Comparison of QuantiFERON‐TB Gold In‐Tube Test and Tuberculin Skin Test for Identification of Latent Mycobacterium tuberculosis Infection in Healthcare Staff and Association Between Positive Test Results and Known Risk Factors for Infection
Objective. We compared a whole-blood interferon‐γ release assay (QuantiFERON‐TB Gold In‐Tube test, hereafter “QFT–in tube test”) with a tuberculin skin test (TST) to determine which test more accurately identified latent Mycobacterium tuberculosis infection in healthcare staff .
Methods. A total of 481 hospital staff members were recruited from 5 hospitals in Melbourne, Australia. They provided information about demographic variables and tuberculosis (TB) risk factors (ie, birth or travel in a country with a high prevalence of TB, working in an occupation likely to involve contact with M. tuberculosis or individuals with TB, or being a household contact of an individual with a proven case of pulmonary TB). The QFT–in tube test and the TST were administered in accordance with standardized protocols. Concordance between the test results and positive risk factors was analyzed using the κ statistic, the McNemar test, and logistic regression.
Results. A total of 358 participants had both a TST result and a QFT–in tube test result available for comparison. There were fewer positive QFT–in tube test results than positive TST results (6.7% vs. 33.0%; 
). Agreement between the tests was poor (71%; 
). A positive QFT–in tube test result was associated with birth in a country with a high prevalence of TB, the number of years an individual had lived in a country with a high prevalence of TB (ie, the effect of each additional year, treated as a continuous variable), and high‐risk occupational contact. A positive TST result was associated with older age, receipt of bacille Calmette‐Guérin (BCG) vaccination, and working in an occupation that involved patient contact. Receipt of BCG vaccination was most strongly associated with discordant results in instances in which the TST result was positive and the QFT–in tube test result was negative.
Conclusion. In a population of healthcare staff with a low prevalence of TB and a significant rate of BCG vaccination, a positive QFT–in tube test result was associated with the presence of known risk factors for TB exposure, whereas a positive TST result was more strongly associated with a prior history of BCG vaccination.
Received February 11, 2008; accepted October 8, 2008; electronically published February 2, 2009.
The need for enhanced infection control and improved screening and prevention programs for healthcare staff1 has been increased by high rates of latent Mycobacterium tuberculosis infection (LTBI) among healthcare workers (HCWs)2‐4 and the emergence of tuberculosis (TB) strains that are resistant to all first‐line anti‐TB agents (so‐called extensively drug‐resistant, or XDR, TB) and thus disease that has no effective therapy and high mortality rates.5 Some advocate a return to bacille Calmette‐Guérin (BCG) vaccination as a means of protection.6
Current screening methods for HCWs rely on the use of the tuberculin skin test (TST).7 However, the usefulness of the TST is limited by a lack of specificity when used to test people with a history of BCG vaccination, the need for a return visit to interpret the test result, and the subjectivity involved in interpreting results.8,9 These problems are of concern when testing groups with a low prevalence of TB and high rates of BCG vaccination, such as HCWs in high‐income countries, and it is likely that a significant number of positive TST results in this group are the result of prior BCG vaccination.10‐12
New diagnostic tests for LTBI that detect in vitro interferon‐γ production in response to M. tuberculosis–specific antigens eliminate the potential for false‐positive results caused by previous BCG vaccination.13,14 The QuantiFERON‐TB Gold In‐Tube (Cellestis; hereafter “QFT–in tube”) is one such test; it has been shown to be comparable to the TST in terms of its sensitivity in detecting active disease and identifying infection among recent contacts of individuals with TB.14,15 Evidence is still accumulating regarding its suitability and sensitivity for occupational screening.16
In this study, we evaluated the QFT–in tube test by comparing it with the TST as a method for screening healthcare staff for LTBI in metropolitan hospitals in Melbourne, Australia, where the percentage of individuals with positive TST results has previously been shown to be 20%.3 We used the correlation between the 2 test results and the presence or absence of known risk factors for TB exposure, as well as the extent of such exposure, to determine which of the tests more accurately detected latent infection.
