ORIGINAL STUDIES

Interferon-␥ Release Assay for the Diagnosis of Latent

Tuberculosis in Children Younger Than 5 Years of Age
Ivan Pavić, MD,* Renata Zrinski Topić, PhD,† Miljenko Raos, MD,‡ Neda Aberle, PhD,§
and Slavica Dodig, PhD†

investigations. IGRAs are based on the detection of IFN-␥ released

Background: There are limited data available on interferon-␥ release
assay (IGRA) performance in children up to 5 years of age, with docu-
mented exposure to active tuberculosis (TB). The aim of this study was to
evaluate (1) the influence of infectivity of adult source cases on test results,
(2) the impact of age, and (3) the level of agreement, between IGRA and
tuberculin skin test (TST) results.
Methods: A total of 142 Bacille Calmette-Guerin-vaccinated children up
to 5 years of age were investigated because of a history of exposure to
active TB. QuantiFERON-TB Gold In-Tube IGRA (QFT) and TST assays
were performed.
Results: Test results were significantly influenced by positive finding of
cavitary lesions (QFT, odds ratio 关OR兴 ⫽ 6.15; TST, OR ⫽ 7.48) and
positive acid-fast bacilli (QFT, OR ⫽ 4.01; TST, OR ⫽ 4.47) in active TB
contacts. QFT resulted in 1 indeterminate response (0.7%), attributable to
low mitogen. There was no evidence for age having any effect on QFT
performance. The 2 tests showed a moderate overall concordance (89%;
␬ ⫽ 0.591) at a TST cutoff value of ⱖ10 mm.
Conclusions: Association of positive QFT and TST results with risk
factors for infection in child contacts (presence of cavitary lesions and
acid-fast bacilli smear positivity in index cases) suggests that both the tests
have good diagnostic accuracy. However, there was significant discord
between results of the 2 tests that could not be definitively resolved. Thus,
in a high-risk population of children up to 5 years of age, both tests (QFT
and TST) should be performed and the child should be considered infected
if either or both tests are positive.
Key Words: children, interferon-␥ release assay, latent tuberculosis
infection, tuberculin skin test
(Pediatr Infect Dis J 2011;30: 866 – 870)
C hildren, who have been in close contact with an adult with
smear-positive tuberculosis (TB), have a high risk of infection
with Mycobacterium tuberculosis. Although most infected chil-
dren initially have latent TB infection (LTBI; a subclinical infec-
tion with M. tuberculosis without clinical, radiologic, or bacterio-
logic evidence of TB), in children up to 5 years of age, the risk of
developing active TB in the 2 years post infection is 20% to 40%.
The risk then decreases as age increases.1
Development of ex vivo T-cell-based interferon (IFN)-␥
release assays (IGRAs) offered significant progress for TB contact
Accepted for publication April 20, 2011.
From the *Children’s Hospital Zagreb, University Hospital “Sestre milosrd-
nice,” Zagreb, Croatia; †Department of Clinical Laboratory Diagnosis,
Srebrnjak Children’s Hospital, Zagreb, Croatia; ‡Pediatric Department,
Srebrnjak Children’s Hospital, Zagreb, Croatia; and §Pediatric Department,
General Hospital “Dr. Josip Benčević,” Slavonski Brod, Croatia.
The authors have no funding or conflicts of interest to disclose.
Address for Correspondence: Ivan Pavić, MD, Children’s Hospital Zagreb,
