Helicobacter ISSN 1523-5378

13
C-Urea Breath Test for the Diagnosis of Helicobacter pylori
Infection in Children: A Systematic Review and Meta-Analysis
Yelda A. Leal,* Laura L. Flores, Ezequiel M. Fuentes-Panana, Roberto Cedillo-Rivera* and Javier Torres
*
Unidad de Investigacion Medica Yucatan (UIMY), Unidad Medica de Alta Especialidad de Merida, Instituto Mexicano del Seguro Social, Merida,
Yuc, Mexico, Division of Pulmonary and Critical Care Medicine, San Francisco General Hospital and the University of California, San Francisco, CA,
USA, Unidad de Investigacion Medica en Enfermedades Infecciosas y Parasitarias (UIMEIP), Hospital de Pediatra Centro Medico Nacional Siglo
XXI, Instituto Mexicano del Seguro Social, Mexico, DF, Mexico

Keywords
13C-urea breath test, Helicobacter pylori,
diagnosis, systematic review, meta-analysis.
Reprint requests to: Yelda A. Leal, Unidad de
Investigacion Medica Yucatan, Unidad Medica
de Alta Especialidad del Centro Medico
Nacional Ignacio Garca Tellez Merida
Yucatan, Instituto Mexicano del Seguro
Social, Calle 34 #439 x 41 Colonia Industrial
(Ex-terrenos del Fenix, Hospital T1), Merida,
Yucatan 97150, Mexico.
E-mail: yelda_leal03@yahoo.com.mx
Abstract
Background: The 13C-urea breath test (13C-UBT) is a safe, noninvasive and
reliable method for diagnosing H. pylori infection in adults. However, the test
has shown variable accuracy in the pediatric population, especially in young
children. We aimed to carry out a systematic review and meta-analysis to
evaluate the performance of the 13C-UBT diagnostic test for H. pylori
infection in children.
Methods: We conducted a systematic review of the PubMed, Embase and
Liliacs databases including studies from January 1998 to May 2009. Selec-
tion criteria included studies with at least 30 children and reporting the
comparison of 13C-UBT against a gold standard for H. pylori diagnosis.
Thirty-one articles and 135 studies were included for analysis. Children were
stratified in subgroups of <6 and 6 years of age, and we considered vari-
ables such as type of meal, cutoff value, tracer dose, and delta time for the
analysis.
Discussion: The 13C-UBT performance meta-analyses showed 1, good accu-
racy in all ages combined (sensitivity 95.9%, specificity 95.7%, LR+ 17.4,
LR) 0.06, diagnostic odds ratio (DOR) 424.9), 2, high accuracy in children
>6 years (sensitivity 96.6%, specificity 97.7%, LR+ 42.6, LR) 0.04, DOR
1042.7), 3, greater variability in accuracy estimates and on average a few
percentage points lower, particularly specificity, in children 6 years (sensi-
tivity 95%, specificity 93.5%, LR+ 11.7, LR) 0.12, DOR 224.8). Therefore,
the meta-analysis shows that the 13C-UBT test is less accurate for the diag-
nosis of H. pylori infection in young children, but adjusting cutoff value,
pretest meal, and urea dose, this accuracy can be improved.

Helicobacter pylori is a gram-negative micro-aerophilic
bacterium that infects over 50% of the worldwide popu-
lation and is associated with the development of peptic
ulcer, gastric cancer, or MALT lymphoma. Although
severe disease is manifested during adulthood, the infec-
tion is usually acquired during childhood [13]. In adults
with severe symptoms, invasive diagnostic methods are
used, and in children, no clear association with disease
or symptoms has been confirmed. Diagnosis of H. pylori
in children can be made with both invasive and non-
invasive tests. Invasive tests require gastrointestinal
endoscopy to obtain biopsies of the gastric mucosa and
include culture, histology, and rapid urease test (RUT).
Although most of the invasive tests are considered gold
standards for the diagnosis of H. pylori infection because
they directly document the presence of the bacteria,
these tests are not recommended for use in children, and
noninvasive tests are desirable options [46].
Helicobacter pylori produces large amounts of urease
(urea-amidohydrolase), an enzyme that is essential for
the bacteria to colonize the acidic stomach environment
[79]. The strong activity of this enzyme is used as a
marker for the presence of H. pylori in different diag-
nostic methods, and the most widely used is the urea
breath test (UBT) [1013].
The UBT is an accurate noninvasive method for
assessing H. pylori infection status in adults. This test
detects the activity of urease in the presence of

