Prevalence and Risk of Latent Tuberculosis Infection in Primary Health Care Workers at La LisaMunicipality, Havana, Cuba
Dalila Martínez-Hall, Susana Borroto-Gutiérrez, Lellanis Arroyo-Rojas, Edilberto González-Ochoa*
Pedro Kourí Institute of Tropical Medicine, Havana City, Cuba
Background: Tuberculosis transmission is a risk to health care workers (HCW) Objectives: To determine the prevalence of latent tuberculosis infection (LTBI), the annual risk of infection and its possible association with some predictive variables. Methods: A cross-sectional study in 1063 HCW at La Lisa municipality policlinics, Havana, was carried out during 2009-2011; a Tuberculin Skin Test (TST) was applied, along with a structured questionnaire exploring general and on exposition data. The reading was performed in 905 HCW at 72 hours, defining as positive ≥10mm induration and tuberculin conversion as ≥10 mm increasing in the second TST of those previously negative. To prove the possible influence of predictive variables on the presence of infection, a bivariate analysis and a multivariate model were applied. Results: The prevalence of LTBI was 15,2%, with 3,7% of AIR. The highest prevalence in Berovides Policlinic (19,3%) and the lowest in Fernández (10,5%). The highest annual risk of infection was in Labra (6,4%) and the lowest in Fonseca (1%). The variables possibly associated to LTBI were: belonging to Labra and Berovides policlinics, having contact with tuberculosis patients, being ≥20 years working there and belonging to the following occupational groups: administrative, nurse, maintenance staff and health technicians. Conclusions: The transmission of tuberculosis infection, even in low levels, is possibly associated to the labor environment; so it would be advisable to value the application of preventive chemotherapy in these personnel and to maintain the appropriate measures of infection control in the institutions.
Tuberculosis, Latent Tuberculosis Infection, Prevalence, Primary Health Care
Received: April 9, 2015
Accepted: June 25, 2015
Published online: July 16, 2015
@ 2015 The Authors. Published by American Institute of Science. This Open Access article is under the CC BY-NC license. http://creativecommons.org/licenses/by-nc/4.0/
Transmission of Mycobacterium tuberculosis is a risk in health-care settings. The magnitude of risk varies by setting, occupational group, the prevalence of tuberculosis (TB) in the community, the population served and the effectiveness of TB infection control measures. The risk may be higher in areas where patients are treated before diagnosis and initiation of TB treatment. 1
In Cuba, TB incidence rates are among the lowest in the region. The national crude incidence rate in 2013 was 6,4 per 100 000 population and it has maintained below 7 during the last ten years, with the lowest in 2009 (5,9 per 100 000 population). 2
The National TB Control Program (NTCP) takes actions to eliminate the disease as a public health problem, based on a differentiated intervention on the most vulnerable groups of acquiring the disease, including health care workers (HCW).3
Havana is the capital and most densely populated province, with the higher economic importance in Cuba, where main TB generating outbreaks are concentrated. That behavior largely determines the results of the country, notifying about a quarter part of the national incidence.2
La Lisa municipality, as part of strengthening of the NTCP, is immersed in an active screening for early detection of TB and other respiratory diseases. Despite this, even 60% of cases are diagnosed in secondary health care level.3 In 2008 the municipality was ranked as high risk in the stratification performed to execute the actions derived of the strengthening of the NTCP funded by The Global Fund.4 After that, a decreasing in the incidence from 12 cases during 2009 (8,6 x 100 000 population) to 9 cases during 2010 (6,2 x 100 000) and 6 cases during 2011 (3,9 x 100 000) was observed, with a decrease of about 3 cases per year and 4,7% annual in the incidence rate. (Martínez D. Un-published data)
In Cuba there have been some recent studies evaluating the TB risk in hospital workers of different categories,5-7 but have not been any studies on workers of Primary Health Care level, which is currently diagnosing around 40% of TB cases and gives medical treatment and monitoring to 100% of the cases under Directly Observed Treatment Short-course (DOTS).