University Hospital “Sestre milosrdnice,” Klaićeva 16, 10000 Zagreb,
Croatia. E-mail: ivanpavic@net.hr.
Copyright © 2011 by Lippincott Williams & Wilkins
ISSN: 0891-3668/11/3010-0866
DOI: 10.1097/INF.0b013e318220c52a
from T lymphocytes upon stimulation with M. tuberculosis-spe-
cific antigens which are absent from Bacille Calmette-Guerin
(BCG) and most nontuberculosis mycobacteria.2 Although the
tuberculin skin test (TST) has been useful in clinical practice, it has
known limitations such as variable specificity, reproducibility, and
cross-reactivity with nontuberculosis mycobacteria and BCG vac-
cination.3 A major benefit of the IGRAs is their high specificity
even in those who have received BCG vaccination.4 However,
little is known about the performance of IGRAs in children less
than 5 years of age.5–7
The aim of the present study was to evaluate an IGRA for
diagnosis of LTBI in BCG-vaccinated children up to 5 years of
age, with documented exposure to active TB, in Croatia, a middle-
income country of central Europe, where the average annual
incidence rate of TB is 25/100,000 (13.8/100,000 in children up to
14 years of age).8 The objectives were to assess (1) the influence
of infectivity of the adult source case (duration of contact, cavitary
lesions, acid-fast bacilli 关AFB兴 sputum smear, culture status) on
test results; (2) the impact of age on results of IGRA; and (3) the
level of agreement between IGRA and TST results.
METHODS
Study Participants
We investigated 142 consecutive children 1 month to 5
years of age between January 2008 and December 2009 at Chil-
dren’s Hospital in Zagreb and General Hospital in Slavonski Brod,
Croatia, because of a history of exposure to a case of active TB.
All study participants had received compulsory BCG vaccination
at birth (BCG vaccine SSI-Danish strain 1331), according to the
national vaccination calendar, which was verified at the examina-
tion by written record and the presence of the scar. All children
were tested at least 12 weeks after exposure.
Inclusion criteria were pediatric patients ⱕ5 years of age and
a documented exposure to a case of active TB. Close contact was
defined as household contact with aggregate exposure to a patient
with active TB of not less than 40 hours in closed rooms, and distant
contact was defined as occasional or unclear exposure time of less
than 40 hours during the presumed period of infectiousness.9 Exclu-
sion criteria were children ⬎5 years, immunocompromised children,
inadequate blood sampling, and diagnosis of active TB.
Diagnostic procedures were performed in accordance with
the Declaration on Human Rights from Helsinki 1975 and Seoul
amendments 2008, and study approval was obtained from the
Hospital Ethics Committee.
IFN-␥ Assay
The commercial QuantiFERON-TB Gold In-Tube IGRA
(QFT; Cellestis Ltd., Chadstone, Australia) was performed accord-
ing to the manufacturer’s instructions. Blood samples for QFT
were drawn under standardized condition in our hospital at the
same day as TST. The cutoff value for positive IFN-␥ response
(TB antigen minus nil) was ⱖ0.35 IU/mL as recommended by the
manufacturer, and the test was considered indeterminate if the

