allow equal time for patients to have completed their antituberculosis therapy, person-time was calculated beginning 270 days after the initial tuberculosis diagnosis through a second tuberculosis event or the last recorded date in their medical record. All recurrent tuberculosis diagnoses reported in the medical records were abstracted, but only the first recurrent tuberculosis diagnosis was evaluated in the current analysis. We calculated recurrent tuberculosis rates for population subgroups and determined risk factors associated with recurrence.
Baseline and follow-up CD4 data were those reported closest in time to the initial episode of tuberculosis and the time of a recurrent tuberculosis diagnosis, respectively. CD4 cell counts are routinely determined between three and four times per year for HIV-infected patients in Rio de Janeiro, thus patients had as many as 34 recorded CD4 cell counts. All CD4 cell counts were used in the analyses as discussed below. ART was treated as a time-dependent covariate, coded as zero until the initiation of treatment and coded as one thereafter.
Differences in categorical variables were assessed by χ2 tests. The Mann–Whitney U-test was used to compare between-group distributions for unpaired data and the Wilcoxon signed-rank test for paired data. Exact confidence intervals (CI) for incidence rates were calculated based on the Poisson distribution. Follow-up time started at 270 days following first tuberculosis diagnosis and ended on the earliest of date of a second recorded tuberculosis diagnosis (recurrence) or last recorded clinic date (censored without recurrence). Kaplan–Meier curves for the primary outcome were generated. Cox proportional hazards regression models evaluated unadjusted and adjusted associations of outcome of initial tuberculosis treatment, ART, time on ART as a time-dependent variable, age at initial tuberculosis diagnosis, sex, and baseline, and time-dependent CD4 cell counts.
Unadjusted proportional hazards models were fit using each risk factor of interest as the only variable in the model. Using covariates significant at the 0.10 level in the unadjusted models, we generated two fully adjusted proportional hazards models. In model 1, CD4 at initial tuberculosis diagnosis was included and in model 2, CD4 was included as a time-dependent variable. All analyses were conducted using SAS (Version 9.1; SAS Institute Inc., Cary, North Carolina, USA) and Stata (Version 9.1; STATA Corp., College Station, Texas, USA).
There were 15 252 HIV-infected patients in the THRio database as of 31 December 2007. Of these, 1961 had a tuberculosis diagnosis prior to their first recorded clinic date and were excluded from the current analysis. Among the remaining 13 291 patients (60 713 person years), there were 1396 patients with at least one incident tuberculosis diagnosis for an incidence rate of 2.3/100 person years.
Three hundred and sixteen tuberculosis patients were excluded because they had fewer than 270 days of follow-up after their initial tuberculosis diagnosis. Among the remaining 1080 tuberculosis patients, 96 (8.9%) developed a recurrent diagnosis at least 270 days later. The median time between diagnoses was 2.4 years [interquartile range (IQR), 1.5–4.1 years]. Multiple recurrent diagnoses were made in 18 patients (14 patients had two recurrences; four patients had three recurrences), but analyses were limited to first recurrent tuberculosis diagnosis.
Among the 96 patients with recurrent tuberculosis, 63 (66%) were male and the median age at first tuberculosis diagnosis was 33.9 years (Table 1). Patients without a recurrent event were slightly older (median age, 36.8 years; P = 0.004). Fewer patients with recurrence received ART prior to their recurrent diagnosis compared with patients without recurrence (P < 0.001). Treatment of the initial episode of tuberculosis was completed by fewer patients with recurrence compared with patients without recurrent tuberculosis (P = 0.02).
The 1080 patients contributed 3370 years of follow-up and the incidence rate of recurrent tuberculosis was 2.8/100 person years. Incidence did not differ by sex, though older patients had decreased rates of recurrence compared with younger patients (Table 2). Patients receiving ART had half the risk of recurrence compared with those not receiving ART [incidence rate ratio (IRR), 0.51; 95% CI, 0.29–0.93]. Patients with incomplete treatment at their initial tuberculosis diagnosis had an incidence of recurrence of 9.0/100 person years (95% CI 5.0–14.8) and were three times as likely to have a recurrent diagnosis (IRR, 3.60; 95% CI 1.92–6.33) compared with those who completed treatment at initial diagnosis (Fig. 1). Patients whose tuberculosis treatment status was unknown had similar rates of recurrence [incidence rate (IR), 3.1/100 person years] compared with those with treatment completion (IR, 2.5/100 person years; IRR, 1.25; 95% CI 0.44–2.85).