As part of the TB risk assessment in HCW from primary health care of La Lisa, in this study the prevalence of latent tuberculosis infection (LTBI), the annual risk of infection and the possible association with predictor variables was determined.
Setting: La Lisa municipality is located at southwest of Havana province; with a geographical area of 37,1 Km2. The total population varied from 133,255 inhabitants in 2009 to 133,520 in 2011 and the population density ranged from 3588 inhabitants per km2in 2009 to 3595 in 2011.8 It is divided into seven People´s Councils (spatial administrative division of the municipal government), that are served by five health areas (territorial health units) 9: Aleida Fernández, Cristóbal Labra, Elpidio Berovides, Pulido Humarán and Pedro Fonseca.
Design: A cross-sectional study was conducted during the period 2009-2011.
Study population: Of the 1294 HCW available (working at the time of the study), all those who agreed to participate were included once did not have any of the following exclusion criteria: pregnancy, acute infectious disease. The total number of participants was 1063.
Procedures: In 2009 they underwent a Tuberculin Test (TST) and those negative, were tested again a year later to assess tuberculin conversion. The TST was performed according to WHO standard technique 10 by two trained and standardized nurses, using 2 Tuberculin Units (0.1 ml) of PPD RT-23 from Evans. The reading was performed at 72 hours, considering as positive those indurations ≥10 millimeters. Individuals without response (0 mm) were considered as non-reactors, and those with ≥15 mm in diameter as hyper-reactors. The increasing in ≥10 mm in the second TST in those patients previously negative was considered tuberculin conversion.
In addition, we applied a structured questionnaire to explore the possible association with the presence of infection of the independent variables: age, sex, occupational category (administrative, nurses, doctors, maintenance personnel, services, health technicians and not health technicians), departments, working time in the center, history of contact with TB patients and presence of BCG vaccination scar. The history of contact with TB patients was considered exposure to TB; workers with BCG vaccination scar were considered vaccinated. The departments were grouped into six areas: Medical Family Clinics, Consulting Specialties, Diagnosis and Treatment, Maintenance, Clerk, Services and Leader’s cabinets.
Statistical analysis: LTBI prevalence rates for each variable were calculated and homogeneity tests using Kruskal-Wallys for samples and Levene for variances were applied. The induration diameter average was analyzed according to the presence or not of BCG scar, taking into account five categories: nonreactors (0 mm), negative reactors (1-9), negative (0-9), positive (≥ 10 mm) and hyper-reactors (≥ 15 mm). The Annual Risk of Infection (ARI) was estimated using the direct method, by calculating the tuberculin conversion during a year (% of individuals who converted in the second TST up to the total of those negative in the first one).
To test the possible influence of the independent variables on the presence or absence of infection, bivariate analysis between each of the independent covariates and the response variable (presence of infection) using Prevalence Risk Ratio (PRR) 11 was initially performed, with reference to the lower prevalence for each covariate. The PRR > 1 was considered significant with CI- 95% and p <0.05.
A multivariate binary logistic regression model by backward method (Wald) was then applied introducing in the model the results with p <0.05 in bivariate analysis. Calculations were performed with the statistical software EpiInfo 2002 (Centers for Disease Control and Prevention, Atlanta, GA, USA), tests of homogeneity in XLSTAT 2009 and logistic regression in SPSS 11.5 (SPSS Inc., Chicago, IL).
Ethical considerations: informed consent from all study participants was obtained. This research was approved by the Ethics Committee of the Pedro Kourí Institute (IPK), Havana, Cuba.