866 | www.pidj.com The Pediatric Infectious Disease Journal • Volume 30, Number 10, October 2011
The Pediatric Infectious Disease Journal • Volume 30, Number 10, October 2011 Diagnosis of Latent Tuberculosis

value of the positive-control well (mitogen minus nil) was less
than 0.5 IU/mL, and/or nil negative control was more than 8 IU/L.
The TST was performed with 2 tuberculin units of standard-
ized purified protein derivative solution (Tuberculin PPD RT 23,
Statens Serum Institute, Copenhagen, Denmark) injected into the
volar aspect of the forearm and transverse induration was mea-
sured by trained healthcare workers 68 to 72 hours later.10 As per
Croatian guidelines for BCG-vaccinated children with a docu-
mented exposure to a case of active TB, induration ⱖ10 mm was
considered positive.10
Statistical Analysis
Categorical variables were presented as number and per-
centages. Variables with normal distribution were described by
arithmetic mean (x៮ ) and standard deviation, and those not showing
normal distribution were presented by median (M) and interquar-
tile range. Student t test and Mann-Whitney U test were used for
comparison of dependent variables (for normal distribution and
asymmetric distribution, respectively). Concordance between
IGRA and TST was tested using interrater agreement. Agreement
was quantified by kappa (␬) coefficient of agreement with ␬ values
from 0.81 to 1.00 indicating very good concordance; 0.61 to 0.80,
good concordance; 0.41 to 0.60, moderate concordance; 0.21 to
0.40, fair concordance; and ⬍0.20, poor concordance. Potential
predictors, ie, intimate contact/exposure time with the adult patient
with active TB, positive findings in adult source case of cavitary
lesions, and AFB sputum smear and culture status were identified
through multivariate linear regression models. Relations were
expressed as odds ratios and 95% confidence intervals. Correlation
of the study variables was expressed by coefficient of correlation
(variables with normal distribution) or Spearman coefficient of rank
correlation (asymmetric distribution). P values ⬍0.05 were consid-
ered statistically significant.11 Data processing was performed using
MedCalc software (Medisoftware, Mariakerke, Belgium).
RESULTS
During a 2-year period, 142 children, 85 males and 57
females, average age 29 ⫾ 16 (x៮ ⫾ standard deviation) months,
were studied. In total, 61% (87/142) of children had a known close
household contact with a TB patient. A total of 24 children (17%)
had a positive TST, and 18 (13%) children had a positive IGRA.
Of the 142 children, 1 (0.7%), a 15-month-old boy who was
diagnosed with pneumonia at the time of sampling had an inde-
terminate QFT result (TST was negative) due to a mitogen re-
sponse below the allowable cutoff. Laboratory evaluation in that
child revealed C-reactive protein to be 0.3 mg/L and leukocyte
count 16.31 ⫻ 109/L; with 41% neutrophils, 45% lymphocytes,
5% monocytes, and 9% eosinophils. The child’s chest radiograph
showed an infiltrate in the right lower lobe. Further standardized
diagnostic workup excluded the diagnosis of TB. Data from this
child were not included in further statistical analyses.
Multiple linear regression analysis revealed significant as-
sociations between positive QFT and TST results in children and
infectivity indicators (ie, cavitary lesions, AFB sputum smear) of
the adult source cases (Table 1). Positive TST and QFT results,
respectively, were 7.48 times (P ⬍ 0.0001) and 6.15 times (P ⬍
0.0001) more likely if the source case had cavitary lesions. Among
contacts of AFB smear-positive source cases, the likelihood of
positive result increased by a factor of 4.47 (P ⫽ 0.0001) for TST,
and 4.01 (P ⫽ 0.0005) for QFT. Close contact was not a significant
predictive factor for both TST (P ⫽ 0.090) and QFT (P ⫽ 0.171),
as was not the culture status of the index cases, P ⫽ 0.321 and
P ⫽ 0.404, respectively.
QFT-positive/TST-positive concordance was found in 14/
141 (10%) of children, and QFT-negative/TST-negative concor-
dance was found in 112/141 (79%) of children. There were 15/141
(11%) children with discordant QFT/TST results, 4 of them QFT
positive only, and 11 TST positive only. These 2 tests showed an
overall concordance of 89% (␬ ⫽ 0.591; 95% confidence interval,
0.475– 0.686). Chest radiograph examination was performed on all
children with positive TST or QFT, and it was normal in all of
them. Only those 18 children who were QFT positive were given
isoniazid. All study participants were followed up till the time of
preparation of this manuscript; none of them have been reported as
developing TB.

TABLE 1. Influence of Infectiousness Indicator (Close Contact, Cavitary Lesions,

Acid-fast Bacilli Sputum Smear, and Culture Status) of Adult Source Case on TST
and QFT Results for the 141 Children Contacts

