Tuberculosis (TB) is one of the most common opportunistic infections in people living with HIV (PLHIV).1,2 HIV infection significantly increases the risk of progression from latent TB infection to active TB disease.3 TB can complicate HIV infection regardless of immune status or antiretroviral use4–8 and is the leading cause of death for PLHIV in resource-limited settings.9 The incidence of culture-confirmed TB has been shown to be up to 19 times higher in PLHIV compared with those without HIV infection.10 HIV-associated TB is also a major contributor to the increasing global burden of TB.9,11
The World Health Organization (WHO) recommends screening PLHIV for active TB disease and, in those without TB disease, empirically treating PLHIV for latent TB infection with isoniazid preventive therapy (IPT).12 IPT benefits individuals by reducing morbidity from active TB disease and potentially benefits communities by reducing transmission of TB. Many countries have not yet implemented IPT despite many studies confirming the efficacy of IPT.13–17 One of the main reasons is concern about how to accurately exclude active TB disease in PLHIV9 before administering IPT.
Randomized controlled trials have consistently shown that IPT reduces TB incidence only in PLHIV with a positive tuberculin skin test (TST).18 In PLHIV, IPT may also reduce mortality.19 WHO's 2010 policy on TB prevention recommends that TST is not required before initiating IPT because of challenges with implementing TST in resource-limited settings; rather, TST should only be used where it is already routinely available.20 TST has been shown to be feasible in some resource-limited settings, especially among clients attending HIV voluntary counseling and testing services.21,22 Compared with a strategy of providing IPT to all PLHIV, providing IPT to only those PLHIV with positive TST could conserve resources and reduce drug-related toxicity because only those PLHIV most likely to benefit would receive IPT and require long-term follow-up.
In a large study that we conducted in Thailand, Cambodia, and Vietnam, we found that asking about a combination of symptoms performs well at identifying PLHIV at low risk of TB disease, results that have been confirmed by a meta-analysis and incorporated into WHO policy.20 We also found that, in patients with a positive symptom screen, mycobacterial culture was the most sensitive next test to identify patients with TB disease; an alternative strategy is to perform smear microscopy, chest radiography, and CD4 count testing to identify patients with a positive symptom screen who are at highest risk of TB disease and most likely to benefit from culture.23
Previous studies of how to screen for active TB in PLHIV before administering IPT have primarily analyzed PLHIV regardless of TST status. Because TST is important for identifying patients likely to benefit from IPT and TB prevalence may differ by TST status, we evaluated the impact that baseline TST has on the performance characteristics of the standardized symptom and laboratory TB screening and diagnosis strategies among PLHIV.
Enrollment of Study Participants
PLHIV who came for various services at the Thai Red Cross Anonymous Clinic in Bangkok, Thailand, were enrolled consecutively from September 2006 to January 2008. This study was part of a larger parent study to define an algorithm for TB screening and diagnosis in PLHIV in Southeast Asia.23 Of all PLHIV enrolled in the larger study, only those PLHIV enrolled at the Thai Red Cross Anonymous Clinic had TST performed. IPT was prescribed for those with positive TST after active TB was excluded.
Patients were eligible if they had documented HIV infection, were aged more than 6 years old, had not undergone TB screening with chest radiography or sputum smears in the previous 3 months, had not received TB treatment or IPT in the past year, and had taken no medications with anti-TB activity within the past month. Informed consent was obtained from all study participants or their parents/guardians. The study was approved by the human subjects review committees at the Thailand Ministry of Public Health, Chulalongkorn University, and the United States Centers for Disease Control and Prevention.
Clinical and Laboratory Assessments
Study participants underwent a standardized interview by the study co-ordinator and a physical examination by a medical doctor. Chest radiography, complete blood count, CD4 count testing, and plasma HIV-1 RNA level were also performed. Three sputum specimens were collected over 2 days along with 1 specimen each of urine, blood and stool. Lymph node aspiration from the largest palpable node was carried out if physical examination revealed peripheral lymph nodes >1 cm (or >2 cm for inguinal nodes). All specimens, with the exception of blood, underwent concentrated Ziehl–Neelsen stain and were analyzed under a microscope for acid-fast bacilli. Mycobacterial culture was done on Lowenstein–Jensen media and Mycobacterial Growth Indicator Tube. Blood specimens were inoculated directly into Myco-F-Lytic bottles (BD, Franklin Lakes, NJ) and placed into an automated blood culture instrument (BACTEC 9120/9240 system). Positive cultures were identified as Mycobacterium tuberculosis (MTB) using niacin production and nitrate reduction tests or the AccuProbe Mycobacterium tuberculosis complex assay (TB AccuProbe; Gen-Probe Incorporated, San Diego, CA).
