TITLE

Prevalence of LTBI among high risk populations: Cross sectional community-based study of high risk, migrant colonia residents along the south Texas-Mexico border.

Abstract

Identification and treatment of latent tuberculosis infection (LTBI) among high risk populations is an essential component of the Centers for Disease Control and Prevention (CDC) strategy for elimination of tuberculosis (TB) in the United States. Along the Lower Rio Grande Valley on the Texas-Mexico border there is a large population of migrant Hispanics. In 1997, the TB incidence rate for the 14 counties along the Texas-Mexico border was 15.3 cases per 100,000 population. Over 93% of the patients were Hispanics; half were born outside the US, and of these 93% were born in Mexico. Three hundred and forty thousand live in “colonias”, which are informal settlements that may lack potable water and wastewater systems. Colonia residents are at high risk of LTBI and TB, with an estimated TB incidence rate of 26 cases per 100,000 persons. Thirty-six percent of the population living in the colonias are children, approximately one third live below the poverty line, and 15-25% have diabetes. Most residents of newer colonias have poor access to health care and live under crowded conditions. One third of the colonia residents are immigrants from Mexico and other countries where TB is endemic. Many visit Mexico frequently, contributing to the 1.6 million daily border crossings across the busiest border in the world. Importantly, many are migrant agricultural workers forming part of the seasonal flow of immigrants to the northern United States. In fact, Mexico is the country of origin for 23% of all foreign-born persons with TB in the US. In this study the prevalence of latent TB infection (LTBI) will be determined in colonia residents from the Texas-Mexico border using the tuberculin skin test (TST) and the newly licensed QuantiFERON assay. A one-stage random cluster design will be used to obtain a representative sample. Individuals with LTBI will be referred to Hidalgo County Health Department where they will be individually assessed and treated appropriately. Analysis will include population-based estimates of LTBI prevalence, including geographic information system coding (GIS) analysis, treatment adherence, and analysis of possible causes of discordance between the TST and QuantiFERON assay. These results will be further analyzed for associations with sociodemographic and health related variables. This unique study will provide the first population-based prevalence data from a large, rapidly expanding, high risk migrant community within the United States.

RESEARCH IDEA:

THIS PROJECT WAS DEVELOPED IN RESPONSE TO A CALL FOR APPLICATIONS THAT PROVIDED THE FOLLOWING INFORMATION:

LTBI Prevalence

As the rate of active TB in the US has decreased, targeted testing for LTBI to identify persons at high risk for TB has become an essential component of CDC’s TB elimination strategy. Identifying and treating LTBI among high risk populations is necessary for eliminating the reservoir of persons with LTBI who will progress to active disease.

In the CDC TB elimination plan, it is assumed that the problem of LTBI in the US is considerable. The elimination plan states that there are an estimated 10 million to 15 million Americans with latent TB infection. However, there is currently no good estimate of LTBI prevalence among specific high-risk populations. There is also little information on the percentage of patients who receive and complete treatment for LTBI.

An estimate of LTBI burden among high risk populations will serve as a baseline for future studies assessing the impact of targeted screening programs on LTBI prevalence. It will also provide information on progress towards TB elimination. Data on LTBI patients who receive and complete treatment will provide information on the effectiveness of LTBI treatment programs. Both components of the study will inform future resource allocation for targeted screening programs.

LTBI Diagnosis

The TST is used as the initial screening test for both LTBI and TB. Despite its widespread use, it has a number of important limitations (e.g. false-positive results, error in placement of test, subjectivity in test reading, return visit required, etc.). A new in vitro diagnostic test for LTBI, the IFN-g assay, has recently been approved for use in the United States. Unlike the TST, the IFN-g assay requires a single patient visit, assesses responses to multiple antigens simultaneously, does not boost amnestic immune responses, eliminates the subjectivity of the TST, and can be completed in less than 24 hours. The fact that the IFN-g assay includes avian PPD to discriminate responses due to reactivity from non-TB mycobacteria and that BCG vaccination has less effect on the IFN-g assay than does TST suggests that the IFN-g assay may have a lower rate of false-positives. However, because there is no “gold standard” for LTBI testing, this remains unclear.