Up to 905 read TST, 15.2% were positive. The higher prevalence of reactors was at the Berovides polyclinic (19.3%), and the lowest in Fernández (10.5%), with statistically significant difference between groups (p = 0.02). The mean diameter of induration (including non-reactors) was 3.8 mm (SD 5.5) for the municipality, and ranged from 3 mm in Fernández and Fonseca to 5 mm in Berovides; 45.5% were non-reactors, figure ranged from 37.4% in Berovides to 53% in Fonseca. (Table 1)
Most HCW (91.3%) are vaccinated and comparing the results of TST with a history of BCG vaccination (Table 2), no significant differences for non-reactors, nor for total negative or positive were found, including hyper-reactors between vaccinated and unvaccinated.
From 821 workers (90.7%) that were negative, 30 (3.7%) converted to positive in the second TST. The lowest rate of tuberculin conversion was in the polyclinic Fernández (1%) and the highest in Labra (6.4%). (Table 3)
In bivariate analysis of possible factors that might be related to the presence of LTBI in HCW of La Lisa, values for sex, previous BCG vaccination and age were not significantly greater than one, depending groups used to compare these variables. Working in polyclinics Labra (RRP=1.74; p=0.027) and Berovides (RRP=1.84; p=0.017) were found with a statistically significant RRP, with reference to Fernández which had the lowest prevalence of LTBI. (Table 4)
Workers who reported previous contact with TB cases had a prevalence 2.76 times higher than those who did not report such a contact (p=0.000). Also significant was the fact of having been working for more than two years in the primary health care of La Lisa, compared with those who worked less than two years.
The administrative staff (PRR=2.75; p=0.010), nurses (PRR=2.62; p=0.001), maintenance personnel (PRR=3.50; p=0.000), and health technicians (PRR = 2.06; p=0.023) had significantly higher prevalence than physicians, occupational group with the lowest prevalence of reactors (7.6%).
Grouping HCW by work areas, PRR was significantly higher in Medical Family Clinics (PRR = 2.38; p=0.021) and maintenance workers (PRR=3.91; p=0.009) compared with leaders cabinets that had the lowest PRR.
In multivariate analysis, the best model found also included as variables associated with the presence of LTBI: work in the polyclinics Labra and Berovides, referred contact with TB cases, working more than 20 years in the institution, and belonging to the following occupational categories: administrative, nurses, health technicians and maintenance workers (Table 5).
The prevalence of LTBI and the tuberculin conversion as measure of ARI in primary health care workers of La Lisa is much lower than that found in some hospitals in the province. However, it was higher than that reported by the municipality in all risk groups.
The transmission of M. tuberculosis in primary health care workers may be associated with risk of exposure to patients seen, the prolonged work in this sector and some occupational categories and work areas where workers are more exposed to Mycobacterium tuberculosis.
TST was used to diagnose ITBL in this research, because the Interferon Gamma Release Assay (IGRA) is not yet in use by the Cuban NTCP since the high costs of this modern test. Even that, the recently published WHO Guidelines for management of LTBI recommends the use of both test, taking in account the advantage and disadvantage of them.12
In this research all TST were performed by two nurses who were kept constant until the end of it, which is a strength that increases the reliability of the results. It is useful to repeat the TST two weeks after the first survey to study the "Booster" phenomenon. 13 It is a limitation of this study not having made it to those not reactors cases for logistical reasons; but considering that a high cut-off point was used, it is expected that most individuals with reactions due to BCG or non-tuberculous mycobacteria don´t react more than 9 mm with the booster.14
The prevalence of TST reactors among primary health care workers in La Lisa is higher than that found in the risk population surveyed in the municipality during 2009-2011, in which positivity of 6.4% on average has been reported up to the total TST conducted by the NTCP. (Unpublished data: Indicators to Strengthen NCTP, Municipal Unit of Hygiene and Epidemiology, La Lisa, 2010.) They were similar in the five health areas, with values between 5.2% in Labra and 7.3% in Fernández.
During the national research in vulnerable groups, the TST positivity found was 2.9%; while in Havana province and its high-risk municipalities were 3.3 and 5.4% respectively. (Unpublished data: Ministry of Public Health. Indicators of Strengthen PNCTB Cuba, 2010.)