2011 Blackwell Publishing Ltd, Helicobacter 16: 327337 327

13
C-UBT H. pylori Diagnosis in Children
isotopically labeled urea. If bacteria are present in the
stomach, the urea is hydrolyzed into ammonia and
carbon dioxide (CO2); the isotopically labeled CO2 diffuses
into the blood and is eliminated from the lungs in the
exhaled breath. There are two carbon isotopes used for
UBT, nonradioactive 13C and radioactive 14C [1012].
The exhaled 13CO2 is usually measured by isotope ratio
mass spectrometry (IRMS). More recently, a less expen-
sive method using nondispersive infrared spectrometry
(NDIRS) has demonstrated similar analytical accuracy
as the IRMS UBT. Results are reported as delta over
baseline (DOB) values of the measured 13CO2 12CO2
ratio [14,15]. The urea hydrolysis rate (UHR) is a new
alternative form of expressing results, and it is a func-
tion of anthropometric variables, which determine the
rate of endogenous 12CO2 production; thus, UHR can
normalize the DOB values [16]. The 13C-urea breath
test (13C-UBT) consistently shows high sensitivity and
specificity for the diagnosis of H. pylori infection in
adult populations, ranging from 95 to 100% [17].
The 13C-UBT has become the most convenient method
for use in children because it is a noninvasive method
and uses 13C, a stable and nonradioactive isotope [46].
The Canadian Consensus Group concluded that 13C-UBT
is a reliable noninvasive test in children, although it is
acknowledged that the test is less accurate in younger
children [18]. In addition, the European H. pylori Study
Group also recommended the 13C-UBT for diagnosis and
also for assessment of eradication of the infection after
antimicrobial treatment. However, no standard protocol
is available and results differ across laboratories
[46,10,18]. A previous 13C-UBT meta-analysis based on
study populations of mostly adults included 12 studies
and reported summary estimates of sensitivity 96.5%,
specificity 96%, likelihood ratio LR+ 24 and LR) 0.04
[19]. Studies in children have shown less consistent
performance, with values of sensitivity and specificity
ranging from 80 to 100%. No published meta-analysis
has focused on children [14,2022].
Owing to the aforementioned observations, we car-
ried out a systematic review and meta-analysis to eval-
uate the performance of the 13C-UBT for diagnosis of
H. pylori infection in children. Data were analyzed to
estimate sensitivity and specificity, as well as additional
accuracy values relevant to clinical practice.
Materials and Methods
Study Identification
We followed standard guidelines and methods for
systematic reviews and meta-analysis [2325]. We
included studies published between December 1997 to
328
Leal et al.
May 2009 in the PUBMED, EMBASE, and LILACS
databases. The search terms used were H. pylori
Children, Breath Test, Diagnostic, 13C-UBT,
Sensitivity, Specificity and Detection. For the final
set of selected publications, we included articles available
online and direct communications with the correspond-
ing authors when the article was not available online.
Study Selection
We included articles that met the following predeter-
mined criteria: 1, comparison of 13C-UBT H. pylori test
with a reference standard (we considered as the refer-
ence standard the following tests: H. pylori culture,
histologic examination, or RUT); 2, evaluation of a
minimum of 30 participants; 3, study report available in
either English or Spanish language (although our initial
literature search had no language restrictions); and 4,
presentation of the actual numbers of true positives,
true negatives, false positives, and false negatives. With
the numeric information, we calculated the following
values: sensitivity, specificity, LR+, LR), diagnostic odds
ratio (DOR), and their corresponding 95% confidence
intervals (95% CI) [26,27]. Reviews, letters to the
editor, opinions, and recommendations about the diag-
nosis of H. pylori infection in children does not fulfill
the meta-analysis quality assessments of diagnostic
studies [23,25] and therefore were excluded from this
meta-analysis. The extracted data from all articles
included in the meta-analysis were reviewed by one
expert, and for quality purposes, a second reviewer
independently extracted data from a subset of the
articles (15 (48%)). We observed a 98% agreement
between the two reviewers, and the remaining discrep-
ancies in the interpretation were resolved by consensus.
Extracted data were organized in an Excel database,
which was cross-checked for input errors.
Assessment of Study Quality
The quality of the studies was evaluated using recom-
mended dimensions for quality assessment of diagnostic
studies [28] addressing the following questions: 1, was
the urea breath test considered a gold standard on itself
and results were not compared with any of the refer-
ence standard tests defined above? 2, Was the whole
sample or a randomly selected subset of the sample
confirmed using the reference standard? 3, Was the
13
C-UBT performed by technicians who were unaware
of (i.e., blinded to) the reference standard results? 4,
Did the study recruit prospectively children suspected
of having H. pylori infection (i.e., case-control vs
cross-sectional design)?
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337