Because of its recent approval in the US, the IFN-g assay has not been widely evaluated. A comparison of the two methods will provide important information on the potential use of the IFN-g assay as a valuable alternative to the TST.

LITERATURE REVIEW WITH REFERENCED LITERATURE:

Background and Need for Support

Although the overall rate of active tuberculosis (TB) has decreased in the US, a high incidence persists along the US-Mexico border, with new cases identified among recent immigrants from Mexico. In the US, Mexican immigrants account for 66% of the Hispanics, and for 23% of all foreign birth TB cases (1-3). The border is porous, with frequent migration, relocation of families, travelers and migrant workers contributing to diffusion of health problems. Reactivation and spread of TB is facilitated in a population made vulnerable by health disparity, poverty, stress of migration, crowding and concomitant disease. The border population has many recent immigrants from south Mexico where reported rates of active TB are extremely high (Table) (4). A substantial population of newly arrived immigrants settle in semi-rural communities called ‘colonias‘ concentrated along the Lower Rio Grande Valley (LRGV) on the Texas-Mexico border, where there are an estimated 340,000 Texas residents in 1,436 colonias (5-7). Residents are 97% Hispanic, mostly with low socioeconomic status, and are living under crowded conditions in poor housing. Converging factors contribute to the high incidence of TB: residents tend to be young (37% are children), one-third are immigrants from countries with a high incidence of TB; and the primary language is Spanish, often compromising seeking and adhering to treatment. Most lack medical insurance or have poor access to healthcare, and many are undocumented (5-10). Many seek care from private physicians in Mexico, thereby escaping the national surveillance and treatment systems in both countries, and contributing to both underreporting and the increase in drug resistance. The high incidence of diabetes in this Hispanic population further elevates the risk for TB (11-13). Reports of TB in different areas of the US have been traced back to the Mexico-US border(14), which is the busiest in the world. While the burden of TB has decreased from incidence rates/100,000 of 15.3 in 1997 to 12.2 in 2000 (2;15), the rate of TB in the colonias is at least 4.4 times higher compared with the US rate (Table) (10;16;17). Over 90% of all these cases are Hispanic and 93% of those born outside the US are from Mexico (2). More serious are the 70% higher rate of isoniazid resistance and the 100% higher rate of rifampin resistance among the TB patients living along the Texas-Mexico border when compared with non-border patients (2). In this population, targeted testing for latent tuberculosis infection (LTBI) and prompt treatment are essential to eliminate the reservoir of persons who will progress to active disease. Estimating the prevalence of LTBI and ensuring adherence to treatment in border residents is key to insuring the continued control of TB in Texas and further north.

There is no gold standard for identification of LTBI, and many patients are only recognized when they become sufficiently sick to seek health care. By this time spread may be substantial. Until now, LTBI has been defined by a reactive tuberculin skin test (TST), which is limited by poor specificity (18). False positives are due to a previous exposure to environmental mycobacterium and/or BCG vaccination. The latter effect wanes at approximately 8% per year (19). False negatives (sensitivity) are a concern in immunosuppressed individuals, a population that is highly vulnerable to TB. The FDA has recently approved a new in vitro diagnostic test, the QuantiFERON (CSL, Melbourne, Australia). Compared to TST, this assay appears to have a superior specificity, eliminates subjectivity, does not boost anamnestic immune responses, is completed in 24 hours, and requires a single patient visit. The literature indicates that a recent BCG vaccination (within 6 months) compromises specificity in about 20% of the assays, but a longer term effect has not been evaluated (20). Thus, it appears that QuantiFERON may present false positives for a shorter period of time after vaccination, when compared to TST. The extent to which environmental mycobacteria affect the QuantiFERON assay is uncertain, though it is designed to distinguish M. tuberculosis from M. avium complex antigens, and therefore should yield higher specificity than TST. The QuantiFERON has not been specifically evaluated in immunosuppressed subjects, but it is likely to have a similar behavior since it is also dependent upon an intact cell-mediated immune response. The two assays have not been extensively compared, but most of the initial studies suggest poor concordance, with kappa values ranging between 0.16 and 0.73 depending on the study population (21-23). More importantly, disagreement appears to be more common in high-risk individuals where chemoprophylaxis is needed the most (21). Thus, further analysis of discordant cases is essential.