However, both LTBI and tuberculin conversion are well below those reported in hospitals HCW in the province. Pedro Kourí Institute (IPK), also located in La Lisa, has reported LTBI prevalence from 33.5% to 55.2% in HCW, with 14.6% of tuberculin conversion. 5,15 This is logical considering that in IPK more than 40 TB patients are hospitalized in average during a year, besides the National TB Reference Laboratory is located inside its facilities. Therefore, the exposure risk is higher to its staff, while in the whole La Lisa municipality were handled only 34 TB cases during the period 2007-2011, averaging 7 cases per year. Desegregating by health areas, this average drops to 1-2 cases per year in Fernández and Humarán, and less than one case in Fonseca. (Unpublished data: Municipal Unit of Hygiene and Epidemiology, La Lisa. Indicators of the National TB Control Program, 2009-2011.)
Another published study of the Pneumology Hospital in Havana, reports 50.8% of LTBI in the staff, with 35.9% of tuberculin conversion, the highest among the studies performed in the country. This is consistent with the higher risk level of exposure to their HCW, as this is the national reference center for the treatment of complicated TB cases.16
However, our figures are similar to those found in a surgical teaching hospital of Santiago de Cuba with 15.4% of TST reactors in their staff. 6Despite this is a provincial hospital, in the lower prevalence of LTBI and tuberculin conversion rate also might influence the TB incidence in the general population, which is lower in Santiago de Cuba province (about 5 x 105) compared with Havana, which contributes about 25% of the country's cases annually and has kept rates above 8 x 105 in the last five years. 2
Most LTBI prevalence studies published in the world are in hospital workers, and a systematic review by Joshi R et al. found that the LTBI prevalence in HCW was 54% in average;17 in another review by Menzies et al. of 42 papers a LTBI prevalence of 63% in low/middle-income countries and 24% in high-income countries was found.18
Significantly higher were the PRR in two health areas, Berovides and Labra, where more TB cases were treated in the period analyzed (10 cases each), and where only TB cases were found in HCW (one in each polyclinic). Berovides and Labra are the polyclinics that serve the most densely populated areas of the municipality, the council of St. Augustine by Berovides, and the councils Alturas de La Lisa and Versalles-Coronela by Labra.
In this study we found higher LTBI prevalence in maintenance personnel, administrative and nurses, which partially differs from other studies published. Nurses usually are among the highest risk, which is consistent with studies published in both national 6-7 and international literature.19-23 Surprisingly, we found doctors with the lowest prevalence of LTBI, what could be explained because in practice, nursing staff is in daily and close contact with patients on treatment application, while doctors evaluate patients twice a month.
In a study in primary health care workers in Russia a higher positivity was found in nurses (32.8%) and physicians (25.5%), although a high positivity in administrative and other support staff was also found. 20 In another study in Portugal a high LTBI positivity in employees with functions that are not supposed to involve risk, as administrative and assistants, was also found. 22
Regarding workspaces, surprisingly, the maintenance personnel, who perform tasks unrelated directly with patients, appears associated with LTBI. Much of maintenance personnel are workers with low educational and social level who could have habits such as alcoholism,23 which involves greater risk than to healthy people. Furthermore, the higher prevalence in workers with occupations involving less risk may be related to infection in the community or in other health facilities where they have worked before.
However, the areas of diagnosis and treatment were not associated. This may be because in these polyclinics, despite around 2335 sputum smears were tested in 2007-2011, only 0.08% were positive, and both the perception of risk and the implementation of preventive infection control measures are generally higher in these areas.
Another area associated to LTBI was Medical Family Clinics (MFC), where monitoring and strictly supervised care is given to all TB patients, specifically by nurses, which precisely had one of the highest LTBI prevalence, not like that in physicians who had the lowest numbers. It is necessary to note that a large number of doctors and nurses in the MFC have been in internationalist health missions in countries with high TB burden, so it is also possible that many of these nurses have become infected during those stays and no in their MFC.