Methods
Recruitment
The study was conducted at 5 hospitals in the Southern Health Care Network (Melbourne, Australia) and was approved by the Human Research Ethics Committees of Melbourne Health and the Southern Health Care Network. All hospital staff (physicians, nurses, allied health staff, patient services assistants, and clerical staff) were invited to participate in the study; enrollment was encouraged by use of posters, public address system announcements, and ward visits, with the assistance of each hospital’s infection prevention and control unit. Staff members who were pregnant or who had previously experienced severe reaction or allergy to purified protein derivative were excluded from receiving the TST but were eligible to undergo the QFT–in tube test. Staff who had a previously documented positive TST result and did not wish to have another TST were also eligible to undergo the QFT–in tube test.
Assessment of Previous Exposure and Risk Factors
Information on age, sex, occupation, other illnesses, history of past TST results, country of birth, date of arrival in Australia, history and duration of overseas travel, and known exposure to TB was collected via interviews. BCG vaccination status was ascertained by self‐reported vaccination history and/or the detection of a BCG vaccination scar on visual inspection by the interviewer; where possible, BCG vaccination status was confirmed with previous staff vaccination records. For each participant, the country of birth and the locations of overseas travel were divided into the following 3 groups on the basis of published TB prevalence rates: high‐prevalence countries (⩾100 cases of TB per 100,000 population), intermediate‐prevalence countries (10‐100 cases of TB per 100,000 population), and low‐prevalence countries (<10 cases of TB per 100,000 population).17,18
We also defined 5 groups that were at high risk for exposure to TB. In the absence of a “gold standard” diagnostic test for LTBI, we used the risk of exposure to TB (defined by these risk groups) as a proxy “gold standard” with which to compare the 2 tests. These groups were as follows: (1) those born in a high‐prevalence country, (2) those with a lifetime history of travel to a high‐prevalence country for more than 12 months, (3) those whose occupation included a high likelihood of TB contact (ie, work in a bronchoscopy suite, work as a respiratory or infectious diseases physician, work on a dedicated TB ward, microbiology or pathology work involving M. tuberculosis, or performance of autopsies), (4) those with high‐risk occupational contact (ie, more than 10 hours total contact with a patient known to have TB, without respiratory precautions), and (5) those who were household contacts of an individual with a proven case of pulmonary TB. An individual's total number of risk factors and the duration of time that person had lived in a high‐prevalence country were later used in our analysis as surrogate markers for the extent of exposure.
TST
A “one‐step” tuberculin strategy was chosen because this was the strategy currently in use for TB screening in the Southern Health Care Network. The TST was performed by use of the Mantoux method, in accordance with Australian guidelines,19 which recommend 10 IU of purified protein derivative (Commonwealth Serum Laboratories); the test result was interpreted 48‐72 hours after administration by use of the palpation method. The test was performed by trained nursing staff who were blinded to the results of the QFT–in tube test and the questionnaire. A positive TST result was defined as induration of 10 mm or greater.20
QFT–In Tube Test
The QuantiFERON‐TB Gold assay, also referred to as the “second‐generation” QFT assay or QFT‐2G, uses selected M. tuberculosis antigens or peptide‐simulating antigens, including early secreted antigenic target 6 and culture filtrate protein 10 . A newer version of this assay, known as the QuantiFERON‐TB Gold In‐Tube test or “third‐generation” QFT assay, was used for this study. This test involves the collection of blood samples in tubes prefilled with antigen (typically a negative control tube, an M. tuberculosis–antigen tube, and an optional mitogen tube), which simplifies laboratory procedures.21 The QFT–in tube test was performed and results interpreted in accordance with the manufacturer’s guidelines.22 Kits were purchased from the manufacturer (Cellestis) and were processed at the Cellestis laboratory in Carnegie, Australia.
Statistical Analysis
Statistical analysis was performed using Stata, version 8.0 (Stata). Concordance between the TST results and the QFT–in tube test results was examined by use of the κ statistic; κ values of less than 0.4 indicate poor agreement, values of 0.4‐0.75 indicate good agreement, and values greater than 0.75 indicate excellent agreement.23
We estimated the increase in the likelihood of a positive test result for each unit of exposure to the risk factors defined above and used matched‐pair logistic regression to assess the significance of the difference in the associations between the tests. The McNemar test and multiple logistic regression analysis were used to determine which factors were significantly associated with a positive result. A multinomial logistic model was used to determine the factors associated with discordant test results (using concordant results as the comparator). Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs).