We are enthusiastic about being the first interdisciplinary group of scientists and public health workers from the LRGV to join to develop a team effort to work on a serious local health risk. The LRGV community is anxious to participate in TB prevention, since most are aware of the threat it poses to their health. The design of the study is based on the use of local, trained community workers (promotoras). Promotoras are a well established resource in this area, receiving formal training locally, and used effectively in many research and intervention programs. The study is unique in other respects. We have in these colonias a population in which TB spreads within a minority, severely disadvantaged, semi-rural community within the borders of the United States in a way that we have not seen for a century or more. Spread within such a community is the natural history of TB. Without understanding the dynamics of the disease and how to control spread at the community level, we will not be able to stem the continuing tide of TB from Mexico and Latin America.

SPECIFIC AIMS

Program Specific Methodology and Approach

Specific aims:

Aim 1 Determine the prevalence of latent TB infection (LTBI) in a South Texas immigrant population.

Aim 2 Determine the correlation between the tuberculin skin test (TST) and the QuantiFERON in the participants and investigate possible causes of discordance between the assays.

Aim 3 Determine the rate of completion of treatment in those participants referred for assessment and treatment.

Aim 4 Determine the sociodemographic factors associated with completion of treatment.

METHODS: STUDY POPULATION, RESEARCH DESIGN AND DESCRIPTION OF TECHNIQUES TO BE USED

Study design: A cross-sectional study of LTBI prevalence will be performed among residents of colonias located in the neighborhoods of the cities of Palmhurst and Alton in Hidalgo county (Figs. 1-2). The specific geographic area is defined by the following roads: Palm Dr (West), Mile 2 Rd (South), Taylor Rd (East) and Mile 6 Rd (North) (Fig. 3). This area was selected because it has young and actively growing colonias, and contains a high proportion of i) recent immigrants, ii) frequent travelers between Mexico and the US, and/or iii) close contacts of recent arrivals from Mexico or other Latin American countries where TB is endemic. The definition of a newly arrived migrant as less than five years resident has little relevance in this community, where migrants with extended families on both sides of the border move continuously in both directions, often changing residence repeatedly.

Sampling strategy: Based on 1996 data from the Texas Water Development Board, there are 115 colonias located in the area defined above. We proposed to divide these into 54 colonia clusters that are homogeneous in population size, each having 271 + 34 (average + standard deviation) persons. A simple random sampling technique will be used to select the clusters to be studied. A simple one-stage cluster sampling design of this population with an expected coverage of 100% of the selected colonia clusters will be performed. Assuming a prevalence of LTBI between 15-25% (based on unpublished local health department data), using a type I error level of 5%, a maximum relative difference from the true prevalence of 15%, and assuming that 40% of the population are adults (18 years or older), we will require 3 colonia clusters (24). However, we anticipate a non-participation rate of 30% that will require an increase in the number of colonia clusters to 7. Based on these estimates we will need to screen approximately 1964 individuals that represent 786 adults (24).

Participant selection: All household members 18 years of age or over will be invited to participate. Exclusion of children will ensure that most participants are foreign born, that the population we study is relatively homogenous in terms of TB risk, and will exclude a high rate of false positive TST due to recent BCG vaccination in this pediatric population.

Informed consent and IRB. Upon agreement, all adult members of the household will be approached and invited to participate by responding to a short questionnaire, giving a blood sample, and receiving a TST. Informed consent in writing will be requested for the questionnaire, for the TST and for the blood sample. All positive tests will be reported to the patient, and they will be referred to the Hidalgo County Health Department TB program for follow up, assessment and treatment as appropriate. All documents will be in English and Spanish. This study will commence upon receipt of the Human Subject Review Committees from CDC, the Texas Department of Health, and the University of Texas Health Science Center Houston Committee for the Protection of Human Subjects (CPHS).