The incidence of LTBI in HCW is related to TB incidence in the general population of the area. A study published by Vries et al. reported in the Netherlands that 42% had acquired the infection in the hospital, 28% in the community and 30% abroad. 24 The risk of infection is more related to the contact proximity to the infection source than with time of exposure. 25-29
We found association between a very long working time (over 20 years) in the areas and the presence of LTBI, which is consistent with national 16 and international 21,29 literature reviewed about the influence of prolonged exposure on HCW.
In Cuba, BCG vaccination coverage is nearly 100% at birth, and until 1992 a reactivation dose was applied to 9 years old scholar children3, so it is justified that a high percentage of respondents have BCG scar. Nevertheless, a high percentage of vaccinated were not reactors.
Because the post-vaccination allergy declines over time 30-31, and in Cuba it was 3-4 mm in children under 15 years more than 30 years ago 32-33, a high cut-off point could provide a high likelihood to differentiate with minimal error, a large proportion of natural allergies (including cross non-tuberculosis mycobacterial infections) an post-vaccination allergies.
In a compilation of international data of 1980-2005 on the effect of BCG vaccination on TST results in general population found a mediocre performance of TST in people of all ages and from vaccinated countries with low TB incidence 34, but it was a good way of detecting LTBI in vaccinated adults from countries with intermediate and high incidence of TB. However, a recent study in Sweden 35 on health students, the specificity of TST with different cut-offs points were estimated, and for ≥ 10 mm this was 97.5%, while for ≥ 15 mm was 99.3%.
In our study, despite most workers are vaccinated, no difference for both cut-offs points (≥ 10 and ≥ 15 mm) between vaccinated and unvaccinated were found, and BCG vaccination was not predictor of tuberculosis infection in multivariate analysis, reinforcing the idea that using a high cut-off point does differentiate infected from uninfected. More than 30 years have elapsed since BCG vaccination for most of the HCW studied. Taking in account that BCG only protect against severe TB in infants, this protection is no longer active for the studied population.
In conclusion, the prevalence of Latent TB Infection and the Annual Risk of Infection in Primary Health Care workers of La Lisa are higher in areas where more cases of TB are served, but lower than in other health care levels in Havana province. The transmission of tuberculosis infection in workers of Primary Health Care in La Lisa, although at low levels, is possibly associated with the work environment, so we suggested assessing the implementation of preventive chemotherapy to the infected personnel at higher risk and maintaining appropriate infection control in the institutions. Even that is recommended by the NTCP3, infection control and TB surveillance plans in HCW at primary health care are not yet applied. This should be a priority to implement by the NTCP in the next years to reach the TB elimination as is planned.
|Health Areas||N||No reactors (PDT=0 mm)||positive TST (≥ 10 mm)||Induration|
Source: Surveys performed.
|Diameter average (mm)||With BCG Scar||Without BCG Scar||Total||P values|
|1 a 9||282||19.5||31||22.6||313||19.8||0.37|
|0 a 9||632||43.7||60||43.8||692||43.7||0.98|
Source: Surveys performed.
|Health areas||Previous TST negative||Tuberculin Conversion|
Source: Surveys performed
TST- Tuberculin Skin Test.
|Variables||TST positive||%||TST negative||%||PRR||P value|
|Health Areas||A. Fernández||20||10.5||171||89.5||Reference|
|Working time (years)||≤ 2||19||8,3||209||91,7||Reference|
|No health Technics||9||10.8||74||89.2||1.43||0.403|
|Specialized medical office||14||15,2||78||84.8||1.87||0.144|
|Diagnostic y Treatment||39||15,9||206||89.1||1.96||0.073|
Source: Surveys performed.
TST- Tuberculin Skin Test, MFC- Medical Family Clinics.
|Variable||p-values||OR||C.I. 95,0% OR|
|Polyclinic C. Labra||0.027||2.04||1.1-3.8|
|Polyclinic E. Berovides||0.003||2.62||1.4-5.0|
|Working time > 20 years||0.011||8.52||1.6-44.1|
Source: Surveys performed.