We investigated the association between CD4 cell count and recurrent tuberculosis in several ways. The median CD4 cell count at first tuberculosis diagnosis for patients who had a recurrent diagnosis was 197 mm−3 (IQR, 108–366) compared with 192 mm−3 (IQR, 91–341) for patients who did not experience a recurrence (P = 0.99). For those with recurrence, there was no change between the median baseline CD4 cell counts and median CD4 cell count at the time of their recurrent diagnosis (P = 0.80) (Table 3). Among patients who did not have a recurrent tuberculosis diagnosis, the median CD4 cell count doubled from 192 mm−3 (IQR, 91–341) at initial tuberculosis diagnosis to 382 mm−3 (IQR, 221–579) at the end of follow-up (P < 0.001). Patients whose CD4 cell counts declined or remained the same were more than five times as likely to have a recurrent diagnosis compared with patients with an increasing CD4 cell count (IRR, 5.61; 95% CI 3.55–9.08).
Multivariate Cox proportional hazard modeling was consistent with the univariate results (Table 4). In model 1, CD4 cell count measured at first tuberculosis diagnosis was not predictive of increased risk of recurrence. With CD4 cell count observed as a time-dependent variable in model 2, a CD4 cell count between 200 and 349 (cells mm−3) reduced recurrence 65% compared with those with a CD4 cell count less than 200 (cells mm−3) [adjusted hazard ratio (aHR), 0.35; 95% CI, 0.20–0.60]. Similar trends were seen with a CD4 cell count between 350 and 499 (cells mm−3) (aHR, 0.42; 95% CI, 0.24–0.74) and CD4 cell count greater than 500 (cells mm−3) (aHR, 0.25; 95% CI, 0.14–0.48). Use of ART reduced recurrence by 50% in both models. Patients with incomplete treatment at initial tuberculosis diagnosis had a three-fold increased risk of recurrence in both models, whereas those with unknown treatment results had a minimal increase that was not statistically significant in either model.
We found recurrent tuberculosis occurred in 8.9% of HIV-infected patients in Rio de Janeiro, at a rate of 2.8/100 person years. This rate was slightly higher than the overall tuberculosis incidence in all HIV-infected patients in this cohort . The major predictors of recurrent tuberculosis were failure to complete initial tuberculosis therapy, not receiving ART and lower CD4 cell counts. Older age (40–49 years) was associated with reduced risk of recurrence.
Although recurrence was greatest among those who did not complete treatment at initial diagnosis (17.2%) in our study, recurrent tuberculosis occurred frequently in treatment completers (8.1%) and those whose treatment status was unknown (8.5%) as well. Our overall recurrence rate of 2.8/100 person years and subgroup rate of 2.5/100 person years in patients with known treatment completion are one fourth of that detected in a cohort of South African mineworkers . However, tuberculosis rates in South African gold-mine workers are considerably higher than those of HIV-infected patients in Rio de Janeiro, Brazil primarily due to pulmonary fibrosis. A community study in Cape Town, South Africa where tuberculosis incidence was 313/100 000, reported recurrence rates of 2.7/100 person years following successful treatment and 6.5/100 person years among treatment failures, though HIV status was unknown for most patients. These rates compare with our rates in Brazil, though our recurrence among treatment failures was even higher (9.0/100 person years).
A study in the United States found that tuberculosis patients who were treated with suboptimal regimens had a six-fold increased risk of recurrence compared with those completing an optimal regimen , though few of their patients had HIV/AIDS and HIV/AIDS was not associated with recurrence. Another US study in the pre-HIV era reported that 77% of patients with recurrent tuberculosis did not initially receive antituberculosis therapy, had been prescribed inadequate or inappropriate therapy, or had been nonadherent to their prescribed regimen . In this study, all patients started therapy with the standard Brazilian three-drug regimen, which is effective for drug-susceptible disease but may not be adequate for drug-resistant tuberculosis. Although directly observed therapy has been increasingly used in Rio de Janeiro, not all patients would have had access to this modality and we cannot calculate the proportion of doses that individual patients may have missed, though we are able to determine which patients completed treatment.