Field studies and access to the community: The key strategy in accessing the community is employment of promotoras as described above. Promotoras are mainly women who live in the neighborhood of the colonias, are already acquainted with this population, and are formally trained to participate in community intervention programs. The promotoras will be directly supervised by a Licensed Vocational Nurse (LVN) who has had previous experience with TB control in Hidalgo County. This field team will be based at the Hidalgo County Health Department (HCHD) which is ten minutes from the chosen field site. This is where the county TB program is housed, and the field team will be supervised by the director of that program, Ms. Rosa Morales. The promotoras and the LVN will visit the selected houses. If there is difficulty locating all the adult members of the household, repeat visits will be made.

Specimen collection and laboratory assays: Heparinized blood samples for the QuantiFERON assay will be drawn prior to application of the tuberculin skin antigen, and the tubes will be transported to the TB laboratory in Harlingen and incubated with the QuantiFERON antigens within 12 hours of collection (see details below). This blood sample will also be used for immediate measurement of random glucose and cholesterol using dipstick techniques. The latter results will be given in writing, and for free as an inducement to participate in the study, and if abnormal, the patient will be directed towards full medical assessment to a free clinic in Alton held by volunteer physicians and other health workers, including Dr. Brian Smith. The glucose levels may also provide information complementing the self-reported diabetes that will be requested in the study questionnaire. The TST will be administered by the Mantoux method by either the LVN or the promotoras that have received direct or indirect training at the course to be offered by CDC in New Jersey. The assay will be applied, read and interpreted according to the instructions received (see details below). Regular quality assurance checks will be established. The results for both immunoassays will be recorded immediately so any discrepancies are noted within one week after testing. The subjects with negative TST and positive QuantiFERON will be retested for TST 7-10 days later to detect a booster effect of the tuberculin (25). That is, the first tuberculin injection may boost the immune response induced by a past exposure to M. tuberculosis or BCG vaccination, resulting in a positive TST result.

Diagnostic tests: LTBI will be evaluated using the tuberculin skin test (TST) and the QuantiFERON assay, exactly as described in Task Order #5 instructions. Members of our research team will travel for training courses in each of these techniques, to assure maximum reproducibility between different study sites. Essentially, TST will be administered by the Mantoux method using 0.1 mL of Tubersol (Connaught Labs, Toronto, Ontario) and transverse induration of the site will be measured 48-72 h later. An induration cutoff of 10 mm will be considered positive, unless a risk-stratified interpretation following the American Thoracic Society (ATS)/CDC guidelines is necessary (26).

The QuantiFERON will be done exactly as indicated by the manufacturer. Briefly, within 12 h of collection, the blood will be incubated for 12-24 h at 37°C with the standard antigens provided in the kit. These include a saline negative control, purified protein derivative from M. tuberculosis (human PPD) as the experimental antigen, PPD from M. avium (avian PPD) as specificity control, and phytohemagglutinin as a mitogen to assess the immune status of the subject. After incubation, the plasma samples will be collected and stored frozen for weekly batch analysis of interferon γ (IFN-γ) induction. Calculation of a positive result for M. tuberculosis and M. avium will be done using the defined criteria in the kit. Cases with reactivity to M. avium will be categorized as negative for M. tuberculosis. Cases without mitogen induced proliferation (mitogen – saline control is less than 0.5 IU) will be considered “indeterminate”.

Data entry and analysis: The data from the questionnaires, TST testing and QuantiFERON results will be entered using Microsoft Access 2002 database (Microsoft Corporation). Data will be entered daily in the field and weekly in the laboratory, and real time analysis will return information to the field to assist in monitoring quality and performance on a weekly basis. Discordant TST/QuantIFERON (positive TST, negative QuantIFERON) results will result in repeat TST of those individuals within 7 to 10 days. Statistical analysis will be performed using SAS software (SAS Institute Inc. Cary. 2001). Personal identifiers will only be available in the master document that will be kept locked at the UTH-SPH, and will not be accessible in the database. The prevalence of LTBI will be estimated using an unbiased estimator known as the expansion estimator (24). Socio-demographic characteristics will be analyzed using summary and variability measures for parametric or nonparametric statistics as appropriate.