Leal et al.
Meta-Analysis
This was performed with a fixed-effects model using
MetaDiSc Beta-1.4 software (Universidad Complutense,
Madrid, Spain) [29]. Our analysis focused on the
following summary measures of diagnostic accuracy:
sensitivity (true positive results), defined as the propor-
tion of tests to be correctly identified as H. pylori posi-
tive by a reference standard (culture and or histology);
specificity (true negative results), defined as the propor-
tion of tests to be correctly identified as H. pylori nega-
tive by a reference standard; LR+ (likelihood ratio),
which measures how many times a positive test is more
likely found in infected versus noninfected children;
LR), which measures how many times a negative result
is more likely found in infected versus noninfected
children; and the DOR, defined as the ratio of the odds
of a positive test result occurring in children with infec-
tion compared with children without infection. The
DOR combines sensitivity and specificity into a single
measure of test accuracy, ranging from zero to infinity,
with higher values indicating better discriminatory test
performance or higher accuracy [27]. Forest plots were
created to display estimates of sensitivity and specificity
and to examine heterogeneity (variability across
studies). We summarized the joint distribution of the
true positive rate (TPR) and the true negative rate
(1-false positive rate (FPR)) with a summary receiver
operating characteristic (SROC) curve. The SROC curve
in studies of diagnostic accuracy represents the relation-
ship between TPR and FPR across studies when test per-
formance is evaluated at varying diagnostic thresholds
[24,30]. The overall diagnostic performance of a test
can be judged by the position and appearance of the
SROC curve, which is fitted by using a regression
model. The area under the curve (AUC) represents an
overall summary measure of the curve and the tests
overall ability to accurately distinguish cases from non-
cases; an AUC of one represents perfect discriminatory
ability. The Q* index is the highest point in the SROC
curve that intersects the antidiagonal and represents a
summary of test performance where sensitivity and
specificity are equal. A Q* index of 1.0 indicates perfect
accuracy (with sensitivity and specificity both equal to
100%). Both AUC and Q* range from 0 to 1, and
higher values indicate better test performance [30].
Exploration of Heterogeneity by Age Subgroups
To evaluate heterogeneity across studies, we assessed
differences in cutoff values, age, test protocols, and
study quality. We also identified differences in test
design and used protocols, and therefore, we defined
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337
13
C-UBT H. pylori Diagnosis in Children
subgroup dividing studies into: 1, tracer dose (unique
dose vs per kg of body weight dose), 2, test meals (citric
acid vs glucose vs fatty meal vs juice vs no meal),
3, fasting time (2 vs 4 hours vs overnight), 4, breath
collection method (bag vs straw vs mask), 5, delta time
(15 vs 20 vs 30 vs 40 vs 60 minutes), 6, cutoff value
(fixed by ROC curve vs not fixed and ranging between
2.0 and 14.0), 7, isotope ratio measurement method
(IRMS vs NDIRS), 8, reported test value (DOB vs UHR),
and 9, time period for restricting prior intake of other
drugs (2 vs 4 vs 8 weeks). To explore the role of age as
a source of heterogeneity in the accuracy of 13C-UBT
results, the age group was further divided into two
subgroups: children 6 and >6 years old, and each age
subgroup was considered against all variables men-
tioned previously. Heterogeneity was assessed using the
chi-squared test with MetaDiSc Beta-1.4 software
(Universidad Complutense, Madrid Spain) [29].
Results
Study Selection
Figure 1 shows the study selection process. The search
in the selected databases retrieved 468 potentially rele-
vant references. After screening titles and abstracts, 236
English and Spanish articles were selected for full-text
review. Only 51 of these articles met the eligibility crite-
ria, and from them, eight were excluded because of the
lack of adequate information about sensitivity and spec-
ificity, and another 12 because of the lack of adequate
definitions of true H. pylori positive and negative status.
In the end, 31 articles met the eligibility criteria and
were included in the meta-analysis [14,2022,3157].
Characteristics of Included Studies
In 23 (74.2%) of the 31 articles included in the meta-
analysis, H. pylori culture and histology were used as
the gold standard [14,2022,3336,3840,4250,5254];
in four (12.9%) histology combined with RUT was the
gold standard [31,32,51,56]; and four articles (12.9%)
reported histology alone as gold standard [37,41,55,57].
We did not find differences in the performance of UBT
when compared against H. pylori culture or histology,
and thereby both tests were considered a suitable gold
standard. For all articles included, the 13C-UBT did not
form part of the reference standard.
Two papers in Spanish were included [42,43]. Twelve
articles (38.7%) reported at least single-blind interpreta-
tion of the 13C-UBT and reference standard results,
while 19 (61.3%) did not mention the blinding status
[14,2022,3135,37,3943,47,50,52,53]. All articles
329
13
C-UBT H. pylori Diagnosis in Children Leal et al.

Potentially relevant references (titles/abstracts)

identified from all searches (n = 468)

References excluded after initial screening (n = 232)

Articles selected for full-text review (n = 236)

Articles excluded (n = 185)

Reasons for exclusion
Duplicated publications in databases = 23
Non-English and non-Spanish language = 55
Letters to editor/case report = 15
Reviews = 26
No comparison against reference standard = 10
Treatment monitoring study = 56

Articles met eligibility criteria (n = 51)

Articles excluded (n = 20)

Reasons for exclusion
Inadequate sensitivity and specificity data = 8
Inadequate definition of H. pylori (+) and () = 12

31 articles included in meta-analysis of 13C-UBTdetection test for H.

pylori infection in children (135 studies)

Figure 1 Flowchart indicating the selection of cases and reasons for exclusion.

reported studies that used a case-control design and pro-
spective data collection. The median sample size was
98.6 participants (interquartile range 58.5155). A total
of 4037 children were included in the meta-analysis,
from these, 1527 were H. pylori positive (cases) and 2510
were H. pylori negative (controls) by the gold standard.
Sixteen (51.6%) articles reported the evaluation of more
than one variation in the 13C-UBT protocol, and in these
cases, each comparison was counted as a separate study.
Thus, a total of 135 test comparisons (hereafter referred
to as studies) were included in the meta-analysis.
13
Overall Accuracy of the C-Urea Breath Test
We included all 135 studies in the initial meta-analysis.
Because all summary measures except DOR were signif-
icantly heterogeneous (Table 1), we explored possible
causes for heterogeneity; first, we analyzed variation in
test protocols, and then we divided the data into two
subgroups younger and older than 6 years old.
Exploration of Heterogeneity
Among the variation in 13C-UBT protocols assessed,
delta time, fasting time, type of meal, tracer dose,
and type of reported results seemed to be associated
with differences in DOR. The remaining four variables
(breath collection method, fixed cutoff value, isotope
ratio measurement method, and restriction period
prior intake of other drugs) had not association.
A delta time of 20 minutes had the best accuracy,
giving DOR estimated values 3.1 times those obtained
from delta times 15 minutes (705.5 vs 223.0), and
1.3 and 2.5 times those obtained from 30 and
60 minutes delta time (515 and 279.3, respectively).
Fasting time of 4 hours produced DOR estimates that
were 3.8 fold higher than those corresponding to
2 hours (627.2 vs 166.5) and 2.1 fold higher than
those corresponding to overnight fasting (295.3).
Protocols that did not include a meal produced DOR
estimates that were 8.3-fold higher than glucose
(1274.3 vs 153.4), 3.6-fold higher than a fatty meal
(357.9), 2.4-fold higher than apple grape orange juice
(525.4), and twofold higher than citric acid (648).
UHR values produced DOR estimates that were
1.3-fold higher than DOB results (849.4 vs 622.8).
Finally, a unique tracer dose of 13C-urea produced
DOR estimates 2.5-fold higher than when the dose
was adjusted to the body weight of the tested
children (479.4 vs 194.5).