The use of ART and increases in CD4 cell count were both strongly associated with protection against recurrence. ART has been shown to reduce the incidence of first episodes of tuberculosis in HIV-infected patients [14,19–22], and it also clearly reduces the risk of recurrence of other opportunistic diseases, such as Mycobacterium avium[23,24], cryptococcosis , and Pneumocystis jiroveci[23,24]. In the present study, CD4 cell counts at the time of initial tuberculosis diagnosis did not predict the likelihood of recurrence. Rather, change in median CD4 cell count after the initial diagnosis was important. For those who experienced a recurrence, the median CD4 cell counts were essentially unchanged, whereas for those who did not have recurrent disease the median CD4 cell count during follow-up doubled. Lawn et al. found that among HIV-infected patients receiving ART in Cape Town, those who developed tuberculosis had significantly impaired CD4 cell responses compared with patients who remained free of tuberculosis. These results underscore the importance or restoration of cellular immunity as a mechanism for reducing the risk of tuberculosis and other opportunistic diseases. We did not have sufficient numbers of viral load determinations to assess the impact of changes in viral burden on risk of recurrence, but others have suggested that this may be an important factor in the probability of experiencing opportunistic infections while receiving ART .
A substantial proportion of recurrent tuberculosis in HIV-infected patients and in high-incidence countries has been attributed to exogenous reinfection . We did not have the capacity to investigate the extent of exogenous reinfection in our cohort, as Mycobacterium tuberculosis isolates were not available for genotyping. However, based on the increased risk seen in patients who failed to complete their initial treatment regimen, we expect that many cases were relapses rather than reinfections. Nonetheless, a proportion of our recurrent cases could have been due to reinfection, and it is likely that ART and higher CD4 cell counts protect against progression of a new infection to active disease.
Provision of secondary isoniazid preventive therapy after treatment completion has been shown to reduce a subsequent tuberculosis diagnosis among HIV-infected populations in high incidence settings [27–29]. In our population, rate of recurrent tuberculosis was higher than the rate of initial tuberculosis (2.8 versus 2.3/100 person years). These results suggest that secondary preventive therapy may also be warranted in this setting to reduce the overall tuberculosis burden. Churchyard et al. suggest that secondary preventive therapy may only be effective for persons with only one previous episode of tuberculosis due to the increased risk of drug resistance after multiple recurrences. We do not have drug resistance data for our patients, but levels of resistance are likely higher in retreatment cases than in new cases. However, in another study in Rio de Janeiro, treatment success in retreatment cases given community-based directly observed therapy was 83%, suggesting that the prevalence of drug resistance is not extremely high . Research on the role of resistance in recurrent tuberculosis in this setting would be useful.
Because we found that recurrent tuberculosis was more common in patients whose CD4 cell counts did not increase, targeting these individuals for secondary preventive therapy might be an efficient strategy. One potential strategy for secondary preventive therapy would be to provide secondary isoniazid preventive therapy (IPT) to those persons with declining CD4 cell counts over a 6-month period following treatment completion for initial tuberculosis diagnosis. Timing of secondary IPT needs further investigation, as does determining the longevity of protection provided by secondary IPT in various populations and settings.
Interestingly, in our study population, tuberculosis incidence rates among those with a negative tuberculin skin test (TST) (2.0/100 person years) or an unknown TST result (2.4/100 person years) were similar to the entire population, suggesting that IPT should not be limited to those with a known positive TST [31–36]. The World Health Organization recommends IPT for HIV-infected patients in regions with high tuberculosis prevalence regardless of TST results [37,38].
Despite the success of ART in reducing tuberculosis incidence in HIV-infected populations, rates remain unacceptably high. Tuberculosis recurrence is an important contributor to the overall tuberculosis burden in settings in which coinfection is common, even where ART is available. Interventions to reduce initial and recurrent tuberculosis in HIV-infected populations are crucial if the burden of HIV-associated tuberculosis is to decline significantly.
J.E.G. conducted the statistical analysis, drafted the manuscript, and helped design the THRio study. B.D. helped design and implement the THRio study. B.S.K. implemented the THRio study and supervised data management. S.C.C. helped implement and design the THRio study. A.G.P. performed statistical analyses. L.H.M. helped design the THRio study design, provided statistical supervision, and helped write the manuscript. R.D.M. helped design the THRio study and assisted in data analyses. R.E.C. secured funding, helped design the THRio study, and assisted in writing of the manuscript. V.S. implemented the THRio study, helped with interpretation of the data, and writing of the manuscript. All authors reviewed and approved the final manuscript.
Sources of financial support: Support for this work was provided by the Bill & Melinda Gates Foundation grant for the Consortium to Respond Effectively to the AIDS-TB Epidemic (CREATE) and National Institutes of Health grants AI066994 and AI001637.
1. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, et al
. The growing burden of tuberculosis
: global trends and interactions with the HIV epidemic. Arch Intern Med 2003; 163:1009–1021.