Tuberculin Skin Test
TST was performed in all PLHIV by placing 5 units of purified protein derivative, produced by Queen Saovabha Memorial Institute of the Thai Red Cross Society in Bangkok, intradermally on the ventral surface of the forearm. Studies conducted in Thailand have found that Thai Red Cross tuberculin produces reactions indistinguishable from international standard purified protein derivative-S.24 The diameter of induration was read in millimeter by a study coordinator (S.P.) 2–4 days later. Study participants with an induration ≥5-mm diameter were defined as TST positive.
In the parent study conducted in Cambodia, Thailand, and Vietnam,23 the total sample size was calculated based on the number of patients needed to demonstrate the desired level of screening and diagnostic test characteristics across a range of populations, each with different true prevalence of TB. Assuming a TB prevalence of 5% and estimating the sensitivity of 95%, the specificity of 80%, and the precision of 5%, we would need to enroll 1460 patients across the 3 countries. The study team oversampled in each country to allow the possibility to have country-specific estimates.
A diagnosis of active TB disease was given if PLHIV had at least 1 MTB culture-positive specimen. To calculate adjusted odds ratios (aOR) for factors associated with a positive TST, variables significant in bivariate analysis at P <0.20 and factors hypothesized a priori to be associated with a positive TST were entered in a multiple logistic regression model, and a final model was chosen through stepwise automated variable selection. We assessed all factors for colinearity; for factors that were colinear, we retained the factor with largest odds ratio and which we judged subjectively to be most clinically meaningful.
The number of patients with positive or negative TST status who were at high risk of active TB disease and would therefore require MTB culture to diagnose active TB disease was calculated under separate conditions including: (1) when positive symptom screening was used alone (a positive symptom screen was defined as patient report of at least 1 of any cough, any fever or night sweats lasting 3 or more weeks in the preceding 4 weeks); (2) when positive laboratory screening was added (a positive laboratory screen was defined as positive sputum smear or abnormal chest radiography or CD4 count <350 cells/mm3); (3) when positive laboratory screening was used alone. The symptom questions and laboratory tests were chosen based on their performance as risk-stratification tools in the larger parent study.23 Our algorithm assumes that patients with smear-positive TB undergo mycobacterial culture because nontuberculous mycobacteria is common in Thai PLHIV. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of using TST alone, symptom screening alone, and TST plus symptom screening to diagnose active TB disease were calculated.
A total of 657 PLHIV were screened, 632 were eligible for enrollment, 630 were enrolled, and 604 were analyzed for this study (Fig. 1). Of the 630 PLHIV enrolled, 26 (4%) could not be analyzed because they were missing TST results (either not tested or not read). PLHIV with missing TST results had higher CD4 count (P value = 0.02) and were more likely to be on antiretroviral therapy (P = 0.02) than those with TST results (data not shown). There was no difference between the groups regarding sex, plasma HIV-1 RNA level, HIV risk factors, symptoms, positive sputum smear, or positive culture rates.
Fifty-eight percent of participants were male (Table 1). The median age was 32 years (interquartile range, IQR: 28–38, minimum age was 17 years). Twenty percent reported receiving antiretroviral therapy. Median CD4 count was 306 cells per cubic millimeter (IQR: 191–466). Plasma HIV-1 RNA was undetectable in 13.6%. Injection drug use (IDU) was reported by 4.3%.