OR: Odds Ratio , CI: Confidence Interval
ARI: Annual Risk of Infection
DOTS: Directly Observed Therapy Short-Course
HCW: Health Care Workers
IPK: Pedro Kourí Institute
LTBI: Latent Tuberculosis Infection
NTCP: National Tuberculosis Control Program
TST: Tuberculin Skin Test
MFC: Medical Family Clinics
 Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings, 2005. MMWR 2005; 54(17):1-151.
 Ministry of Public Health. Annual Health Statistics Report 2009-2013. National Health Statistics Bureau. Minsap. Acceded 12/11/2014. Available in: http://www.sld.cu/sitios/dne/
 Ministry of Public Health. National Program and Proceedings for Tuberculosis Prevention and Control. Havana, Cuba. 2013. ECIMED. Available in: http://www.sld.cu/galerias/pdf/sitios/tuberculosis/manual.pdf
 Ministry of Public Health. Strengthening the national TB Control Program in the Republic of Cuba. Acceded 11/03/2014 Available in: http://portfolio.theglobalfund.org/es/Grant/Index/CUB-708-G03-T#tab2
 Borroto S, Fernández R, Castro O, González E, Armas L. Evaluación del Riesgo de Tuberculosis en los Trabajadores del Instituto Pedro Kourí. Med. y Seguridad del Trab. 2000; (87):69-76.
 Borroto S, Gámez D, Martínez Y, Ferrer A I, Velásquez Y, Llanes M et al. Latent tuberculosis Infection among health care workers at a general hospital in Santiago de Cuba. Int J Tuberc Lung Dis. 2011; 15(11):1510-14
 Díaz A, Dueñas D, Lazo M, Borroto S, González E. Tuberculosis in health care workers of Psiquistric Hospital at Havana, 1997-2003. Rev Panam Infectol. 2005; 7(3):22-26.
 National Office of Statistics and Information. Annual Statistics of La Lisa. Acceded: 15-06-2015. Available at: http://www.onei.cu/aed2013/23La%20Habana/Municipios/12%20%20La%20Lisa/esp/20080618_tabla_cuadro.htm
 Iñigas L. Territorio y contextos en la salud de la población. Simposio Juan Pérez de la Riva in Memorian. 2007.
 KNCV. Guidelines for conducting tuberculin skin test surveys in high-prevalence countries. Second Edition. IJTLD 2011;15(Suppl 1):S1-S26.
 Deddens J, Petersen M. Approaches for estimating prevalence ratios. Occup Environ Med 2008; 65:501-506.
 World Health Organization. Guidelines on the management of latent tuberculosis infection. WHO 2015. WHO/HTM/TB/2015.01. Available at: http://www.who.int/tb/publications/latent-tuberculosis-infection/en/
 Thompson N J, Glassroth J L, Snider D E, Farer L S. The booster phenomenon in serial tuberculin testing. Am Rev Respir Dis 1979; 119:587–97.
 Farhat M, Greenaway C, Pai M, Menzies D. False-positive tuberculin skin tests: what is the absolute effect of BCG and non-tuberculous mycobacteria? Int J Tuberc Lung Dis. 2006 Nov;10(11):1192-204.
 Borroto S, Urbino A, Martínez A. Prevalence of tuberculin reactors in the staff of "Pedro Kourí" Institute for Tropical Medicine, 1996-2000. Rev Cubana Med Trop. 2006; 58(1):85-89.
 Borroto S, Sevy J, Fumero M, González E, Machado D. Tuberculosis risk assessment in the staff of the National University Pneumologic Hospital of Havana. Rev Cubana Med Trop. 2012; 64(1):55-60.