2. Mallory KF, Churchyard GJ, Kleinschmidt I, De Cock KM, Corbett EL. The impact of HIV infection on recurrence of tuberculosis
in South African gold miners. Int J Tuberc Lung Dis 2000; 4:455–462.
3. Sonnenberg P, Murray J, Glynn JR, Shearer S, Kambashi B, Godfrey-Faussett P. HIV-1 and recurrence, relapse, and reinfection of tuberculosis
after cure: a cohort study in South African mineworkers. Lancet 2001; 358:1687–1693.
4. Johnson J, Kagal A, Bharadwaj R. Factors associated with drug resistance in pulmonary tuberculosis
. Indian J Chest Dis Allied Sci 2003; 45:105–109.
5. Fujiwara PI, Cook SV, Rutherford CM, Crawford JT, Glickman SE, Kreiswirth BN, et al
. A continuing survey of drug-resistant tuberculosis
, New York City, April 1994. Arch Intern Med 1997; 157:531–536.
6. Garcia-Garcia ML, Ponce de LA, Jimenez-Corona ME, Jimenez-Corona A, Palacios-Martinez M, Balandrano-Campos S, et al
. Clinical consequences and transmissibility of drug-resistant tuberculosis
in southern Mexico. Arch Intern Med 2000; 160:630–636.
7. Verver S, Warren RM, Beyers N, Richardson M, van der Spuy GD, Borgdorff MW, et al
. Rate of reinfection tuberculosis
after successful treatment is higher than rate of new tuberculosis
. Am J Respir Crit Care Med 2005; 171:1430–1435.
8. van Rie A, Warren R, Richardson M, Victor TC, Gie RP, Enarson DA, et al
. Exogenous reinfection as a cause of recurrent tuberculosis
after curative treatment. N Engl J Med 1999; 341:1174–1179.
9. Panjabi R, Comstock GW, Golub JE. Recurrent tuberculosis
and its risk factors: adequately treated patients are still at high risk. Int J Tuberc Lung Dis 2007; 11:828–837.
10. Driver CR, Munsiff SS, Li J, Kundamal N, Osahan SS. Relapse in persons treated for drug-susceptible tuberculosis
in a population with high coinfection with human immunodeficiency virus in New York City. Clin Infect Dis 2001; 33:1762–1769.
11. Benator D, Bhattacharya M, Bozeman L, Burman W, Cantazaro A, Chaisson R, et al
. Rifapentine and isoniazid once a week versus rifampicin and isoniazid twice a week for treatment of drug-susceptible pulmonary tuberculosis
in HIV-negative patients: a randomised clinical trial. Lancet 2002; 360:528–534.
12. Aber VR, Nunn AJ. Factors affecting relapse following short-course chemotherapy. Bull Int Union Tuberc 1978; 53:260–264.
13. Pulido F, Pena JM, Rubio R, Moreno S, Gonzalez J, Guijarro C, et al
. Relapse of tuberculosis
after treatment in human immunodeficiency virus-infected patients. Arch Intern Med 1997; 157:227–232.
14. Golub JE, Saraceni V, Cavalcante SC, Pacheco AG, Moulton LH, King BS, et al
. The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis
incidence in HIV-infected patients in Rio de Janeiro, Brazil. AIDS 2007; 21:1441–1448.
15. Moulton LH, Golub JE, Durovni B, Cavalcante SC, Pacheco AG, Saraceni V, et al
. Statistical design of THRio: a phased implementation clinic-randomized study of a tuberculosis
preventive therapy intervention. Clin Trials 2007; 4:190–199.
16. Castelo Filho A, Kritski AL, Barreto AW, Lemos ACM, Netto AR, Gueimarares CA, et al
. II Brazilian consensus on treatment of tuberculosis
– Brazilian guidelines for tuberculosis
2004 [in Portuguese]. J Bras Pneumol 2004; 30(Suppl 2):S62–S64.
17. Selassie AW, Pozsik C, Wilson D, Ferguson PL. Why pulmonary tuberculosis
recurs: a population-based epidemiological study. Ann Epidemiol 2005; 15:519–525.
18. Kopanoff DE, Snider DE Jr, Johnson M. Recurrent tuberculosis
: why do patients develop disease again? A United States Public Health Service cooperative survey. Am J Public Health 1988; 78:30–33.
19. Jones JL, Hanson DL, Dworkin MS, DeCock KM. HIV-associated tuberculosis
in the era of highly active antiretroviral therapy. The Adult/Adolescent Spectrum of HIV Disease Group. Int J Tuberc Lung Dis 2000; 4:1026–1031.