TABLE 1-a Characteri...Image Tools
Any cough, any fever, and night sweats lasting 3 or more weeks in the preceding 4 weeks were present in 40.6%, 29.6%, and 2.3% of participants, respectively; 314 (52.0%) had at least 1 of these symptoms. An abnormal chest radiograph was found in 14.4%. One or more sputum smears were positive for AFB in 15 (2.5%), and 33 (5.5%) had at least 1 culture positive for MTB, including 32 (97%) from sputum and 1 from blood alone. Of the 32 patients with positive sputum cultures, 13 had cultures positive from additional sites as follows: 9 stool only, 1 lymph node only, 1 urine only, 1 blood and urine, and 1 urine and stool.
TABLE 1-b Characteri...Image Tools
Tuberculin Skin Test
A positive TST result was found in 151 (25.0%) patients. Among those with a positive TST, the median diameter of induration was 20.0 mm (IQR: 14.5–27.0). In multivariate analysis (Table 2), factors associated with a positive TST included having a CD4 count ≥350 cells per cubic millimeter as compared with <350 cells per cubic millimeter [aOR: 2.4, 95% confidence interval (CI): 1.4 to 4.0, if CD4 count 350–500 cells/mm3; aOR: 3.8 and CI: 2.3 to 6.3, if CD4 count >500 cells/mm3], self-reported IDU history (aOR: 5.2; CI: 2.2 to 12.7), odynophagia in the past 4 weeks (aOR: 0.4; CI: 0.2 to 0.8), and having a positive culture for MTB (aOR: 11.7; CI: 4.2 to 32.4).
Algorithm to Screen for Active TB Disease in PLHIV Stratified by TST
Of 151 TST-positive PLHIV, 22 (14.6%) were diagnosed with TB disease. Using an approach of only symptom screening, 80 (52.9%) of 151 would screen positive and need MTB culture because they had at least 1 of the 3 symptoms included in the symptom screening questions; 18 (22.5%) of those screening positive would be diagnosed with TB disease (Fig. 2). Among the 71 TST-positive but symptom screen negative PLHIV, 4 TB cases (18.2% of all TB cases among TST positives) would be missed.
Using an approach of symptom screening, followed by laboratory screening (ie, sputum smear, chest radiography, and CD4 count) to identify patients at high risk of TB disease, 51 (63.8%) of the 80 TST-positive, symptom screen positive PLHIV would need MTB culture to confirm or rule out active TB disease. Among the 80 TST-positive and symptom screen positive PLHIV, 2 TB cases would be missed using this screening algorithm.
If all 151 TST-positive PLHIV underwent only laboratory screening, instead of symptom screening, 79 (52.3%) would screen positive and require MTB culture. Three TB cases among all 151 TST-positive PLHIV would be missed by this approach.
Among 453 TST-negative PLHIV, 11 (2.4%) were diagnosed with TB disease. Using an approach of only symptom screening, 234 (51.6%) of 453 would screen positive and need MTB culture; 10 (4.3%) of those with a positive symptom screen would be diagnosed with TB disease. Among the 219 TST-negative and symptom screen negative PLHIV, 1 (9.1% of all TB cases among TST negatives) TB case would be missed.
Using an approach of symptom screening, followed by laboratory screening, 161 (68.8%) of 234 TST-negative, symptom screen positive PLHIV would need MTB culture. Among the 234 TST-negative but symptom screen positive PLHIV, no TB cases would be missed by this screening algorithm.
If all 453 TST-negative PLHIV underwent only laboratory screening first, instead of symptom screening, 306 (67.5%) would screen positive and require MTB culture. None of the TB cases would be missed by this approach.
Performances of TST and Symptoms to Screen for Active TB Disease
Using TST alone to identify active TB disease had 66.7% sensitivity, 77.4% specificity, 14.6% PPV, and 97.6% NPV (Table 3). In contrast, symptom screening alone regardless of TST status had 84.8% sensitivity, 49.9% specificity, 8.9% PPV, and 98.3% NPV. A combination of these approaches (TST, followed by symptom screening among those that were TST positive) had 54.5% sensitivity, 89.1% specificity, 22.5% PPV, and 97.1% NPV. A combination of symptom screening followed by laboratory screening regardless of TST results had 78.8% sensitivity, 67.4% specificity, 12.3% PPV, and 97.8% NPV; overall, 35.1% of 604 PLHIV would undergo MTB culture to correctly diagnose 78.8% of 33 active TB cases.
