Lessons for TB from the COVID-19 response: qualitative data from Brazil, India and South Africa.

BACKGROUND: Brazil, India and South Africa are among the top 30 high TB burden countries globally and experienced high rates of SARS-CoV-2 infection and mortality. The COVID-19 response in each country was unprecedented and complex, informed by distinct political, economic, social and health systems contexts. While COVID-19 responses have set back TB control efforts, they also hold lessons to inform future TB programming and services. METHODS: This was a qualitative exploratory study involving interviews with TB stakeholders (n = 76) in Brazil, India and South Africa 2 years into the COVID-19 pandemic. Interview transcripts were analysed using an inductive coding strategy. RESULTS: Political will - whether national or subnational - enabled implementation of widespread prevention measures during the COVID-19 response in each country and stimulated mobile and telehealth service delivery innovations. Participants in all three countries emphasised the importance of mobilising and engaging communities in public health responses and noted limited health education and information as barriers to implementing TB control efforts at the community level. CONCLUSIONS: Building political will and social mobilisation must become more central to TB programming. COVID-19 has shown this is possible. A similar level of investment and collaborative effort, if not greater, as that seen during the COVID-19 pandemic is needed for TB through multi-sectoral partnerships.

CONTEXTE: Le Brésil, l'Inde et l'Afrique du Sud figurent parmi les 30 pays les plus touchés par la TB dans le monde et ont connu des taux élevés d'infection et de mortalité dus au SARS-CoV-2. La réponse au COVID-19 dans chacun de ces pays a été sans précédent et complexe, en raison de contextes politiques, économiques, sociaux et de systèmes de santé distincts. Si les réponses au COVID-19 ont fait reculer les efforts de lutte contre la TB, elles permettent également de tirer des enseignements pour les futurs programmes et services de lutte contre la TB. MÉTHODES: Il s'agit d'une étude exploratoire qualitative comprenant des entretiens avec des acteurs de la lutte contre la TB (n = 76) au Brésil, en Inde et en Afrique du Sud, 2 ans après le début de la pandémie de COVID-19. Les transcriptions des entretiens ont été analysées à l'aide d'une stratégie de codage inductive. RÉSULTATS: La volonté politique ­ qu'elle soit nationale ou infranationale ­ a permis la mise en œuvre de mesures de prévention généralisées au cours de la riposte au COVID-19 dans chaque pays et a stimulé les innovations en matière de prestation de services mobiles et de télésanté. Les participants des trois pays ont souligné l'importance de la mobilisation et de l'engagement des communautés dans les réponses de santé publique et ont noté que l'éducation et l'information sanitaires limitées constituaient des obstacles à la mise en œuvre des efforts de lutte contre la TB au niveau communautaire. CONCLUSIONS: La volonté politique et la mobilisation sociale doivent occuper une place plus centrale dans les programmes de lutte contre la TB. La conférence COVID-19 a montré que c'était possible. Un niveau d'investissement et de collaboration similaire, voire supérieur, à celui observé lors de la pandémie de COVID-19 est nécessaire pour lutter contre la TB par le biais de partenariats multisectoriels.

Factors associated with non-completion of TB preventive treatment in Brazil.

BACKGROUND: Among Brazilian initiatives to scale up TB preventive therapy (TPT) are the adoption of the 3HP regimen (12 weekly doses of rifapentine and isoniazid [INH]) in 2021 and the implementation in 2018 of the TPT surveillance information system. Since then, 63% of the 76,000 eligible individuals notified completed TPT. Recommended regimens in this period were 6H, 9H (6 or 9 months of INH) and 4R (4 months of rifampicin).OBJECTIVE: To analyse the factors associated with TPT non-completion.METHODS: We analysed the cohort of TPT notifications from 2018 to 2020. Robust variance Poisson regression model was used to verify the association of TPT non-completion with sociodemographic, clinical and epidemiological variables.RESULTS: Of the 39,973 TPT notified in the study period, 8,534 (21.5%) were non-completed, of which 7,858 (92.1%) were lost to follow-up. Age 15-60 years (relative risk [RR] 1.27, 95% confidence interval [95% CI] 1.20-1.35), TPT with isoniazid (RR 1.40, 95% CI 1.19-1.64) and Black/mixed race (RR 1.17, 95% CI 1.09-1.25) were associated with a higher risk of non-completion.CONCLUSION: Individuals in situations of social and financial vulnerability such as being Black/pardo race, younger and on longer TPT regimens were more likely to be associated with TPT incompletion.