20. Badri M, Wilson D, Wood R. Effect of highly active antiretroviral therapy on incidence of tuberculosis
in South Africa: a cohort study. Lancet 2002; 359:2059–2064.
21. Lawn SD, Badri M, Wood R. Tuberculosis
among HIV-infected patients receiving HAART: long term incidence and risk factors in a South African cohort. AIDS 2005; 19:2109–2116.
22. Girardi E, Antonucci G, Vanacore P, Palmieri F, Matteelli A, Iemoli E, et al
in HIV-infected persons in the context of wide availability of highly active antiretroviral therapy. Eur Respir J 2004; 24:11–17.
23. Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med 1998; 338:853–860.
24. Grinsztejn B, Veloso VG, Pilotto JH, Campos DP, Keruly JC, Moore RD. Comparison of clinical response to initial highly active antiretroviral therapy in the patients in clinical care in the United States and Brazil. J Acquir Immune Defic Syndr 2007; 45:515–520.
25. Lawn SD, Bekkar LG, Middlekoop K, Myer L, Wood R. Impact of HIV infection on the epidemiology of tuberculosis
in a peri-urban community in South Africa: the need for age-specific interventions. Clin Infect Dis 2006; 42:1040–1047.
26. Moh R, Danel C, Messou E, Ouassa T, Gabillard D, Anzian A, et al
. Incidence and determinants of mortality and morbidity following early antiretroviral therapy initiation in HIV-infected adults in West Africa. AIDS 2007; 21:2483–2491.
27. Fitzgerald DW, Desvarieux M, Severe P, Joseph P, Johnson WD Jr, Pape JW. Effect of posttreatment isoniazid on prevention of recurrent tuberculosis
in HIV-1-infected individuals: a radomised trial. Lancet 2000; 356:1470–1474.
28. Churchyard GJ, Fielding K, Charalambous S, Day JH, Corbett EL, Hayes RJ, et al
. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS 2003; 17:2063–2070.
29. Haller L, Sossouhounto R, Coulibaly IM, Dosso M, Kone M, Adom H, et al
. Isoniazid plus sulphadoxine-pyrimethamine can reduce morbidity of HIV-positive patients treated for tuberculosis
in Africa: a controlled clinical trial. Chemotherapy 1999; 45:452–465.
30. Cavalcante SC, Soares EC, Pacheco AG, Chaisson RE, Durovni B, DOTS Expansion Team. Community DOT for tuberculosis
in a Brazilian favela: comparison with a clinic model. Int J Tuberc Lung Dis 2007; 11:544–549.
31. Whalen CC, Johnson JL, Okwera A, Hom DL, Huebner R, Mugyenyi P, et al
. A trial of three regimens to prevent tuberculosis
in Ugandan adults infected with the human immunodeficiency virus. Uganda-Case Western Reserve University Research Collaboration. N Engl J Med 1997; 337:801–808.
32. de Pinho AM, Santoro-Lopes G, Harrison LH, Schechter M. Chemoprophylaxis for tuberculosis
and survival of HIV-infected patients in Brazil. AIDS 2001; 15:2129–2135.
33. Halsey NA, Coberly JS, Desormeaux J, Losikoff P, Atkinson J, Moulton LH, et al
. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis
in HIV-1 infection. Lancet 1998; 351:786–792.
34. Pape JW, Jean SS, Ho JL, Hafner A, Johnson WD Jr. Effect of isoniazid prophylaxis on incidence of active tuberculosis
and progression of HIV infection. Lancet 1993; 342:268–272.
35. Mwinga A, Hosp M, Godfrey-Faussett P, Quigley M, Mwaba P, Mugala BN, et al
. Twice weekly tuberculosis
preventive therapy in HIV infection in Zambia. AIDS 1998; 12:2447–2457.
36. Gordin F, Chaisson RE, Matts JP, Miller C, de Lourdes Garcia M, Hafner R, et al
. Rifampin and pyrazinamide vs isoniazid for prevention of tuberculosis
in HIV-infected persons: an international randomized trial. JAMA 2000; 283:1445–1450.
37. World Health Organization/Global Tuberculosis
Programme and UNAIDS. Policy statement on preventive therapy against tuberculosis in people living with HIV [policy statement]
. Wkly Epidemiol Rec
38. World Health Organization. Interim policy on collaborative TB/HIV activities. WHO/HTM/TB/2004 330 2004.