330 2011 Blackwell Publishing Ltd, Helicobacter 16: 327337

 13
Leal et al. C-UBT H. pylori Diagnosis in Children

Table 1 Description of studies included in meta-analysis and hetero- A Sensitivity (95% CI)
geneity test De Carvalho L (2003) 0.88 (0.47 1.00)
Machado R (2004) 0.93 (0.68 1.00)
Machado R (2004) 1.00 (0.03 1.00)
Machado R (2004) 0.93 (0.66 1.00)
Summary of test Test for Frenck R (2006) 1.00 (0.80 1.00)
Imrie C (2001) 1.00 (0.54 1.00)
accuracy values heterogeneityap Imrie C (2001)
Imrie C (2001)
1.00
1.00
(0.54 1.00)
(0.54 1.00)
Imrie C (2001) 0.67 (0.22 0.96)
Accuracy measure (95% CI) value Imrie C (2001) 1.00 (0.54 1.00)
Imrie C (2001) 0.67 (0.22 0.96)
Dondi E (2006) 0.93 (0.78 0.99)
13 Dondi E (2006) 0.93 (0.78 0.99)
C-UBT in 135 studies (Hp+ n = 4616; Hp) n = 7949) Kinderman A (2000) 1.00 (0.72 1.00)
Sensitivity 95.9 (95.396.4) .049 Yang H (2005) 1.00 (0.75 1.00)
Yang H (2005) 1.00 (0.75 1.00)
Yang H (2005) 1.00 (0.75 1.00)
Specificity 95.7 (95.396) <.001 Yang H (2005) 1.00 (0.75 1.00)
Yang H (2005) 1.00 (0.75 1.00)
LR+ 17.4 (14.620.7) <.001 Elitsur Y (2009) 1.00 (0.54 1.00)
LR) 0.06 (0.050.07) .033 Kawakami E (2002) 0.89 (0.52 1.00)

DOR 250.2 (120.4520.2) .683

Pooled sensitivity = 0.95 (0.91 0.97)
13
C-UBT in children 6 years in 21 studies (Hp+ n = 242; Hp) n = 1159) 2
c = 23.96; d.f. = 20 (p = .2440)
2
0 0.2 0.4 0.6 0.8 1 Inconsistency (I ) = 16.5 %
Sensitivity 95.0 (91.597.4) .244 Sensitivity
Specificity 93.5 (92.194.9) <.001 B Sensitivity (95% CI)
De Carvalho L (2003) 1.00 (0.92 1.00)
LR+ 11.7 (8.316.7) <.001 Machado R (2004) 0.96 (0.87 1.00)
LR) 0.12 (0.080.18) .570 Machado R (2004) 1.00 (0.75 1.00)
Machado R (2004) 0.95 (0.83 0.99)
DOR 224.8 (123.9407.9) .999 Frenck R (2006) 0.86 (0.70 0.95)
Imrie C (2001) 0.94 (0.83 0.99)
Imrie C (2001) 0.78 (0.63 0.88)
13
C-UBT in children >6 years in 11 Studies (Hp+ n = 493; Hp) n = 1010) Imrie C (2001) 1.00 (0.93 1.00)
Imrie C (2001) 0.98 (0.89 1.00)
Sensitivity 96.6 (94.598.0) .503 Imrie C (2001) 0.92 (0.80 0.98)
Imrie C (2001) 1.00 (0.93 1.00)
Specificity 97.7 (96.698.6) .054 Dondi E (2006) 0.95 (0.90 0.98)
Dondi E (2006) 0.98 (0.94 1.00)
LR+ 42.6 (22.281.9) .035 Kinderman A (2000) 0.88 (0.74 0.96)
Yang H (2005) 0.90 (0.79 0.96)
LR) 0.04 (0.030.06) .885 Yang H (2005) 0.85 (0.73 0.93)
DOR 1402.7 (686.52865.8) .991 Yang H (2005) 0.90 (0.79 0.96)
Yang H (2005) 0.95 (0.86 0.99)
Yang H (2005) 0.95 (0.86 0.99)
a Elitsur Y (2009) 0.79 (0.49 0.95)
Chi-squared and p value. Kawakami E (2002) 0.96 (0.78 1.00)

CI, Confidence Interval; DOR, diagnostic odds ratio; LR, likelihood ratio;
13
C-UBT, 13C-urea breath test.
Pooled specificity = 0.94 (0.92 0.95)
2
c = 60.53; d.f. = 20 (p = .0000)
2
0 0.2 0.4 0.6 0.8 1 Inconsistency (I ) = 67.0 %
Specificity

Accuracy of 13C-UBT for Diagnosis of H. pylori
Infection in Children 6 Years Old
Twenty-one studies assessed the diagnostic accuracy of
13
C-UBT in children 6 years of age and younger and
included 1401 children, 242 H. pylori positive and 1159
H. pylori negative. Table 1 shows the overall accuracy
measures in this group. Figure 2A,B shows the forest
plot of sensitivity and specificity estimates. Variations in
both estimates were notable: sensitivity ranged from 50
to 100%, whereas specificity varied from 61 to 100%.
The corresponding SROC (Fig. 3) shows an area under
curve of 0.97 and a Q* of 0.93 indicating good accu-
racy; the summary DOR value was 224.8 (95% IC,
123.9407.9) and although heterogeneity appears to be
negligible (p = .999), the test for heterogeneity of speci-
ficity estimates across studies showed it to be extreme
(p < .0001). To identify factors associated with this con-
siderable heterogeneity, we stratified into subgroups
according to variation in the 13C-UBT protocols and
compared their performance. In children under 6 years
old, tracer dose of 50 mg, juice meal, and test value
cutoff 6.0 showed better overall performance as seen
in Table 2.
Figure 2 Forest plot for the sensitivity (A) and specificity (B) of the
13
C-urea breath test achieved in children 6 years old and younger.
The squares and lines represent the point estimates and 95% CIs,
respectively. The size of the squares indicates the study size. The dia-
mond at bottom denotes pooled estimated values.
Accuracy of 13C-UBT for Diagnosis of H. pylori
Infection in Children >6 Years Old
Eleven studies assessed the performance of 13C-UBT in
the subgroup of children older than 6 years and
included 1503 children, 493 H. pylori positive, and 1010
H. pylori negative. Almost all 11 studies show nearly
perfect estimates of sensitivity and specificity (Table 1).
Figure 4A,B shows the corresponding forest plot, and
both estimated values ranged from about 94 to 100%.
Figure 5 shows the SROC for this subgroup, with
AUC = 0.99, and Q* = 0.97; the curve shows a clear
trade-off between sensitivity and specificity. In addition,
the summary DOR was 1402.7 (95% CI, 686.52865.8)
and it showed negligible heterogeneity (p = .991). The
test for heterogeneity across studies had a p-value of
.05 (Table 1).