One other approach would be to perform TST first, followed by laboratory screening in TST-positive PLHIV and symptom and laboratory screening in TST-negative PLHIV. This approach had 87.9% sensitivity, 63.0% specificity, 12.1% PPV, and 98.9% NPV; overall, 240 (39.7%) of 604 PLHIV would undergo MTB culture to correctly diagnose 29 (87.9%) of 33 active TB cases. Compared with symptom screening alone, this approach could increase the number of active TB cases correctly diagnosed (87.9% vs. 84.8%) and decrease the need for MTB culture (39.7% vs. 52.0%) among PLHIV.
Another approach would be to perform MTB culture on all TST-positive PLHIV and on TST-negative PLHIV with positive symptom screening and either positive sputum smear or abnormal chest radiography. This approach had 87.9% sensitivity, 70.8% specificity, 14.8% PPV, and 99.0% NPV. Compared with the previous approach, this approach could correctly diagnose a similar proportion of active TB cases (87.9%) although reducing the need for MTB culture to 32.4%. The number of PLHIV who needed sputum smear and chest radiograph reduced from 63.7% to 38.7% and from 61.3% to 38.7%, respectively.
Although infrequently used in the high-burden TB countries, TST can aid with care for PLHIV by helping to screen for active TB disease and by identifying patients most likely to benefit from IPT. Our study found that one-fourth of PLHIV attending an urban clinic in Thailand had a positive TST and that patients with a negative TST and negative symptom screening were at very low risk of having active TB disease.
The roles of TST in the diagnosis of active TB in resource-limited settings remains uncertain. In settings where the prevalence of latent TB is high, the use of TST may be limited by their poor specificity.25 The sensitivity and specificity reported from previous studies, which evaluated the use of TST to diagnose active TB, ranged from 69%–94% and 76%–88%, respectively.25–27 When used alone in our study, TST provided 66.7% sensitivity and 77.4% specificity for active TB diagnosis.
In our study population, we found that a positive TST identified a subset of PLHIV with a high prevalence of TB disease (14.6%) and that symptom screening did not perform particularly well in this group, missing 18.2% of TB cases among TST-positive PLHIV. Symptom screening also did not perform well in excluding TB disease in TST-positive PLHIV; the NPV was 94.4% (data not shown), given the low prevalence of TB disease (5.6%) in TST-positive PLHIV with negative symptom screening. Our findings present a challenge for public health and clinical decision-making: those most likely to benefit from IPT had the highest risk of TB disease and the greatest difficulty being ruled out for TB disease. In our study, screening could be maximally improved among TST-positive PLHIV by performing mycobacterial culture of sputum for all TST-positive PLHIV before initiating IPT. Another option that has recently been validated and endorsed by WHO involves testing sputum using a fully automated nucleic acid amplification test (eg, GeneXpert); although less sensitive than liquid culture, Xpert has a sensitivity that exceeds smear microscopy, provides results in less than 2 hours, requires minimal training, and has minimal interoperator variability.28 Our study suggests that places in which TST is routinely available should consider evaluating a strategy of screening PLHIV with TST and, among those that are TST positive, performing Xpert or culture before providing IPT.
Because PLHIV who are TST negative do not benefit from IPT, 75% of PLHIV in our clinic would not be expected to benefit from IPT; rather, the highest priority for TB prevention in this group would be to exclude active TB disease. We found symptom screening—which has been validated both in Southeast Asia and globally—performs extremely well in this group, with a NPV of 99.5%. Therefore, in a theoretical population of 1000 PLHIV in our center, a strategy of performing TST first then culture (or other sensitive TB diagnostic evaluation) for TST-positive PLHIV or for TST-negative, symptom screen positive PLHIV would require diagnostic evaluations for 63.7% of patients (250 among TST positives; 387 among TST negative but symptom screen positives) and no further evaluation for 36.3%. After diagnostic evaluations, 213 (85.4%) of 250 TST-positive patients (with no active TB disease) would receive IPT.
Our proposed approach to perform MTB culture in all TST-positive PLHIV and to perform symptom screening followed by sputum smear and chest radiography in TST-negative PLHIV increased the number of active TB cases correctly diagnosed by 3.1% and decreased the need for MTB culture by 19.6%, compared with the use of symptom screening alone. A formal cost-effectiveness study of our algorithm should be performed to evaluate whether our strategy would benefit patients and public health at an acceptable cost compared with the approach most likely to be used in resource-limited settings (ie, symptom screening of all PLHIV, provision of IPT to all patients that are symptom screen negative, and diagnostic evaluations for all patients that are symptom screen positive).