Results
A total of 481 hospital staff members participated in the study, all of whom had a QFT–in tube test performed. The characteristics of the study subjects are shown in Table 1. Of 481 participants, 32 (6.7%) had a positive QFT–in tube test result, 441 (91.7%) had a negative result, and 8 (1.6%) had an indeterminate result.
A total of 364 participants (75.7%) had a TST performed and interpreted during the study. Of these, 120 (33.0%) had a positive test result when a cutoff of 10 mm induration was used, 73 (20.0%) participants showed induration of 15 mm or greater, and 39 (10.7%) showed induration of 20 mm or greater. Four people had a severe reaction to purified protein derivative that involved ulceration or blistering, requiring treatment. One hundred seventeen (24.3%) of the participants did not have a TST result available; 61 (12.7%) had a previously documented positive TST result and declined to repeat the TST, 8 (1.7%) declined the TST on the basis of a previous severe reaction, and 47 (9.8%) failed to have the test result interpreted within 72 hours, despite repeated attempts to contact them (1 additional participant did not have a TST performed because of having recently received a measles‐mumps‐rubella vaccination, which can interfere with TST results).
Of the 364 participants who had TST results available, 6 had an indeterminate QFT–in tube test result; therefore, the final sample size for comparison of the tests was 358 participants. Table 2 shows the test results stratified by BCG vaccination status. The results analyzed below include only participants who had results available for both tests.
Agreement Between the Test Results
There were fewer positive QFT–in tube test results than positive TST results (6.7% versus 33.0%; OR, 19.6 [95% CI, 8.6‐54.]; 
). Five subjects with positive QFT–in tube test results had negative TST results. The agreement between the QFT–in tube test and the TST was generally poor (71% when a cutoff of 10 mm induration was used; 
), but it improved slightly when higher cutoff points were used for the TST (82% for a cutoff of 15 mm induration [
] and 89% for a cutoff of 20 mm induration [
]). Agreement for tests of unvaccinated subjects was 92% when a cutoff of 10 mm induration was used (
) and 97% when a cutoff of 15 mm induration was used (
); agreement for tests of BCG‐vaccinated subjects at the same cutoff points was 66% (
) and 79% (
), respectively. For each 5‐mm increase in induration, an increasing proportion of subjects had a positive QFT–in tube test result (Figure 1).
Figure 1. Percentage of subjects with a positive QuantiFERON‐TB Gold In‐Tube (QFT–in tube) test result, according to size of tuberculin skin test (TST) induration.
Correlation with Risk Factors
Table 3 shows the odds ratio for a positive QFT–in tube test result relative to various exposures. The factors associated with a positive QFT–in tube test result were birth in a high‐prevalence country, the number of years an individual had lived in a high‐prevalence country, and high‐risk occupational contact.
For the TST, receipt of BCG vaccination, an occupation involving patient contact (as opposed to hospital occupations that do not involve such contact, eg, clerical positions), and a greater number of years lived in a high‐prevalence country were associated with a positive test result. Staff who reported having had a previous TST were more likely to have a positive result for the TST administered as part of this study (
) and staff whose occupation involved patient contact were more likely to report a history of previous TST (
).
The results of a multinomial logistic regression for discordant results are shown in Table 4. High‐risk occupational contact and birth in a high‐prevalence country showed the strongest association with discordant test results in instances in which the QFT–in tube test result was positive and the TST result was negative. Receipt of BCG vaccination and having an occupation that involved patient contact were most strongly associated with discordant results in instances in which the TST result was positive and the QFT–in tube test result was negative.
Correlation With Extent of Exposure
The relationship between positive results for each test and an individual's total number of risk factors is shown in Figure 2, and the relationship between positive test results and the number of years participants had lived in a high‐prevalence country is shown in Figure 3. A positive QFT–in tube test result correlated with increased exposure for both the number of years individuals lived in a high‐prevalence country (OR, 1.5 [95% CI, 1.16‐1.92]; 
) and their total number of risk factors (OR, 2.7 [95% CI, 1.49‐4.71]; 
). Increased exposure as measured by duration of travel in a high‐prevalence country or time worked in a high‐risk occupation was not significantly associated with positive results for either test.