2011 Blackwell Publishing Ltd, Helicobacter 16: 327337 331

13
C-UBT H. pylori Diagnosis in Children Leal et al.

Sensitivity SROC Curve analysis show that 1, 13C-UBT has good overall perfor-
1
Symmetric SROC
mance with sensitivity and specificity of 95%, LR+ of
0.9 AUC = 0.9795
SE(AUC) = 0.0050 17.4, LR) of 0.06, and DOR of 424.9; 2, the overall
Q* = 0.9363
0.8 SE(Q*) = 0.0093 accuracy of 13C-UBT in children older than 6 years of
0.7 age (11 studies) showed the best performance with sen-
0.6 sitivity of 96.6% and specificity of 97.7%; this is also
0.5
reflected in the high LR+, LR), and DOR estimates of
42.6, 0.04, and 1042.7, respectively; 3, 13C-UBT per-
0.4
formed in children under 6 years of age (21 studies)
0.3
showed less consistent and lower accuracy measures
0.2
with an overall sensitivity of 95%, specificity of 93.5%,
0.1 LR+ of 11.7, LR) of 0.12, and DOR of 224.8; 4, adjust-
0 ments in the cutoff value (6.0), the intake of juice
0 0.2 0.4 0.6 0.8 1
1-specificity meal pretest, and labeled urea dose (50 mg) may be
required to improve the accuracy in this group of
Figure 3 Summary of receiver operator curve (SROC) for 13C-urea children.
breath test achieved in children 6 years old and younger. Each solid
square represents an individual study in the meta-analysis. The curve
is the regression line that summarizes the overall diagnostic accuracy. Analysis of Heterogeneity
AUC = area under curve, SE (AUC) = standard error of AUC,
Q* = index defined by the point of the SROC curve where the sensitivity Since the original description by Graham of the 13C-UBT
and specificity are equal; SE (Q*) = standard error of Q* index. to diagnose H. pylori infection [10], several modifications
have been proposed; still criteria for test performance
and interpretation of results in children are not yet suffi-
Discussion
ciently defined. Potential explanations for the consider-
able heterogeneity in the accuracy of the 13C-UBT across
Principal Findings
studies of children, and in particular, in the frequency of
This systematic review identified 135 studies that false positives, include the following: 1, urease activity
addressed the diagnostic accuracy of 13C-UBT for from the oral bacterial flora [58,59], 2, differences in
H. pylori infection in children. The results of the meta- delta time, and 3, variability in cutoff values. To

13
Table 2 Stratified analyses for evaluation of C-UBT accuracy in children 6 years old and younger

Test property Summary of test accuracy values p valuea Summary of test accuracy values p valuea

Studies = 13 Body weight dose 75 50 mg Unique dose 50 mg

Sensitivity 88.8 (80.394.5) .201 93.3 (90.997.2) .998
Specificity 94.0 (91.995.7) .000 96.4 (91.794.5) .462
LR+ 12.8 (7.422.2) .000 22.8 (7.715.3) .676
LR) 0.26 (0.180.39) .756 0.08 (0.090.23) .900
DOR 100.7 (48.5209.0) .981 334.5 (89.8277.3) .873

Studies = 27 Glucose meal Juice meal

Sensitivity 88.9 (79.395.1) .123 94.1 (88.797.4) .812
Specificity 93.5 (91.295.4) .000 95.3 (93.296.9) .030
LR+ 11.6 (6.720.1) .000 15.9 (9.3027.3) .068
LR) 0.28 (0.190.43) .712 0.08 (0.040.15) .946
DOR 89.3 (41.1193.9) .976 292.3 (129.5659.9) .988

Studies = 13 Cutoff value 2.5 Cutoff value 6.0

Sensitivity 100 (91.8100) 1.000 97.6 (91.799.7) .379
Specificity 84.4 (79.588.5) .029 94.6 (91.696.7) .025
LR+ 5.3 (3.58.0) .076 13.4 (6.726.7) .027
LR) 0.11 (0.040.28) .988 0.05 (0.020.13) .984
DOR 56.7 (18.9169.6) .928 398.9 (125.11271.6) .901

a
Test for heterogeneity, chi-squared and p value.
DOR, diagnostic odds ratio; LR, likelihood ratio; 13C-UBT, 13
C-urea breath test.