The addresses of LTBI-positive participants identified in this study, plus the addresses of the active TB cases that have been identified in the last two years by the HCHD will be geocoded on digitized shape network maps GIS-geographic information system. Participants with addresses which are not geocodable will be located using a global positioning system device (GPS). Digitized maps that have been developed from aerial photographs are being kindly provided to us by the State of Texas Attorney’s Office. They precisely indicate the location of each colonia, and census data is in the process of being added. In this way, maps of TB will be prepared and spatial analysis performed using ArcGIS software (ESRI, Redlands, CA), to reveal clustering patterns within the population. This data will visually indicate the location of high risk areas for TB within our study area, or other regions of Hidalgo county.

Treatment and follow-up: All participants with a positive TST and/or QuantiFERON will be referred to the Hidalgo County Health Department (HCHD) where they will be further evaluated on an individual basis to determine if chemoprophylaxis should be initiated. We anticipate that over the two year period there will be an average of 10 newly identified LTBI patients per month (total of 250 out of approximately 1000 patients, assuming there is 25% prevalence of positive TST). Close cooperation will be maintained with the HCHD and data concerning treatment and completion will be available from county records. The HCHD TB program physician will assess the patients and make decisions on treatment. Chest x-ray and sputum culture are available. In Hidalgo County, patients receive 9 months of INH (27-29). It is self-administered in all cases except contacts less than 5 yrs old, which are in DOT. In the patients with highest risk of TB (persons with HIV or diabetics undergoing dialysis) the treatment is extended to 12 months. All patients are monitored for toxicity on a monthly basis. The administrator of the HCHD, Mr. Mike Keenen, supports this study and has expressed willingness to receive and evaluate the referred patients (see his letter of support). Dr. Richard Wing, the TDH regional tuberculosis clinician, also agrees that the estimated increase in work load of 15-20 new persons with LTBI will not be a significant addition to the current workload.

Sociodemographic data and analyses: The draft questionnaire will provide the following information: i) socio-demographic characteristics, including age, race, place of birth, residence outside the US, frequency of travel across the border, residence of work, health insurance, history of incarceration, drug use, residence in homeless shelter, amongst others, ii) BCG vaccination and date, (this is particularly relevant in the colonias since most of the adult residents will be Mexican born), iii) self-reported HIV test results, severe immunocompromised conditions due to cancer, or prolonged treatment with immunosuppressive drugs, because these individuals are at high risk for TB, but may present false-negative TST and QuantiFERON results, iv) self-reported diabetes and a random glucose test to establish the proportion of LTBI subjects with this high-risk condition (see specimen collection above), and v) previous LTBI and TB history, including the presence of current symptoms that may indicate undiagnosed active TB (cough of more than 3 weeks duration, night sweats, etc.). Any suspect TB cases will be referred to HCHD as described above. All questionnaires will be available in Spanish and English, depending on the primary language spoken by the interviewed study subjects. The questionnaire will be prepared in English and translated into Spanish and back translated for accuracy, with consensus from CDC. Most of our research team is 100% bilingual. Many of the participants will only speak Spanish.

An association between the treatment outcome and the sociodemographic and health factors will be estimated using logistic regression techniques. We anticipate two logistic regression models: One describing possible explanations for no desire to obtain treatment, and the second describing factors for not completing treatment. These models will be analyzed at periodic time intervals during the grant period.