Infectiousness Indicator Test

Positive/Negative Positive/Negative OR 95% CI P
N N, %

TST, N ⫽ 24/117
Close contact 23/1, 26/2 1.75 0.92–3.35 0.090
87/54
Cavity 23/1, 60/1 7.48 3.41–16.39 ⬍0.0001
38/103
AFB 22/2, 48/2 4.47 2.16 –9.24 0.0001
46/95
Culture status 24/0, 22/0 1.38 0.73–2.59 0.321
109/32
QFT, N ⫽ 18/123
Close contact 17/1, 20/2 1.66 0.80 –3.43 0.171
87/54
Cavity 18/0, 47/0 6.15 2,75–13.77 ⬍0.0001
38/103
AFB 17/1, 37/1 4.01 1.84 – 8.71 0.0005
46/95
Culture status 18/0, 17/0 1.35 0.67–2.74 0.404
109/32
Relations are expressed as number (N) and percent (%) of positive/negative results, odds ratios (OR), and 95% confidence
intervals (95% CI).
TST indicates tuberculin skin test; QFT, QuantiFERON-TB Gold In-Tube interferon-␥ release assay; AFB, acid-fast bacilli.

© 2011 Lippincott Williams & Wilkins www.pidj.com | 867

Pavić et al The Pediatric Infectious Disease Journal • Volume 30, Number 10, October 2011

TABLE 2. Characteristics of All Children (N ⫽ 15) With Discordant QFT/TST Results

Patient Age (mo) Gender Contact TST (mm) IFN-␥ (IU/mL), First IFN-␥* (IU/mL), Second Mitogen (IU/mL), First
†
1 7 M Close 0 0.36 0.36 1.00
2 15 M Distant 2 0.56 — 7.10
3† 50 M Close 4 10.77 5.46 18.28
4† 53 M Close 4 1.72 14.10 27.06
5† 25 F Close ⬍10 5.65 6.06 16.69
6‡ 23 M Close 14 0.07 0.01 10.10
7‡ 6 M Close 11 0.06 0.01 10.32
8‡ 34 M Close 24 0.32 0.07 2.02
9 6 M Close 15 0.02 — 2.64
10 52 M Close 20 0.17 — 6.08
11 19 F Close 12 0.02 — 3.17
12 21 F Distant 20 0.00 — 8.80
13‡ 17 M Close 13 0.02 0.02 24.10
14 57 M Close 11 0.01 — 8.98
15‡ 23 M Close 18 0.07 0.08 17.66
*This value was obtained 4 months after the first IFN-␥ value.
†
Children whose QFT result remained positive.
‡
Children whose QFT result remained negative.
QFT indicates QuantiFERON-TB Gold In-Tube IFN-␥ release assay; TST, tuberculin skin test; IFN-␥, interferon-␥.

FIGURE 1. A, Responses to QFT TB-specific antigens versus age for the 18 child contacts who were test positive. B, Re-
sponses to mitogen-positive control versus age for all children (n ⫽ 141). For both plots the dotted line represents the lin-
ear regression line of best fit, with no significant associations with age.

Table 2 shows characteristics of the 15 children with dis-
cordant QFT/TST results at initial measurement (first). Of them, 2
were distant contacts and 13 were close contacts with an active TB
case. The response to mitogen in all children was normal. At 4
months after the primary determination, QFT was again performed
in 9 of the 15 (60%) children (second); all repeat results were
concordant with the initial results with QFT remaining positive in
4 children (patients labeled with †), and negative in 5 (patients
labeled with ‡).
To investigate the relationship between age and ability to
respond to the TB-specific antigens used in QFT, data from
children showing a positive result were plotted versus age (Fig.
1A). Linear regression showed no effect of age on the magnitude
of response (r ⫽ 0.131, P ⫽ 0.605). Mitogen-positive control
values (IU/mL) for IFN-␥ were plotted against age for all children
(Fig. 1B). Similarly, linear regression analysis demonstrated no
significant association with age and magnitude of response (r ⫽
0.0687, P ⫽ 0.418).
DISCUSSION
The results of the study demonstrate that the presence of
cavitary lesions and AFB smear positivity status in adult source
cases were significantly associated with both TST and QFT results.
There was no evidence for age having any impact on QFT
performance in immunocompetent children. A moderate concor-
dance between TST and QFT was documented.
A major limitation in evaluating tests for the detection of
LTBI is the lack of a gold standard for LTBI. Arguably, the only
proof for LTBI is the later development of active TB. But this can
only be ascertained through longitudinal monitoring of subjects at
risk, which is unethical in most settings and probably especially so
in very young children. Therefore, to estimate the accuracy of a
test for LTBI, a pragmatic approach is to associate test results with
known risk factors for infection. Such factors include determina-
tion of the infectivity of the index case and the proximity of
exposure for contacts.
Little is known about the relationship between infection risk
factors and IGRA results in young children. Prolonged close
contact of children with an adult case of active TB is one of the
highest risk factors for young children to become infected with M.
tuberculosis. On the basis of the results of this study, cavitary
lesions and AFB smear positivity in adult source case were
identified as the most important risk factors for positive QFT
results (as an indicator of LTBI) and positive TST results in
children aged up to 5 years.
Our results are consistent with earlier reports such as those of
Diel et al12 who found strong associations with positive QFT results
and measures of infection risk, foreign origin, coughing of the source