A public health approach to increase treatment of latent TB among household contacts in Brazil.

SETTING: Two consecutive trials were conducted to evaluate the effectiveness of a public health approach to identify and correct problems in the care cascade for household contacts (HHCs) of TB patients in three Brazilian high TB incidence cities.METHODS: In the first trial, 12 clinics underwent standardised evaluation using questionnaires administered to TB patients, HHCs and healthcare workers, and analysis of the cascade of latent TB care among HHCs. Six clinics were then randomised to receive interventions to strengthen management of latent TB infection (LTBI), including in-service training provided by nurses, work process organisation and additional clinic-specific solutions. In the second trial, a similar but streamlined evaluation was conducted in two clinics, who then received initial and subsequent intensive in-service training provided by a physician.RESULTS: In the evaluation phase of both trials, many HHCs were identified, but few started LTBI treatment. After the intervention, the number of HHCs initiating treatment per 100 active TB patients increased by 10 (95%CI - 11 to 30) in the first trial, and by 44 (95%CI 26 to 61) in the second trial.DISCUSSION: A public health approach with standardised evaluation, local decisions for improvements, followed by intensive initial and in-service training appears promising for improved LTBI management.

Cost and cost-effectiveness of tuberculosis treatment shortening: a model-based analysis.

BACKGROUND: Despite improvements in treatment success rates for tuberculosis (TB), current six-month regimen duration remains a challenge for many National TB Programmes, health systems, and patients. There is increasing investment in the development of shortened regimens with a number of candidates in phase 3 trials. METHODS: We developed an individual-based decision analytic model to assess the cost-effectiveness of a hypothetical four-month regimen for first-line treatment of TB, assuming non-inferiority to current regimens of six-month duration. The model was populated using extensive, empirically-collected data to estimate the economic impact on both health systems and patients of regimen shortening for first-line TB treatment in South Africa, Brazil, Bangladesh, and Tanzania. We explicitly considered 'real world' constraints such as sub-optimal guideline adherence. RESULTS: From a societal perspective, a shortened regimen, priced at USD1 per day, could be a cost-saving option in South Africa, Brazil, and Tanzania, but would not be cost-effective in Bangladesh when compared to one gross domestic product (GDP) per capita. Incorporating 'real world' constraints reduces cost-effectiveness. Patient-incurred costs could be reduced in all settings. From a health service perspective, increased drug costs need to be balanced against decreased delivery costs. The new regimen would remain a cost-effective option, when compared to each countries' GDP per capita, even if new drugs cost up to USD7.5 and USD53.8 per day in South Africa and Brazil; this threshold was above USD1 in Tanzania and under USD1 in Bangladesh. CONCLUSION: Reducing the duration of first-line TB treatment has the potential for substantial economic gains from a patient perspective. The potential economic gains for health services may also be important, but will be context-specific and dependent on the appropriate pricing of any new regimen.

Cost-effectiveness of the Xpert® MTB/RIF assay for tuberculosis diagnosis in Brazil.

INTRODUCTION: The Xpert® MTB/RIF assay is being implemented as a substitute for sputum smear microscopy (SSM) in many low and high tuberculosis (TB) burden countries, including Brazil, a country with low multidrug resistance and moderate human immunodeficiency virus co-infection rates. SETTING: Brazilian National TB Programme (NTP). OBJECTIVE AND DESIGN: We estimated the incremental cost-effectiveness ratio (ICER) of Xpert as a substitute for two SSM tests in the diagnosis of drug-susceptible TB. The costs for confirming each additional case and for avoiding treatment due to false-positive empirical diagnoses were estimated. RESULTS: The ICER was US$943 for each additional TB diagnosis and US$356 for each additional TB diagnosis with bacteriological confirmation, assuming 80% specificity of clinical diagnosis using both strategies. For every 100 000 patients with suspected TB, the NTP would spend an additional US$1.2 million per year to confirm 3344 more TB patients. The model was highly sensitive to specificity of clinical diagnosis after a negative test. CONCLUSION: Although the NTP has no threshold for cost-effectiveness, our model can provide support for decision makers in Brazil and other countries with a low prevalence of drug resistance among TB patients. Financial benefit can potentially be expected if physicians rely more on a negative Xpert result and empirical treatment is reduced.