Analysis of immunoassays: While concordance between the TST and QuantiFERON confirms history of exposure to M. tuberculosis, the interpretation of discordance is not thoroughly understood due to the absence of a gold standard for LTBI. In our study population and design, we anticipate being able to complement the body of information regarding discrepant results between the TST and immunoassay. First, our population has a high proportion of BCG vaccinated individuals. Most are immigrants from Mexico and other Latin American countries where BCG is routinely implemented. Many migrate frequently (some of them daily!) between both countries, and receive medical treatment in Mexico where the costs are lower. Thus, the impact of BCG vaccination and when it was administered will be specifically addressed. Second, exposure to environmental mycobacterium is very likely in the colonias due to the low socioeconomic status and the poor hygiene associated with their living conditions. Addressing this issue is possible in part by analyzing the proportion of individuals that are positive for the M. avium specificity control. Third, determine if a negative-TST but positive-QuantiFERON may be an indication of a higher sensitivity in the latter assay. Our study is designed to perform a two-step TST testing to determine the proportion of individuals that become TST-positive upon re-exposure to the tuberculin antigen (27). Fourth, the use of digit preference for recording TST induration has been identified as a possible explanation for discordance between both immunoassays. In the present Task Order the operator-induced errors should be reduced following the training course at CDC. And fifth, the data from immunosuppressed individuals will be recorded, but we anticipate there may not be too many of these cases in our population. An association between discordant results and the sociodemographic characteristics and health condition of the subjects that was recorded in the questionnaire will be estimated using logistic regression techniques.

ANTICIPATED RESULTS OR OUTCOME

Expected outcomes: Anticipated outcomes of the project are:

1. Population based prevalence data for LTBI among high risk immigrant residents of colonias along the US-Mexico border. This is the first study of its kind and will allow us to understand the magnitude and precise nature of the risk of TB in the United States presented by this large migrant population.

2. Information on the factors associated with discordance between the TST and QuantiFERON assay and the advantages and limitations of each method. This information will be essential for the targeted LTBI treatment program because it should provide a more precise indication of the individuals that truly have LTBI.

3. Provide baseline information on the percent of patients who receive and complete treatment for LTBI among colonia residents in the Lower Rio Grande Valley (LRGV). This will be complemented with information on socio-demographic factors associated with LTBI and treatment adherence. Such information will be used to develop strategies to improve the completion rate among colonia residents treated for LTBI and to assess the contribution of incomplete treatment to the development of resistant strains.

4. The data from this study will provide population based information with spatial analysis in a community living a normal unconfined life in which TB circulates in accordance with its well established natural history. Only with this kind of data can effective control programs be established.

Reference List

1. CDC. Preventing and Controlling Tuberculosis Along the U.S.-Mexico Border. MMWR 2001;50:RR1.

2. Taylor J, Suarez L. Prevalence and risk factors of drug-resistant TB along the Mexico-Texas border. Am J Public Health 2002;90(2):271-3.

3. US Census Bureau. Detailed census 2000 supplementary survey data. 2002.
Ref Type: Internet Communication

4. Sanchez-Perez H, Flores-Hernandez J, Jansa J, Cayla J, Martin-Mateo M. Pulmonary tuberculosis and associated factors in areas of high levels of poverty in Chiapas, Mexico. Int J Epidemiol 2001;30(2):386-93.

5. Dutton, RJ, Weldon, M., Shannon, J, Bowcock, C, Tackett-Gibson, M, Blakely, C, Dyer, J, Jayasuriya, B, Worrall, W, and Betru, R. Survey of health and environmental conditions in Texas border counties. TDH No. 79-10828. 1-6-2000. Austin, TX, Texas Department of Health.
Ref Type: Report

6. Demographic Profile of U.S.-Mexico Border. August 28,2000. 2000. New Mexico State University. 2002.
Ref Type: Electronic Citation

7. Panoramic Views of Las Colonias. 2002. Southwest Texas State University. 5-29-2002.
Ref Type: Electronic Citation

8. Pepin, M. Texas colonias: an environmental justice case study. 1998. Our Lady of the Lake Univeristy. 2002.
Ref Type: Electronic Citation

9. Office of Colonia Initiatives-TDHCA. 2002. Texas Department of Housing and Community Affairs. 2002.
Ref Type: Electronic Citation

10. Gordon, Kierstan, , April Burcham, , Hector Galán, , Amanda Hirsch, , Natasha Rodriguez, and Sullivan, Maury. The forgotten Americas: Focus: Las Colonias. 2002. 2000. KLRU
PBS Online. 2002.
Ref Type: Electronic Citation

11. Bastida E, Cuellar I, Villas P. Prevalence of diabetes mellitus and related conditions in a south Texas Mexican American sample. J Community Health Nurs 2001;18(2):75-84.