Our study confirms other previous reports about use and interpretation of TST. First, we demonstrated a very high return rate (99.1%) of Thai PLHIV for TST reading in this study. This finding is consistent with previous studies showing the feasibility of performing TST among Thai PLHIV.29 Second, TST reactivity was greatest in those with relatively preserved immune function.30–33 Third, IDU was highly correlated with positive TST, similar to what has been seen in Russia.31 We hypothesize that, in Bangkok, this is likely due to exposure to jails or other environments in which TB transmission is high.34
Our study is subject to several important limitations. First, most of our participants were adults living in Bangkok. The epidemiology of TB infection and disease in PLHIV in our population may be substantially different than other parts of the world. Second, we used our study population to derive an algorithm for TB screening. Prospective evaluations are needed to validate whether the approach we have described performs under routine conditions and in different settings. Finally, this study was conducted as clinical research with resources not routinely available in most settings. In settings where TST and highly sensitive TB diagnostic tests are not available, other approaches may be needed. Our strategy involves more widespread use of TB diagnostic testing and more limited use of IPT in PLHIV, which is in accordance with the current WHO recommendation for settings where TB diagnostic testing is available.
In summary, we found that TST may be a useful initial screening test among PLHIV both to identify the minority of patients highly likely to benefit from IPT and to identify those at highest risk of active TB disease.
The authors thank Dr Chawalit Wongsuthipol for his clinical work, and Sumanee Nilgate, Punjapon Prasurthsin, and Yuttana Kerdsuk from the Microbiology Department, Faculty of Medicine, Chulalongkorn University, for managing clinical specimens and performing laboratory assessments in our study.
1. Nissapatorn V. Lessons learned about opportunistic infections in Southeast Asia. Southeast Asian J Trop Med Public Health. 2008;39:625–641.
2. Chakraborty N, Mukherjee A, Santra S, et al.. Current trends of opportunistic infections among HIV-seropositive patients from Eastern India. Jpn J Infect Dis. 2008;61:49–53.
3. Whalen C, Horsburgh CR, Hom D, et al.. Accelerated course of human immunodeficiency virus infection after tuberculosis. Am J Respir Crit Care Med. 1995;151:129–135.
4. Manosuthi W, Chaovavanich A, Tansuphaswadikul S, et al.. Incidence and risk factors of major opportunistic infections after initiation of antiretroviral therapy among advanced HIV-infected patients in a resource-limited setting. J Infect. 2007;55:464–469.
5. Corbett EL, Watt CJ, Walker N, et al.. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med. 2003;163:1009–1021.
6. Brinkhof MW, Egger M, Boulle A, et al.. Tuberculosis after initiation of antiretroviral therapy in low-income and high-income countries. Clin Infect Dis. 2007;45:1518–1521.
7. Manabe YC, Breen R, Perti T, et al.. Unmasked tuberculosis and tuberculosis immune reconstitution inflammatory disease: a disease spectrum after initiation of antiretroviral therapy. J Infect Dis. 2009;199:437–444.
8. Duncombe C, Kerr SJ, Ruxrungtham K, et al.. HIV disease progression in a patient cohort treated via a clinical research network in a resource limited setting. AIDS. 2005;19:169–178.
9. WHO. Global Tuberculosis Control: Surveillance, Planning, Financing: WHO Report 2008. Geneva, Switzerland: World Health Organization; 2008. Report No.: WHO/HTM/TB/2008.393.
10. Corbett EL, Bandason T, Cheung YB, et al.. Epidemiology of tuberculosis in a high HIV prevalence population provided with enhanced diagnosis of symptomatic disease. PLoS Med. 2007;4:e22.
11. Williams B, Maher D. Tuberculosis fueled by HIV: putting out the flames. Am J Respir Crit Care Med. 2007;175:6–8.
12. WHO. Interim Policy on Collaborative TB/HIV Activities. Geneva, Switzerland: World Health Organization; 2004. Report No.: WHO/HTM/TB/2004.330.