868 | www.pidj.com © 2011 Lippincott Williams & Wilkins

The Pediatric Infectious Disease Journal • Volume 30, Number 10, October 2011
case, AFB smear positivity, and exposure time. Our results are also in
accord with those of Lewinsohn et al13 who reported that higher
IFN-␥ responses were associated with cavitary disease, sputum AFB
grade, and extent of disease on chest x-ray in the index case.
In the present study, no impact of age on QFT results was
observed. This is in contrast to the results of Kampmann et al14
who found that IGRA responses were lower in children aged ⬍5
years than in children aged 5 to 15 years. Similarly, Diel et al12
reported that the proportion of positive QFT results increases with
age, but this may simply be due to the risk of being infected
increasing with age. Other authors have proposed that very young
children produce, on an average, less IFN-␥ than older children,
but to our knowledge, this is only conjecture.5,15 Our study found
no evidence of impaired performance of QFT with younger age,
which could, in part, be explained by the healthy cohort of children
we studied.
Indeterminate results could be one of the possible problems
in interpreting the results of IGRAs. A study of IGRAs in a diverse
clinical population reported that indeterminate results with the
QuantiFERON Gold test were disproportionately high in children.6
The causes for indeterminate results could be manifold such as
errors prior to analysis or clinical reasons, eg, young age, low
T-cell count,16 immunosuppression,17 or high background re-
sponse due to recent viral infection.18 Despite the very young age
of our cohort, we identified only 1 indeterminate response from the
142 children tested (0.7%), a similar level to that reported by an
Italian group.19 This 1 indeterminate result was attributable to a
low response to mitogen, probably due to acute infection (pneu-
monia), and can be considered to be a true indeterminate. How-
ever, others have found higher indeterminate rates in children
younger than 3 to 4 years, than in older children,15,20 and an
inverse correlation with indeterminate test results and age.17 It
should be emphasized that different types of IGRAs and different
generations of IGRAs may have different rates of indeterminate
results.20 We expect that the reasons for low number of indeter-
minate results in our study was good control of preanalytical and
analytical procedures (ie, good clinical practice and good clinical
laboratory practice) as well as testing an apparently healthy cohort
of young children.
In the present study, a moderate concordance between TST
and IGRA was confirmed (89%, ␬ ⫽ 0.591), which was consistent
with results of other studies performed in children.15,19,21 A
slightly better overall concordance of results was reported for
children younger than 5 years (88%, ␬ ⫽ 0.626), especially in the
subgroup of children older than 2 years (95%, ␬ ⫽ 0.828).7 A
major factor associated with TST/IGRA concordance in these
studies was BCG vaccination, which was thought to be related to
TST-positive discordant results.
Discordance between IGRA and TST indicates that in some
children, one of these tests gave false-positive or false-negative
results. False-positive TST may indicate a lack of specificity for
diagnosing LTBI. In this study, BCG vaccination could be a major
reason for QFT-negative/TST-positive discordance; however, this
was difficult to ascertain as all children had received the vaccine at
birth and positive TST responses were closely associated with risk
factors for infection. The finding that 5 children who were TST-
positive, but QFT-negative, remained QFT-negative when tested 4