 Joshi R, Reingold AL, Menzies D, Pai M. Tuberculosis among health-care workers in low and middle-income countries: a systematic review. PLoS Med 2006; 3(12):494.
 Menzies D, Joshi R, Pai M: Risk of tuberculosis infection and disease associated with work in health care settings. Int J Tuberc Lung Dis. 2007; 11(6):593-605.
 Khanna P, Nikolayevskyy V, Warburton F, Dobson F, Drobniewski F. Rate of latent tuberculosis infection detected by occupational health screening of nurses new to a London teaching hospital. Infect Control Hosp Epidemiol 2009; 30(6):581-4.
 Drobniewski F, Balabanova Y, Zakamova E, Nikolayevskyy V, Fedorin I. Rates of Latent Tuberculosis in Health Care Staff in Rusia. PLoS Medicine 2007; 4(2):55.
 Škodric V, Savic B, Jovanovic M, Pesic I, Videnovic J, Zugic V, et al. Occupational Risk of Tuberculosis among health care workers at the Institute for Pulmonary Diseases of Serbia. Int J Tuberc Lung Dis 2000; 4(9):827-31.
 Torres J, Silva R, Sá R, Cardoso MJ, Nienhaus A. Results of five-years systematic screening for latent tuberculosis infections um healthcare workers in Portugal. J Occup Med Toxicology 2010; 5:22.
 Brito G, Iraizoz AM, Alburquerque F. Screening of alcoholism and bioethical analysis of damage provoked in Cuban population. Rev Cubana Med Gen Integr 2011;27(3):355-370.
 Vries G, Ebek MM, Langbregts-van Weezen C. Heath care workers with tuberculosis infected during work. Eur Resp J 2006; 28:1216-21.
 Hosoglu S, Tanrikulu AC, Dagli C, Akalin S. Tuberculosis among Health Care Workers in a short working period. Am J Infect Control 2005; 33:23-6.
 Jiamjarasrangsi W, Hirunsuthikul N, Kamonratanakul R. Tuberculosis among Health Care Workers at King Chulalongkorn Memorial Hospital, 1988-2002. Int J Tuberc Lung Dis 2005; 11:1253-8.
 Craven R, Wenzel R, Atuk N. Minimizing tuberculosis risk to hospital personnel and students exposed to unsuspected disease. Ann Intern Med 1975; 82: 628-32.
 Franchi A, Richeldi L, Parrinello G, Franco R. Room size is the major determinant for tuberculin conversion. Int Arch Occupational Envir Health 2007; 80:533-8.
 He GX, van den Hof S, van der Werf MJ, et al. Infection control and the burden of tuberculosis infection and disease in health care workers in China: a cross-sectional study. BMC Infect Dis 2010; 10:313.
 Brantsæter A B, Romanus V, Andersen P H, Heldal E. Evidence of protective effect of BCG vaccination in persons at low risk of tuberculosis in Nordic countries. Int J Tuberc Lung Dis 2009; 13:440–45.
 Jasmer R M, Nahid P, Hopewell P. Latent tuberculosis infection. N Engl J Med 2002; 347:1860–66.
 Werner E, Sevy J I. Tuberculin Skin Test with 2 PPD-RT23 units in 1588 scholars between 6 to 14 years old no BCG vaccinated. Bol Hig Epidemiol 1968; 6:411–423.
 Wilson R. Tuberculin sensibility in a 441 scholars group of middle education level. [Specialty thesis on pediatric]. Havana: Army Hospital ‘Carlos J. Finlay’, 1976.
 Joos T, Miller W, Murdoch D. Tuberculin reactivity in bacilli Calmette-Guérin vaccinated populations: a compilation of international data. Int J Lung Dis. 2006; 10(8):883-91.
 Fjällbrant H, Rutqvist A, Widström, Zetterberg G, Ridell M, Larson L-O. Tuberculin skin test reactivity of health care students in a country with a low prevalence of tuberculosis. Int J Tuberc Dis. 2010; 14(10):1272-79.