332 2011 Blackwell Publishing Ltd, Helicobacter 16: 327337

 13
Leal et al. C-UBT H. pylori Diagnosis in Children

A Sensitivity (95% CI) [21]. Also, the administration of 13C-urea in capsules to

De Carvalho L (2003) 0.95 (0.83 0.99)
Frenck R (2006) 0.97 (0.82 1.00) avoid activity of oral bacteria has shown excellent
Ni Y (2000) 1.00 (0.87 1.00)
Kinderman A (2000) 0.98 (0.90 1.00) results in adults; however, this has not been tested in
Yang H (2005) 0.97 (0.86 1.00)
Yang H (2005)
Yang H (2005)
0.99
0.97
(0.92 1.00)
(0.90 1.00)
infants or toddlers [60,61]. Concerning variations in
Yang H (2005) 0.94 (0.86 0.98) delta time, Rowland et al. [33] reported a significant
Yang H (2005) 0.93 (0.84 0.98)
Elitsur Y (2009) 1.00 (0.85 1.00)
Kawakami E (2002) 1.00 (0.72 1.00)
decrease in specificity when samples were obtained at
15 minutes instead of 30 minutes. Finally, the cutoff
value represents a crucial factor for accuracy trade-offs
Pooled sensitivity = 0.97 (0.95 0.98)
2
c = 9.31; d.f. = 10 (p = .5028) of the test, where cutoff values low might increase sensi-
0 0.2 0.4 0.6 0.8 1 Inconsistency (I 2) = 0.0 %
Sensitivity tivity but reduce specificity, and vice versa. There was
B Specificity (95% CI) no uniformity of the cut point for defining positive and
De Carvalho L (2003)
Frenck R (2006)
0.99
0.95
(0.92 1.00)
(0.74 1.00)
negative results across studies included in this meta-
Ni Y (2000) 1.00 (0.87 1.00)
Kinderman A (2000) 0.98 (0.88 1.00)
analysis; this value ranged from 2.0 to 14.0. Some
Yang H (2005) 0.99 (0.95 1.00)
Yang H (2005) 0.93 (0.88 0.96)
authors fixed the cutoff value by the ROC curve; for
Yang H (2005) 0.99 (0.96 1.00)
Yang H (2005) 0.99 (0.96 1.00) example, Kato et al. [34] reported that 3.5% was an
Yang H (2005) 0.99 (0.96 1.00)
Elitsur Y (2009) 0.98 (0.91 1.00) adequate fixed value that achieves 97.8% of sensitivity
Kawakami E (2002) 1.00 (0.66 1.00)
and 98.5% specificity.
To identify factors associated with the considerable
Pooled specificity = 0.98 (0.97 0.99)
2
c = 18.69; d.f. = 10 (p = .0444)
heterogeneity observed in 13C-UBT accuracy estimates
2
0 0.2 0.4 0.6 0.8 1 Inconsistency (I ) = 46.5 % (Table 1), we stratified in subgroups according to varia-
Specificity
tions in the protocol test. However, considerable heter-
ogeneity persisted even after adjusting for variables that
Figure 4 Forest plot for the sensitivity (A) and specificity (B) of the our analysis showed to improve accuracy (20 minutes
13
C-urea breath test achieved in children over 6 years old. The
of delta time, 4 hours of fasting time, no meal, unique
squares and lines represent the point estimates and 95% CIs, respec-
tively. The size of the squares indicates the study size. The diamond
tracer dose, and results reported as UHR values). Heter-
at bottom denotes pooled estimated values. ogeneity persisted in part because the 13CO2 exhaled
depends not only on the 13C-urea dose given and the
amount hydrolyzed by the urease from H. pylori but
Sensitivity SROC Curve also on the individuals CO2 production or the degree
1
Symmetric SROC of 13CO2 diluted within the bodys CO2 and bicarbonate
0.9 AUC = 0.9951
SE(AUC) = 0.0017
Q* = 0.9729
pool, because the CO2 production is known to be influ-
0.8 SE(Q*) = 0.0056
enced by anthropometric characteristic as well as by
0.7 age and sex [16,62]. Thus, young children with rela-
0.6 tively low weight and height may produce smaller
0.5 amounts of endogenous CO2. Accordingly, we analyzed
0.4
two age subgroups, 6 and >6 years of age and further
explored this heterogeneity.
0.3
In the group of children older than 6 years of age,
0.2
the 13C-UBT showed greater accuracy, as shown in
0.1 Table 1. The estimated DOR was six times greater than
0 in the group of children 6 years of age, and heteroge-
0 0.2 0.4 0.6 0.8 1
1-specificity neity appeared to be less extreme (p = .05 for reports
pertaining to older children and p < .001 for reports
Figure 5 Summary of receiver operator curve (SROC) for 13C-urea pertaining to younger children). In children 6 years of
breath test achieved in children over 6 years old. Each solid square age and younger, the shape of the SROC suggested that
represents an individual study in the meta-analysis. The curve is the variability in the thresholds could partially explain
regression line that summarizes the overall diagnostic accuracy.
the heterogeneity (Fig. 3). Our analysis identified the
AUC = area under curve, SE (AUC) = standard error of AUC,
Q* = index defined by the point of the SROC curve where the sensitiv-
following as potentially important sources of hetero-
ity and specificity are equal; SE (Q*) = standard error of Q* index. geneity: 1, tracer dose, 2, pretest meal, and 3, cutoff
value. We found that a unique tracer dose of 50 mg of
13
counteract the oral flora, some authors have C-urea shows greater accuracy than when it was
recommended washing the patients mouth adjusted to body weight (5075 mg were used between
[22,34,39,44,48,51] or the use of a nasogastric tube studies), with a DOR estimated value of 3.3.