Non-completion of latent tuberculous infection treatment among children in Rio de Janeiro State, Brazil.

BACKGROUND: Children with latent tuberculous infection (LTBI) are particularly vulnerable to progression to active tuberculosis (TB), and are thus a priority target for isoniazid preventive therapy (IPT). However, adherence to IPT is poor. We hypothesised that children from poorer families, with reduced access to health care and lack of understanding about the disease are more likely to default from IPT. METHODS: A questionnaire was administered to close child contacts or their parents at the time of prescribing IPT in three cities in Rio de Janeiro State. The children were followed prospectively. Treatment adherence was defined as taking 80% of prescribed doses. RESULTS: Among 1078 children screened for LTBI, 97 (8.9%) did not return for tuberculin skin test (TST) reading; 332 (30.8%) were TST-positive; 115/332 (34.6%) were prescribed IPT, 6 of whom did not initiate treatment and 11 did not adhere during the first 2 months; 25 additional children did not complete IPT. Overall non-completion was four times more frequent among those with lower income. Health care access and knowledge did not improve treatment completion. CONCLUSIONS: Substantial losses to follow-up occurred before IPT prescription; this should be further investigated. Among the children who started isoniazid, low income, but not difficult access or poor knowledge, increased the risk of treatment non-completion.

Patient costs for the diagnosis of tuberculosis in Brazil: comparison of Xpert MTB/RIF and smear microscopy.

SETTING: Manaus and Rio de Janeiro, two Brazilian state capitals with the country's fifth and sixth highest tuberculosis (TB) incidence rates (around 90/100,000 population in 2012). OBJECTIVE: To compare the costs of the Xpert MTB/RIF assay with those of standard care (two smears) in diagnosing TB from the patient's perspective. METHOD: We interviewed 218 patients diagnosed with TB in the previous 4 months by Xpert or smear microscopy. Information on non-medical direct costs for transportation and food, indirect costs such as time spent for diagnostic visits and socio-demographic data were gathered. RESULTS: The median patient income was US$390.24. Median total costs incurred by patients were 54% higher with the smear process than with Xpert (US$25.24 vs. US$16.44, P < 0.000) due to higher indirect and direct costs. Male patients incurred higher indirect costs (U$10.27 vs. US$7.51, P = 0.038), and patients in Manaus incurred higher total costs. CONCLUSIONS: Although the diagnosis and treatment of TB in Brazil are free of charge, non-medical direct and indirect costs for patients may represent important barriers to accessing appropriate care. Compared to standard care, Xpert reduced the financial burden for patients. These findings support the decision to scale-up Xpert technology in the country.

Changes in QuantiFERON®-TB Gold In-Tube results during treatment for tuberculous infection.

SETTING: Randomised trial comparing 9 months of isoniazid with 4 months of rifampicin for the treatment of high-risk tuberculin skin test positive subjects in Rio de Janeiro, Brazil. OBJECTIVES: To compare QuantiFERON®-TB Gold In-Tube (QFT-GIT) responses before and 1, 4 and 9 months after starting treatment for latent tuberculous infection (LTBI) according to adherence to one of the two regimens. DESIGN: Participants in the trial were invited to undergo serial QFT-GIT. Within-subject differences at different time points were analysed as quantitative responses and categorised as positive or negative using different cut-off points. RESULTS: Of 215 participants, 118 completed treatment, of whom 58 underwent all three tests; and 97 did not complete treatment, of whom 10 underwent all tests. After 1 month of treatment, there was no significant difference in QFT-GIT response between the groups. After 4 and 9 months, reversions were more frequent in non-adherent subjects. Marked within-subject fluctuations were observed. No cut-off point could be established at which QFT-GIT responses were consistently positive or associated with adherence or type of treatment. CONCLUSION: Frequent within-subject variability in QFT-GIT responses, not associated with LTBI treatment, makes it difficult for clinicians to interpret QFT-GIT conversions and reversions.

Artificial neural network models to support the diagnosis of pleural tuberculosis in adult patients.