months later is also supportive of at least some of the TST results
being falsely positive. False-negative TST may indicate low skin
reactivity in small children,22 impaired T-cell function,23 or results
erroneously read as negative. An advantage of IGRA, compared
with TST, is the ability to control the quality of work by using
mitogen-positive control.
© 2011 Lippincott Williams & Wilkins
Diagnosis of Latent Tuberculosis
The fact that all negative QFT results remained negative,
and all positive results remained positive 4 months after the initial
measurement, confirmed excellent reproducibility of the QFT,
albeit with reproducibility studies in only 9 children.
There are a number of limitations for our study, including
the lack of exact diameter measurements of induration for the TST
for all children. Responses less than 10 mm were simply recorded
as negative, thus removing our ability to investigate test concor-
dance at lower cutoffs. Reliable information on the exact duration
of exposure and proximity to source cases was lacking, limiting
our ability to investigate these risk factors closely.
The risk of developing active TB following acute infection
in small children is high. Therefore, the early identification of
children infected with the M. tuberculosis is the main factor in
preventing active TB. Children with IGRA-positive/TST-positive
concordance are highly likely to have LTBI, and they should be
treated appropriately. Children with IGRA-positive/TST-negative
discordance are highly likely to be truly infected with M. tuber-
culosis, due to the high specificity of QFT, and should also be
treated appropriately. In IGRA-negative/TST-positive discor-
dance, there is a likelihood that the positive TST result is due to
lower specificity of the test in BCG vaccinated children. Serial use
of IGRAs may be useful to provide a reliable diagnosis that will
help decide if the positive TST is due to infection or a false-
positive reaction. However, such an approach requires further
study and validation.
Our study provides evidence that the QFT test performs
well in children less than 5 years of age and does not appear to be
adversely affected in very young children. Without a gold standard
for LTBI, we cannot determine if one test is more accurate than the
other. Our results suggest that QFT and TST have commensurate
performance in young children. However, given the level of
discordance between QFT and TST identified in this study and the
association of results from both tests with risk factors for infection,
we can only conclude that in a high-risk population of children
aged up to 5 years, both tests, QFT and TST should be performed
and the child should be considered infected if either or both tests
are positive.
REFERENCES
1. Marais BJ, Pai M. Recent advances in the diagnosis of childhood tubercu-
losis. Arch Dis Child. 2007;92:446 – 452.
2. Mazurek M, Jereb J, Vernon A, et al. Updated Guidelines for using
interferon gamma release assays to detect Mycobacterium tuberculosis
infection—United States, 2010. MMWR Recomm Rep. 2010;59:1–25.
3. Menzies RI. Tuberculin skin testing. In: Reichman LB, Hershfield ES, eds.
Tuberculosis: A Comprehensive International Approach. 2nd ed. New
York, NY: Marcel Dekker; 2000:279 –322.
4. Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for
the diagnosis of latent tuberculosis infection—an update. Ann Intern Med.
2008;149:177–184.
5. Dogra S, Narang P, Mendiratta DK, et al. Comparison of a whole blood
interferon-gamma assay with tuberculin skin testing for the detection of
tuberculosis infection in hospitalized children in rural India. J Infect.
2007;54:267–276.