2011 Blackwell Publishing Ltd, Helicobacter 16: 327337 333

13
C-UBT H. pylori Diagnosis in Children
Accordingly, Megraud [44] reported that reducing the
dose from 75 to 45 mg in younger children resulted in
improved specificity. Also, a pretest meal of orange
grape apple juice produced DOR estimates that were
3.3-fold higher than glucose, most likely because gastric
acid stimulates the bacterial urease activity and in vitro
studies have shown that urease activity is almost 40-
fold greater at pH 3.5 compared with pH 7.4 [63]. It
has also been reported that citric acid acts with Ure1 to
counteract the DOB-reducing effect of bicarbonate or
H2-receptor inhibitors [64]; thus, citric acid has demon-
strated good performance in adults. However, citric acid
is not well accepted by children, and apple, orange, or
grape juice seems to be a good alternative. Finally, a
cutoff value 6.0 improved overall performance with a
DOR estimate of 398.9. Yang and Seo [52] reported that
in children older than 6 years of age, the optimal cutoff
value (4.0%) was the same as the manufactures
recommendation for adults, whereas in children aged
6 years or younger, the optimal cutoff value was found
to be 7.0.
The application of the 13C-UBT in children is much
needed because it avoids the use of invasive and painful
procedures, such as endoscopy and blood draw. Also,
gold standard tests (histology, culture, and RUT) might
show false negative results in young children where
colonization of the stomach is commonly weak and
patchy. In this sense, the 13C-UBT might overcome this
problem because it samples the whole stomach for
H. pylori [54,65].
Our results show that the 13C-UBT is a highly accu-
rate and convenient tool for the diagnosis of H. pylori
infection in children older than 6 years of age, but in
younger children, there seems to be less certainty of
how accurately this test performs. In addition to
younger children having lower weight and height, and
thereby producing smaller amounts of endogenous
CO2, our meta-analysis identifies that the younger the
children, the more variable conditions are used in the
13
C-UBT, and this variability is contributing to the less
accuracy found in children 6 years of age. This study
supports that accuracy may approach that observed in
older children if the test is subjected to fixed conditions,
mainly tracer dose, pretest meal, and cutoff value.
It is important to note that because of a lower num-
ber of studies enrolling children 6 years of age, our
meta-analysis could not further subdivide ages 6 and
younger and therefore does not identify the specific age
where the 13C-UBT becomes less accurate. However,
this study found that there are less standardized param-
eters when the test is applied to children 2 years old,
and this correlates with more variable results and lower
test specificity.
334
Leal et al.
A problem with the 13C-UBT is that it is not accessi-
ble in many developing countries because of its cost. In
the USA, the cost per sample has been reported to be
$104 [66], because the IRMS generally used to measure
13
C enrichment in breath samples is very expensive
[15]. To reduce costs, cheaper equipment based on
nondispersive, isotope selective, infrared spectroscopy
(NDIRS), or laser-assisted ratio analysis (LARA) has
recently been developed. These are valid alternatives to
IRMS and will certainly promote the wider use of the
13
C-UBT, which might be especially useful for epidemi-
ologic studies of children, for screening symptomatic
children before endoscopy, and for assessing the effi-
cacy of eradication regimens [14,15]. Furthermore, our
meta-analysis does not show any difference between
the quality of the data obtained through IRMS or
NDIRS measurement.
Clinical and Epidemiologic Implications
Based on the results of this meta-analysis, clinicians
should consider the 13C-UBT to diagnose active H. pylori
infection, particularly in children older than 6 years of
age. However, our results suggest that in children
6 years old and younger, the 13C-UBT is less accurate,
with a specificity of 93.5% and the corresponding LR+
of 11.7. Also, the LR) of 0.12 informs us that infection
cannot be excluded when the test is negative in this age
group [26]. Our results are in agreement with the high
frequency of false positives reported in children younger
than 6 years old. This lower accuracy might have lead
to wrong conclusions in epidemiologic studies and
could partially explain the observed phenomenon of
transient infection in children, attributed to a high
spontaneous clearance rate of H. pylori during infancy
[6770]. Therefore, we should be very cautious when
interpreting results obtained with the 13C-UBT in
children younger than 6, and especially younger than
2 years old. For this age group, results should be
confirmed with a different method, especially in popula-
tions with a low prevalence of H. pylori infection. We
have recently found that a stool antigen test is also reli-
able in children and could be used together with the
13
C-UBT [71,72].
Strengths and Weaknesses of the Review
An important strength of our study is its comprehensive
search strategy. Screening, study selection, and quality
assessment were done independently by two reviewers.
For some studies, we reduced the problem of missing
data by contacting the authors directly. We also
explored heterogeneity and potential publication bias in
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337