BACKGROUND: Clinicians in countries with high tuberculosis (TB) prevalence often treat pleural TB based on clinical grounds, as the availability and sensitivity of diagnostic tests are poor. OBJECTIVE: To evaluate the role of artificial neural networks (ANN) as an aid for the non-invasive diagnosis of pleural TB. These tools can be used in simple computer devices (tablets) without remote internet connection. METHODS: The clinical history and human immunodeficiency virus (HIV) status of 137 patients were prospectively entered in a database. Both non-linear ANN and the linear Fisher discriminant were used to calculate performance indexes based on clinical grounds. The same procedure was performed including pleural fluid test results (smear, culture, adenosine deaminase, serology and nucleic acid amplification test). The gold standard was any positive test for TB. RESULTS: In pre-test modelling, the neural model reached >90% accuracy (Fisher discriminant 74.5%). Under pre-test conditions, ANN had better accuracy compared to each test considered separately. CONCLUSIONS: ANN are highly reliable for diagnosing pleural TB based on clinical grounds and HIV status only, and are useful even in remote conditions lacking access to sophisticated medical or computer infrastructure. In other better-equipped scenarios, these tools should be evaluated as substitutes for thoracocentesis and pleural biopsy.

Tuberculosis treatment outcomes and socio-economic status: a prospective study in Duque de Caxias, Brazil.

A prospective study was conducted to evaluate tuberculosis treatment outcomes according to socio-economic status (SES) using different classification criteria. Patients aged ≥18 years under treatment for ≤8 weeks were interviewed. Outcomes were classified as successful (cure/completed) or unsuccessful (default/failure/death). The overall treatment default ratio was 20.9% and the unsuccessful outcome rate was 24.1%. Unsuccessful treatment was associated with SES according to any criteria used, except for the definition of poverty line. Poverty seems to be hampering the achievement of the World Health Organization targeted 90% cure rate in developing settings.

Factors associated with treatment adherence in a randomised trial of latent tuberculosis infection treatment.

SETTING: Randomised controlled trial of latent tuberculosis infection (LTBI) treatment in 10 clinics in Canada, Saudi Arabia and Brazil. OBJECTIVE: To identify early predictors of LTBI treatment adherence, including pre-treatment characteristics. DESIGN: Patients randomised to 4 months of rifampicin (RMP; n = 420) or 9 months of isoniazid (n = 427) were monitored for adherence using an electronic device. Outcomes were 1) treatment completion, defined as intake of >or=80% of the prescribed doses, and further categorised as completed within the allotted time or not; and 2) treatment regularity, measured by the time interval between doses. Relative risk (RR) and adjusted odds ratios (aOR) of patients' pre-treatment characteristics and adherence at first follow-up visit were calculated. RESULTS: Completion of treatment was higher with RMP (aOR 4.3, 95%CI 2.7-6.8). Early predictors (first follow-up visit) of non-adherence were late first visit attendance (RR for completion in time 0.9, 95%CI 0.8-0.98), >20% of missed doses (RR 0.4, 95%CI 0.3-0.6) and greater variation of hours between doses (0.209 vs. 0.131, P < 0.001). Serious adverse events were not associated with irregularity of treatment. CONCLUSION: The shorter RMP regimen was associated with better adherence. Patients with poor adherence could be identified at the first follow-up visit from their punctuality in follow-up, missed doses and variability of pill-taking.

Uses of tuberculosis mortality surveillance to identify programme errors and improve database reporting.

BACKGROUND: In 2006, 848 persons died from tuberculosis (TB) in Rio de Janeiro, Brazil, corresponding to a mortality rate of 5.4 per 100 000 population. No specific TB death surveillance actions are currently in place in Brazil. SETTING: Two public general hospitals with large open emergency rooms in Rio de Janeiro City. OBJECTIVE: To evaluate the contribution of TB death surveillance in detecting gaps in TB control. METHODS: We conducted a survey of TB deaths from September 2005 to August 2006. Records of TB-related deaths and deaths due to undefined causes were investigated. Complementary data were gathered from the mortality and TB notification databases. RESULTS: Seventy-three TB-related deaths were investigated. Transmission hazards were identified among firefighters, health care workers and in-patients. Management errors included failure to isolate suspected cases, to confirm TB, to correct drug doses in underweight patients and to trace contacts. Following the survey, 36 cases that had not previously been notified were included in the national TB notification database and the outcome of 29 notified cases was corrected. CONCLUSION: TB mortality surveillance can contribute to TB monitoring and evaluation by detecting correctable and specific programme- and hospital-based care errors, and by improving the accuracy of TB database reporting. Specific local and programmatic interventions can be proposed as a result.

Administrative measures for preventing Mycobacterium tuberculosis infection among healthcare workers in a teaching hospital in Rio de Janeiro, Brazil.