Figure 2. Relationship between positive results for the QuantiFERON‐TB Gold In‐Tube test (QFT–in tube) and the tuberculin skin test (TST) and an individual's total number of risk factors. Error bars indicate 95% confidence intervals. OR, odds ratio.
Figure 3. Relationship between positive results for the QuantiFERON‐TB Gold In‐Tube test (QFT–in tube) and the tuberculin skin test (TST) and the number of years participants had lived in a high‐prevalence country. Error bars indicate 95% confidence intervals. OR, odds ratio.
Discussion
To our knowledge, this is the largest study to date that compares the QFT–in tube test and the TST for screening healthcare staff in a developed country, and the first to correlate findings with the extent of exposure to known risk factors. The findings demonstrate that in a population with a low prevalence of TB but a significant rate of BCG vaccination, a positive QFT–in tube test result has a stronger degree of association with the presence of known risk factors for TB exposure than does a positive TST result, whereas a positive TST result has a strong association with a prior history of BCG vaccination (OR, 9.23) and some risk factors for TB. The findings support the hypothesis that BCG vaccination status may significantly affect the proportion of positive TST results and suggest that the QFT–in tube test may be more effective for identifying HCWs with LTBI in countries with a low prevalence of TB by eliminating the possibility of false‐positive results caused by BCG vaccination.
Our results are consistent with those of a Korean study24 that showed the QFT–in tube test to be comparable with the TST in its ability to detect LTBI and showed QFT–in tube test results to be less affected by BCG vaccination status; our results are also consistent with those of a Japanese study25 in which HCWs in a community hospital were assigned a contact score on the basis of their contact time with an index patient with TB and the infectivity of the index patient. In that study, the QFT–in tube test result showed a better association with the contact score than did the TST result.
Our findings are contrary to those of a study conducted in India in which there was good agreement (
) between the QFT–in tube test result and the TST result for HCWs.26 The differences between their findings and ours could be explained by the differing prevalence of exposure to TB in the 2 study populations. If both tests are of comparable sensitivity, the level of agreement between the tests is likely to increase as the TB prevalence (and therefore the number of true‐positive results) in a population increases. Agreement would, therefore, be lower in low‐prevalence populations with a comparable proportion of confounding factors, such as BCG vaccination, that might influence the result of one test (the TST) but not the other (the QFT–in tube test). The high level of agreement between the tests for Indian HCWs also provides indirect evidence that the QFT–in tube test may have a sensitivity similar to that of the TST.
Our findings are also contrary to those of a study of US military personnel that found discordant results to be associated with birth in a high‐prevalence country. The authors of that study concluded that the QFT–in tube test was more specific than the TST when a cutoff of 10 mm induration is used, but also that it may not be as sensitive for the detection of TB as a properly performed TST.27 In our study, not only did a positive QFT–in tube test result show a stronger correlation than a positive TST result with birth in a high‐prevalence country, but this factor was also associated with discordant results in instances in which the TST result was negative. The QFT–in tube test result also showed a stronger correlation with the number of years individuals had lived in a high‐prevalence country when this was used as a marker for the extent of exposure. The odds of a positive TST result appeared to plateau after 20 years spent living in a high‐prevalence country, whereas the odds of a positive QFT–in tube test result continued to rise after that point. This latter finding may be of relevance to the study by Mazurek et al.,27 in which the mean age of participants was only 20 years.