12. Espino DV, Burge SK, Moreno CA. The prevalence of selected chronic diseases among the Mexican-American elderly: data from the 1982-1984 Hispanic Health and Nutrition Examination Survey. J Am Board Fam Pract 1991;4(4):217-22.

13. Hanis CL, Chu HH, Lawson K, Hewett-Emmett D, Barton SA, Schull WJ et al. Mortality of Mexican Americans with NIDDM. Retinopathy and other predictors in Starr County, Texas. Diabetes Care 1993;16(1):82-9.

14. CDC. Prevention and Control of Tuberculosis in Migrant Farm Workers (ACET). 41(RR10). 6-6-1992. MMWR.
Ref Type: Report

15. Division, Tuberculosis Elimination. Tuberculosis in Texas: Statistical report 1998-1999. 1998. Austin, Texas Department of Health (http://www.tdh.state.tx.us/tb).
Ref Type: Report

16. Division, Tuberculosis Elimination. Tuberculosis in Texas: Statistical report 1998-1999. 1998. Austin, Texas Department of Health (http://www.tdh.state.tx.us/tb).
Ref Type: Report

17. Small PM, Fujiwara PI. Management of tuberculosis in the United States. N Engl J Med 2001;345(3):189-200.

18. Targeted tuberculin testing and treatment of latent tuberculosis infection. American Thoracic Society. MMWR Morb Mortal Wkly Rep 2000;49(RR-6):1-51.

19. Menzies D. Interpretation of repeated tuberculin tests. Boosting, conversion, and reversion. Am J Respir Crit Care Med 1999;159(1):15-21.

20. Johnson PD, Stuart RL, Grayson ML, Olden D, Clancy A, Ravn P et al. Tuberculin-purified protein derivative-, MPT-64-, and ESAT-6-stimulated gamma interferon responses in medical students before and after Mycobacterium bovis BCG vaccination and in patients with tuberculosis. Clin Diagn Lab Immunol 1999;6(6):934-7.

21. Mazurek GH, LoBue PA, Daley CL, Bernardo J, Lardizabal AA, Bishai WR et al. Comparison of a whole-blood interferon gamma assay with tuberculin skin testing for detecting latent Mycobacterium tuberculosis infection. Jama 2001;286(14):1740-7.

22. Pottumarthy S, Morris AJ, Harrison AC, Wells VC. Evaluation of the tuberculin gamma interferon assay: potential to replace the Mantoux skin test. J Clin Microbiol 1999;37(10):3229-32.

23. Katial RK, Hershey J, Purohit-Seth T, Belisle JT, Brennan PJ, Spencer JS et al. Cell-mediated immune response to tuberculosis antigens: comparison of skin testing and measurement of in vitro gamma interferon production in whole-blood culture. Clin Diagn Lab Immunol 2001;8(2):339-45.

24. Levy PS, Lemeshow S. Sampling of populations: methods and applications. 3rded. New York: John Wiley & Sons; 2002.

25. Screening for tuberculosis and tuberculosis infection in high-risk populations. Recommendations of the Advisory Council for the Elimination of Tuberculosis. MMWR Recomm Rep 1995;44(RR-11):19-34.

26. Diagnostic Standards and Classification of Tuberculosis in Adults and Children. This official statement of the American Thoracic Society and the Centers for Disease Control and Prevention was adopted by the ATS Board of Directors, July 1999. This statement was endorsed by the Council of the Infectious Disease Society of America, September 1999. Am J Respir Crit Care Med 2000;161(4 Pt 1):1376-95.

27. Targeted tuberculin testing and treatment of latent tuberculosis infection. American Thoracic Society. MMWR Morb Mortal Wkly Rep 2000;49(RR-6):1-51.

28. Cohn DL. Treatment of latent tuberculosis infection: renewed opportunity for tuberculosis control. Clin.Infect.Dis. 2000;31(1):120-4.

29. Update: Fatal and severe liver injuries associated with rifampin and pyrazinamide for latent tuberculosis infection, and revisions in American Thoracic Society/CDC recommendations--United States, 2001. MMWR Morb.Mortal.Wkly.Rep. 2001;50(34):733-5.