13. Pape JW, Jean SS, Ho JL, et al.. Effect of isoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection. Lancet. 1993;342:268–272.
14. Whalen CC, Johnson JL, Okwera A, 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.
15. Halsey NA, Coberly JS, Desormeaux J, et al.. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet. 1998;351:786–792.
16. Hawken MP, Meme HK, Elliott LC, et al.. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS. 1997;11:875–882.
17. Woldehanna S, Volmink J. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database Syst Rev. 2004;(1):CD000171.
18. Akolo C, Adetifa I, Shepperd S, et al.. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database Syst Rev. 2010;(1):CD000171.
19. Samandari T, Agizew TB, Nyirenda S, et al.. 6-month versus 36-month isoniazid preventive treatment for tuberculosis in adults with HIV infection in Botswana: a randomised, double-blind, placebo-controlled trial. Lancet. 2011;377:1588–1598.
21. Aisu T, Raviglione MC, van Praag E, et al.. Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS. 1995;9:267–273.
22. Mugisha B, Bock N, Mermin J, et al.. Tuberculosis case finding and preventive therapy in an HIV voluntary counseling and testing center in Uganda. Int J Tuberc Lung Dis. 2006;10:761–767.
23. Cain KP, McCarthy KD, Heilig CM, et al.. An algorithm for tuberculosis screening and diagnosis in people with HIV. N Engl J Med. 2010;362:707–716.
24. Bharnthong T, Kongwatana C, Premchaiporn P, et al.. A comparison of PPD-TRC and PPD-S by Mantoux test. Thai J Tuberc Chest Dis. 2002;23:23–31.
25. Syed Ahamed Kabeer B, Raman B, Thomas A, et al.. Role of QuantiFERON-TB gold, interferon gamma inducible protein-10 and tuberculin skin test in active tuberculosis diagnosis. PLoS One. 2010;5:e9051.
26. Zhang S, Shao L, Mo L, et al.. Evaluation of gamma interferon release assays using Mycobacterium tuberculosis antigens for diagnosis of latent and active tuberculosis in Mycobacterium bovis BCG-vaccinated populations. Clin Vaccine Immunol. 2010;17:1985–1990.
27. Lee JE, Kim HJ, Lee SW. The clinical utility of tuberculin skin test and interferon-gamma release assay in the diagnosis of active tuberculosis among young adults: a prospective observational study. BMC Infect Dis. 2011;11:96.
28. Helb D, Jones M, Story E, et al.. Rapid detection of Mycobacterium tuberculosis and rifampin resistance by use of on-demand, near-patient technology. J Clin Microbiol. 2010;48:229–237.
29. Khongphatthanayothin M, Avihingsanon A, Teeratakulpisarn N, et al.. INH prophylaxis for TST-positive HIV clients at the Thai Red Cross AIDS Research Centre. Presented at: 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention; July 19–22, 2009; Cape Town, South Africa. 2009. TUPEB 128.
30. Balcells ME, Perez CM, Chanqueo L, et al.. A comparative study of two different methods for the detection of latent tuberculosis in HIV-positive individuals in Chile. Int J Infect Dis. 2008;12:645–652.
31. Vitek E, Gusseinova N, Laricheva N, et al.. Factors associated with positive tuberculin skin test results among HIV-infected persons in Orel Oblast, Russia. Int J Tuberc Lung Dis. 2009;13:829–835.
32. Swaminathan S, Subbaraman R, Venkatesan P, et al.. Tuberculin skin test results in HIV-infected patients in India: implications for latent tuberculosis treatment. Int J Tuberc Lung Dis. 2008;12:168–173.
33. Hiransuthikul N, Hanvanich M, Dore GJ, et al.. Factors associated with tuberculin skin test reactivity among HIV-infected people in Bangkok. Southeast Asian J Trop Med Public Health. 2003;34:804–809.
34. Sretrirutchai S, Silapapojakul K, Palittapongarnpim P, et al.. Tuberculosis in Thai prisons: magnitude, transmission and drug susceptibility. Int J Tuberc Lung Dis. 2002;6:208–214.
© 2012 Lippincott Williams & Wilkins, Inc.