6. Lighter J, Rigaud M, Eduardo R, et al. Latent tuberculosis diagnosis in
children by using QuantiFERON-TB Gold In-tube test. Pediatrics. 2009;
123:30 –37.
7. Okada K, Mao TE, Mori T, et al. Performance of an interferon-gamma
release assay for diagnosing latent tuberculosis infection in children. Epi-
demiol Infect. 2008;136:1179 –1187.
8. EuroTB; the National Coordinators for Tuberculosis Surveillance in the
WHO European Region. Surveillance of tuberculosis in Europe. Report on
tuberculosis cases notified in 2005. Saint-Maurice, France: Institut de veille
sanitaire; March 2007. Available at: http://www.euroTB.org. Accessed
December 7, 2010.
www.pidj.com | 869
Pavić et al The Pediatric Infectious Disease Journal • Volume 30, Number 10, October 2011
9. Behr A, Hopewell PC, Paz EA, et al. Predictive value of contact investi-
gation for identifying recent transmission of Mycobacterium tuberculosis.
Am J Respir Crit Care Med. 1998;158:465– 469.
10. Mardešić D. Tuberculosis. In: Mardešić D, ed. Paediatrics. 6th ed. Zagreb,
Croatia: Školska knjiga; 2000:477– 494.
11. Marusteri M, Bacarea V. Comparing groups for statistical differences: how
to choose the right statistical test? Biochem Med. 2010;20:15–32.
12. Diel R, Loddenkemper R, Meywald-Walter K, et al. Comparative perfor-
mance of tuberculin skin test, Quantiferon-TB-Gold in Tube assay, and
T-Spot. TB test in contact investigations for tuberculosis. Chest. 2009;135:
1010 –1018.
13. Lewinsohn DA, Zalwango S, Stein CM, et al. Whole blood interferon-
gamma responses to Mycobacterium tuberculosis antigens in young house-
hold contacts of persons with tuberculosis in Uganda. PLoS One. 2008;3:
e3407.
14. Kampmann B, Tena-Coki G, Anderson S. Blood tests for diagnosis of
tuberculosis. Lancet. 2006;368:282.
15. Connell TG, Curtis N, Ranganathan SC, et al. Performance of a whole
blood interferon gamma assay for detecting latent infection with Mycobac-
terium tuberculosis in children. Thorax. 2006;61:616 – 620.
16. Grare M, Derelle J, Dailloux M, et al. QuantiFERON(R)-TB Gold In-Tube
as help for the diagnosis of tuberculosis in a French pediatric hospital.
Diagn Microbiol Infect Dis. 2010;66:366 –372.
870 | www.pidj.com
17. Haustein T, Ridout DA, Hartley JC, et al. The likelihood of an indeter-
minate test result from a whole-blood interferon-gamma release assay
for the diagnosis of Mycobacterium tuberculosis infection in children
correlates with age and immune status. Pediatr Infect Dis J. 2009;28:
669 – 673.
18. Stephan C, Wolf T, Goetsch U, et al. Comparing QuantiFERON-tubercu-
losis gold, T-SPOT tuberculosis and tuberculin skin test in HIV-infected
individuals from a low prevalence tuberculosis country. AIDS. 2008;22:
2471–2479.
19. Bianchi L, Galli L, Moriondo M, et al. Interferon-gamma release assay
improves the diagnosis of tuberculosis in children. Pediatr Infect Dis J.
2009;28:510 –514.
20. Bergamini BM, Losi M, Vaienti F, et al. Performance of commercial blood
tests for the diagnosis of latent tuberculosis infection in children and
adolescents. Pediatrics. 2009;123:419 – 424.
21. Kampmann B, Whittaker E, Williams A, et al. Interferon-␥ release assays
do not identify more children with active tuberculosis than the tuberculin
skin test. Eur Respir J. 2009;33:1374 –1382.
22. Raos M, Marković J, Dodig S. Epidemiological and clinical characteristics
of tuberculosis in children up to seven years of age treated in Srebrnjak
Children’ s Hospital. Paediatr Croat. 2008;52:93–97.
23. Soborg B, Ruhwald M, Hetland ML, et al. Comparison of screening
procedures for Mycobacterium tuberculosis infection among patients with
inflammatory diseases. J Rheumatol. 2009;36:1876 –1884.
© 2011 Lippincott Williams & Wilkins