A limitation of the present study was the high rate of loss to follow‐up; 117 (24.3%) of the 481 participants originally recruited did not have a TST performed and/or a TST result interpreted, and a large proportion of these individuals had had positive TST results in the past. Therefore, the possibility of selection bias cannot be excluded. Our study was further limited by the fact that some subcategories contained a only small number of subjects, which may have prevented some relationships from becoming fully apparent in our analysis
A number of different strategies have been suggested for LTBI screening of HCWs, including the use of a 2‐stage testing strategy, in which an initial TST is followed by a QFT–in tube test for those with a positive TST result.28 Our findings suggest that this approach may fail to identify a significant proportion of individuals who may have a negative TST result but would have a positive QFT–in tube test result if tested. Although a strategy that combines the use of both the TST and the QFT–in tube test may be appropriate for patients at high risk for TB reactivation (such as immunosuppressed individuals), implementing this strategy for a low‐risk group such as HCWs would result in increased cost without reducing any of the inherent problems of TST testing.
Our findings support the recent Centers for Disease Control and Prevention guidelines, which suggest that the QFT–in tube test is a viable alternative for a widespread HCW screening program.13,29 If this approach were used, it is likely that far fewer staff would require follow‐up and treatment, which has significant implications for reducing the costs of a screening program, as well as for reducing the morbidity that results from adverse reactions to isoniazid therapy.
Acknowledgments
We thank the participating healthcare workers; Drs. Alan Street, Tony Korman and Jim Black for discussions about the study; the Infection Control Unit at Southern Health and Mr. Francis Tyler for assistance with recruitment and performance of tuberculin skin tests; and Christalla Hajisava for assistance with the manuscript.
Financial support: Department of Human Services Victoria (Public Health Research Projects grant to B.‐A.B.); National Health and Medical Research Council (scholarship to P.V.); Edgar Tattnall Memorial Trust (grant to B.‐A.B.).
Potential conflicts of interest: Cellestis is a listed Australian Security and supplied the QuantiFERON‐TB Gold In‐Tube test kits for the study. Cellestis personnel did not participate in study design, data analysis, or manuscript preparation. None of the authors are holders of Cellestis stock. G.A.J. and D.J. have received study support from Cellestis.
References
- 1. Jereb JA, Klevens RM, Privett TD, et al. Tuberculosis in health care workers at a hospital with an outbreak of multidrug‐resistant Mycobacterium tuberculosis. Arch Intern Med 1995; 155:854‐859.
- 2. 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; 11:593‐605.
- 3. Stuart RL, Bennett NJ, Forbes AB, Grayson ML. Assessing the risk of tuberculosis infection among healthcare workers: the Melbourne Mantoux Study. Melbourne Mantoux Study Group. Med J Aust 2001; 174:569‐573.
- 4. Stuart RL, Grayson ML. Health care workers and tuberculosis. Aust N Z J Med 2000; 30:367‐372.
- 5. Gandhi NR, Moll A, Sturm AW, et al. Extensively drug‐resistant tuberculosis as a cause of death in patients co‐infected with tuberculosis and HIV in a rural area of South Africa. Lancet 2006; 368:1575‐1580.
- 6. Graham M, Howley TM, Pierce RJ, Johnson PD. Should medical students be routinely offered BCG vaccination? Med J Aust 2006; 185:324‐326.
- 7. Moran‐Mendoza O, Marion SA, Elwood K, Patrick DM, FitzGerald JM. Tuberculin skin test size and risk of tuberculosis development: a large population‐based study in contacts. Int J Tuberc Lung Dis 2007; 11:1014‐1020.
- 8. Tissot F, Zanetti G, Francioli P, Zellweger JP, Zysset F. Influence of bacille Calmette‐Guerin vaccination on size of tuberculin skin test reaction: to what size? Clin Infect Dis 2005; 40:211‐217.
- 9. von Reyn CF, Horsburgh CR, Olivier KN, et al. Skin test reactions to Mycobacterium tuberculosis purified protein derivative and Mycobacterium avium sensitin among health care workers and medical students in the United States. Int J Tuberc Lung Dis 2001; 5:1122‐1128.
- 10. Panlilio AL, Burwen DR, Curtis AB, et al. Tuberculin skin testing surveillance of health care personnel. Clin Infect Dis 2002; 35:219‐227.
- 11. Marsh BJ, San Vicente J, von Reyn CF. Utility of dual skin tests to evaluate tuberculin skin test reactions of 10 to 14 mm in healthcare workers. Infect Control Hosp Epidemiol 2003; 24:821‐824.