Tuberculosis (TB) is an occupational disease of healthcare workers (HCWs). Administrative and engineering interventions simultaneously implemented in hospitals of developed countries have reduced the risk of nosocomial transmission of M. tuberculosis. We studied the impact of administrative infection control measures on the risk for latent TB infection (LTBI) among HCWs in a resource-limited, high-burden country. An intervention study was undertaken in a university-affiliated, inner-city hospital in Rio de Janeiro, where routine serial tuberculin skin testing (TST) is offered to all HCWs. From October 1998 to February 2001, the following infection control measures were progressively implemented: isolation of TB suspects and confirmed TB inpatients, quick turnaround for acid-fast bacilli sputum tests and HCW education in use of protective respirators. Among 1336 initially tested HCWs, 599 were retested. The number of TST conversions per 1000 person-months during and after the implementation of these measures was reduced from 5.8/1000 to 3.7/1000 person-months (P=0.006). The most significant reductions were observed in the intensive care unit (from 20.2 to 4.5, P<0.001) and clinical wards (from 10.3 to 6.0, P<0.001). Physicians and nurses had the highest reductions (from 7.6 to 0, P<0.001; from 9.9 to 5.8, P=0.001, respectively). We conclude that administrative measures for infection control can significantly reduce LTBI among HCWs in high-burden countries and should be implemented even when resources are not available for engineering infection control measures.

Two-step tuberculin skin test and booster phenomenon prevalence among Brazilian medical students.

SETTING: Five medical schools in three cities in Rio de Janeiro State, Brazil, with different tuberculosis (TB) incidence rates. OBJECTIVE: To evaluate the prevalence of the booster phenomenon and its associated factors in a young universally BCG-vaccinated TB-exposed population. DESIGN: A two-step tuberculin skin test (TST) was performed among undergraduate medical students. Boosting was defined as an induration > or =10 mm in the second TST (TST2), with an increase of at least 6 mm over the first TST (TST1). The association of boosting with independent variables was evaluated using multivariate analysis. RESULTS: Of the 764 participants (mean age 21.9 +/- 2.7 years), 672 (87.9%) had a BCG scar. The overall booster phenomenon prevalence was 8.4% (95%CI 6.5-10.6). Boosting was associated with TST1 reactions of 1-9 mm (aOR 2.5, 95%CI 1.04-5.9) and with BCG vaccination, mostly after infancy, i.e., after age two years (aOR 9.1, 95%CI 1.2-70.7). CONCLUSION: The prevalence of the booster phenomenon was high. A two-step TST in young BCG-vaccinated populations, especially in those with TST1 reactions of 1-9 mm, can avoid misdiagnosis as a false conversion and potentially reduce unnecessary treatment for latent TB infection.

Novel tests for diagnosing tuberculous pleural effusion: what works and what does not?

Tuberculous pleuritis is a common manifestation of extrapulmonary tuberculosis and is the most common cause of pleural effusion in many countries. Conventional diagnostic tests, such as microscopic examination of the pleural fluid, biochemical tests, culture of pleural fluid, sputum or pleural tissue, and histopathological examination of pleural tissue, have known limitations. Due to these limitations, newer and more rapid diagnostic tests have been evaluated. In this review, the authors provide an overview of the performance of new diagnostic tests, including markers of specific and nonspecific immune response, nucleic acid amplification and detection, and predictive models based on combinations of markers. Directions for future development and evaluation of novel assays and biomarkers for pleural tuberculosis are also suggested.

McNemar chi2 test revisited: comparing sensitivity and specificity of diagnostic examinations.

When evaluating a novel diagnostic examination for clinical use, it should be compared with a reference standard, defined as the best available examination, which may include clinical and laboratory criteria. The novel examination and reference standard's results are usually presented in the form of a 2 x 2 table, which allows calculation of sensitivity, specificity and accuracy. It has been recommended that the measures of statistical uncertainty should be reported, such as the 95% confidence interval, when evaluating the accuracy of diagnostic examinations. Comparing the difference in sensitivity or specificity of a novel examination with the reference standard is important when evaluating its usefulness. The McNemar chi(2) test, used to compare discordance of two dichotomous responses, can be applied for this purpose. However, applying the McNemar test to a 2 x 2 table for comparing the accuracy of examinations is not recommended, since this test is sensitive to the proportion of positive versus negative subjects. Moreover, if the novel examination has higher sensitivity than the one considered as the reference standard, constructing a classic 2 x 2 table would result in low specificity of the novel examination. Thus, in order to compare sensitivities and specificities between examinations, this table is inappropriate and an independent reference standard is necessary. In this article, we propose the use of the McNemar chi(2) test to compare sensitivities between examinations using a 2 x 2 table exclusively among diseased patients, defined by a set of criteria and follow-up of patients. Likewise, specificities can be compared applying the McNemar test among healthy individuals.