- 12. Johnson PD, Stuart RL, Grayson ML, 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; 6:934‐937.
- 13. Mazurek GH, Jereb J, Lobue P, Iademarco MF, Metchock B, Vernon A. Guidelines for using the QuantiFERON‐TB Gold test for detecting Mycobacterium tuberculosis infection, United States. MMWR Recomm Rep 2005; 54(RR‐15):49‐55.
- 14. Pai M, Riley LW, Colford JM Jr. Interferon‐gamma assays in the immunodiagnosis of tuberculosis: a systematic review. Lancet Infect Dis 2004; 4:761‐776.
- 15. Nakaoka H, Lawson L, Squire SB, et al. Risk for tuberculosis among children. Emerg Infect Dis 2006; 12:1383‐1388.
- 16. Villarino ME, Mazurek G. Tuberculosis contacts, concerns, and controls: what matters for healthcare workers? Infect Control Hosp Epidemiol 2006; 27:433‐435.
- 17. World Health Organization. Global Tuberculosis Database 2006. Available at: http://www.who.int/globalatlas/dataQuery/default.asp. Accessed November 12, 2007.
- 18. Broekmans JF, Migliori GB, Rieder HL, et al. European framework for tuberculosis control and elimination in countries with a low incidence: recommendations of the World Health Organization (WHO), International Union against Tuberculosis and Lung Disease (IUATLD) and Royal Netherlands Tuberculosis Association (KNCV) Working Group. Eur Respir J 2002; 19:765‐775.
- 19. The Australian immunisation handbook. 9th ed. Australia: National Health Medical Research Council; 2008:231‐234. Available at: http://www.immunise.health.gov.au/internet/immunise/publishing.nsf/Content/Handbook-home. Accessed January 14, 2009.
- 20. Department of Human Services. Management, control and prevention of tuberculosis: guidelines for health care providers (2002‐2005). Available at:http://www.health.vic.gov.au/ideas/diseases/tb_mgmt_guide. Accessed January 14, 2009.
- 21. Farris A, Branda J. QuantiFERON‐TB Gold assay for tuberculosis infection. Clin Microbiol Newsl 2007;129‐136.
- 22. Cellestis. QuantiFERON‐TB Gold IT Method and Technical Instructions. Melbourne, Australia: Cellestis; 2007. Available at: http://www.cellestis.com/IRM/Content/aust/qtfproducts_tbgoldintube_techinfo.html. Accessed December 12, 2007.
- 23. Ewer K, Deeks J, Alvarez L, et al. Comparison of T‐cell‐based assay with tuberculin skin test for diagnosis of Mycobacterium tuberculosis infection in a school tuberculosis outbreak. Lancet 2003; 361:1168‐1173.
- 24. Kang YA, Lee HW, Yoon HI, et al. Discrepancy between the tuberculin skin test and the whole‐blood interferon gamma assay for the diagnosis of latent tuberculosis infection in an intermediate tuberculosis‐burden country. JAMA 2005; 293:2756‐2761.
- 25. Kobashi Y, Obase Y, Fukuda M, et al. Usefulness of QuantiFERON TB‐2G, a diagnostic method for latent tuberculosis infection, in a contact investigation of health care workers. Intern Med (Tokyo, Japan). 2007; 46:1543‐1549.
- 26. Pai M, Gokhale K, Joshi R, et al. Mycobacterium tuberculosis infection in health care workers in rural India: comparison of a whole‐blood interferon gamma assay with tuberculin skin testing. JAMA 2005; 293:2746‐2755.
- 27. Mazurek GH, Zajdowicz MJ, Hankinson AL, et al. Detection of Mycobacterium tuberculosis infection in United States Navy recruits using the tuberculin skin test or whole‐blood interferon‐γ release assays. Clin Infect Dis 2007; 45:826‐836.
- 28. Diel R, Nienhaus A, Lange C, Schaberg T. Cost‐optimisation of screening for latent tuberculosis in close contacts. Eur Respir J 2006; 28:35‐44.
- 29. Jensen PA, Lambert LA, Iademarco MF, Ridzon R. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health‐care settings, 2005. MMWR Recomm Rep 2005; 54(RR‐17):1‐141.