Pleural fluid ADA, IgA-ELISA and PCR sensitivities for the diagnosis of pleural tuberculosis.

The diagnosis of pleural tuberculosis (pTB) is difficult, and more sensitive and specific techniques are needed. In the period August 1998 to November 2002, we evaluated 132 patients with a pleural effusion submitted to a thoracentesis and pleural biopsy in a tertiary care hospital in Rio de Janeiro, Brazil. Three tests were performed and compared in the pleural fluid: ADA activity measurement, IgA-ELISA for two combined specific Mycobacterium tuberculosis antigens, and polymerase chain reaction (PCR) for detection of M. tuberculosis DNA. Ninety-five patients (72%) were given a final diagnosis of pTB. Overall histopathologic sensitivity was 77%. The sensitivities of pleural fluid culture and AFB smear were 42% and 1%, respectively. Twenty-one (22%) additional patients had a clinical diagnosis of pTB. Median follow-up time of all TB patients after the completion of antituberculous treatment was 13 months. Sensitivities of ADA, IgA-ELISA and PCR were 91%, 78% and 82%, while specificities were 93%, 96% and 85%, respectively. Only ADA sensitivity was significantly higher than the histopathologic examination (McNemar chi2 test; p = 0.002) and also significantly higher than ELISA (p = 0.049), but not higher than PCR (p = 0.143). We conclude that the routine use of ADA activity measurement in pleural fluid can obviate the need for a pleural biopsy in the initial diagnostic approach to pleural effusions, while IgA-ELISA and PCR techniques, potentially more specific tests, need further refinement to improve their accuracy.

Tuberculosis teaching in Brazilian medical schools.

To achieve tuberculosis (TB) control, National Tuberculosis Programme guidelines should be implemented effectively. In a survey conducted in 2005-2006, 33 Brazilian medical school coordinators answered a questionnaire about TB education. The median time dedicated to TB was 27 h (4-119 h), spread over several disciplines, mainly biological and clinical. This included 12 h (0-88 h) of practical activities, mainly in university hospitals (53%). The recommendation to offer human immunodeficiency virus testing for TB patients was taught in only 54% of the schools. TB education in Brazil is fragmented and restricted to a biological approach, while field activities are insufficient and carried out in inadequate settings. Important changes to the TB curriculum are necessary.

Latent tuberculosis infection among undergraduate medical students in Rio de Janeiro State, Brazil.

SETTING: Five medical schools in three cities with different tuberculosis (TB) incidence rates in Rio de Janeiro State, Brazil. OBJECTIVE: To estimate prevalence of and associated factors for latent tuberculosis infection (LTBI) among medical students. DESIGN: A cross-sectional survey was conducted among undergraduate students in pre-clinical, early and late clinical years from schools in cities with low (28/100,000), intermediate (63/100,000) and high (114/100,000) TB incidence rates. Information on socio-demographic profile, previous BCG vaccination, potential TB exposure, co-morbidity and use of respiratory protective masks was obtained. A tuberculin skin test (TST) was performed using the Mantoux technique by an experienced professional. A positive TST, defined as induration > or = 10 mm, was considered LTBI. RESULTS: LTBI prevalence was 6.9% (95%CI 5.4-8.6). In multivariate analysis, male sex (adjusted odds ratio [aOR] 1.8; 95% CI 1.1-3.0), late clinical years (aOR 1.9; 95% CI 1.01-3.5), intermediate TB incidence (aOR 4.3; 95% CI 1.3-14.6) and high TB incidence in the city of medical school (aOR 5.1; 95% CI 1.6-16.8) were significantly associated with LTBI. CONCLUSIONS: The higher prevalence of LTBI in late clinical years suggests that medical students are at increased risk for nosocomial Mycobacterium tuberculosis infection. The implementation of a TB control program may be necessary in medical schools, particularly in cities with higher TB incidence.