Leal et al.
accordance with published guidelines [2325,73]. We
analyzed data within specific subgroups to lessen the
effect of heterogeneity.
We also acknowledge some limitations of this work.
First, we were able to include only English and Spanish
language articles, and this might have introduced some
selection bias into our results. Second, we did not
address the effect of factors such as laboratory infrastruc-
ture, expertise with the test technology and the patient
clinical and histologic manifestations. Although we used
guidelines such as Standards for Reporting of Accuracy
(STARD) [28] to improve the quality of the analysis, our
findings should be interpreted in the context of the
quality and variability of the included studies.
In summary, our meta-analysis supports current
evidence, indicating that the 13C-UBT test is highly
accurate for the diagnosis of H. pylori infection in
children older than 6 years of age, but it is less certain in
younger children. However, our meta-analysis also sup-
ports that the accuracy of the test could be improved
with adjustments in the cutoff value, pretest meal, and
urea dose.
Acknowledgements and Disclosures
We thank Karen J. Goodman, of the University of Alberta
Centre of Excellence for Gastrointestinal Inflammation and
Immunity Research, for her helpful comments and critical
revision of the manuscript. JT is a recipient of an exclusivity
scholarship from Fundacion IMSS, Mexico.
Conflict of Interest
The authors have no competing interests.
Funding
The authors have no support or funding to report.
References
1 Perez-Perez GI, Rothenbacher D, Brenner H. Epidemiology of
Helicobacter pylori infection. Helicobacter 2004;9(Suppl 1):16.
2 Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J
Med 2002;347:117586.
3 Kusters J, van Vliet A, Kuipers E. Pathogenesis of Helicobacter
pylori infection. Clin Microbiol Rev 2006;19:44990.
4 Drumm B, Day A, Gold B, et al. European Society for Paediatric
Gastroenterology, Hepatology and Nutrition. Helicobacter pylori
and peptic ulcer: Working Group Report of the second World
Congress of Pediatric Gastroenterology, Hepatology, and
Nutrition. J Pediatr Gastroenterol Nutr 2004;39(Suppl 2):
S62631.
5 Czinn S, Gold B, Dickinson C, et al. Research agenda for pedi-
atric gastroenterology, hepatology and nutrition: acid-peptic
diseases. Report of the North American Society for Pediatric
Gastroenterology, Hepatology and Nutrition for the Childrens
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337
13
C-UBT H. pylori Diagnosis in Children
Digestive Health and Nutrition Foundation. J Pediatr
Gastroenterol Nutr 2002;35(Suppl 3):S2503.
6 Koletzko S. Noninvasive diagnostic tests for Helicobacter pylori
infection in children. Can J Gastroenterol 2005;19:4339.
7 Mobley H, Island M, Hausinger R. Molecular biology of micro-
bial ureases. Microbiol Rev 1995;59:45180.
8 Stingl K, Altendorf K, Bakker E. Acid survival of Helicobacter
pylori: how does urease activity trigger cytoplasmic pH homeo-
stasis? Trends Microbiol 2002;10:704.
9 Yoshiyama H, Nakazawa T. Unique mechanism of Helicobacter
pylori for colonizing the gastric mucus. Microbes Infect
2000;2:5560.
10 Graham D, Klein P, Evans D Jr, et al. Campylobacter pylori
detected noninvasively by the 13C-urea breath test. Lancet
1987;1:11747.
11 Marshall B, Francis G, Langton S, et al. Rapid urease test in the
management of Campylobacter pyloridis associated gastritis. Am J
Gastroenterol 1987;3:20010.
12 Marshall B, Surveyor I. Carbon-14 urea breath test for the
diagnosis of Campylobacter pylori associated gastritis. J Nucl Med
1988;29:116.
13 Monteiro L, Oleastro M, Lehours P, Megraud F. Diagnosis of
Helicobacter pylori infection. Helicobacter 2009;14:814.
14 Kawakami E, Machado R, Reber M, Patricio F. 13 C-urea
breath test with infrared spectroscopy for diagnosing Helicobacter
pylori infection in children and adolescents. J Pediatr Gastro-
enterol Nutr 2002;35:3943.
15 Parente F, Bianchi P. The (13)C-urea breath test for non-inva-
sive diagnosis of Helicobacter pylori infection: which procedure
and which measuring equipment? Eur J Gastroenterol Hepatol
2001;13:8036.
16 Slater C, Preston T, Weaver L. Is there an advantage in normal-
ising the results of the Helicobacter pylori [13C] urea breath test
for CO2 production rate in children? Isotopes Environ Health Stud
2004;40:8998.
17 Gisbert J, Pajares J. Review article: 13C-urea breath test in the
diagnosis of Helicobacter pylori infection a critical review.
Aliment Pharmacol Ther 2004;20:100117.
18 Bourke B, Ceponis P, Chiba N, et al. Canadian Helicobacter
Study Group Consensus Conference: update on the approach
to Helicobacter pylori infection in children and adolescents an
evidence-based evaluation. Can J Gastroenterol 2005;19:399
408.
19 Roberts A, Childs S, Rubin G, de Wit N. Tests for Helicobacter
pylori infection: a critical appraisal from primary care. Fam Pract
2000;17:S1220.
20 De Carvalho Costa Cardinali L, Rocha G, Rocha A, et al.
Evaluation of [13C] urea breath test and Helicobacter pylori stool
antigen test for diagnosis of H. pylori infection in children from
a developing country. J Clin Microbiol 2003;41:33345.
21 Imrie C, Rowland M, Bourke B, Drumm B. Limitations to
carbon 13-labeled urea breath testing for Helicobacter pylori in
infants. J Pediatr 2001;139:7347.
22 Kato S, Nakayama K, Minoura T, et al. Comparison between
the 13C-urea breath test and stool antigen test for the diagnosis
of childhood Helicobacter pylori infection. J Gastroenterol
2004;39:104550.
23 Deville WL, Buntinx F, Bouter LM, et al. Conducting system-
atic reviews of diagnostic studies: didactic guidelines. BMC Med
Res Methodol 2002;2:921.
24 Lijmer J, Bossuyt P, Heisterkamp S. Exploring sources of
heterogeneity in systematic reviews of diagnostic tests. Stat Med
2002;21:152537.
335
13
C-UBT H. pylori Diagnosis in Children
25 Pai M, McCulloch M, Gorman J, et al. Systematic reviews and
meta-analyses: an illustrated, step-by-step guide. Natl Med J
India 2004;17:8695.
26 Deeks J, Altman D. Diagnostic tests 4: likelihood ratios. BMJ
2004;329:1689.
27 Glas AS, Lijmer JG, Prins MH, et al. The diagnostic odds ratio:
a single indicator of test performance. J Clin Epidemiol
2003;56:112935.
28 Smidt N, Rutjes A, van der Windt D, et al. Reproducibility of
the STARD checklist: an instrument to assess the quality of
reporting of diagnostic accuracy studies. BMC Med Res Methodol
2006;6:1221.
29 Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A.
Meta-DiSc: a software for meta-analysis of test accuracy data.
BMC Med Res Methodol 2006;6:3142.
30 Walter S. Properties of the summary receiver operating charac-
teristic (SROC) curve for diagnostic test data. Stat Med
2002;21:123756.
31 Hino B, Eliakim R, Levine A, et al. Comparison of invasive and
non-invasive tests diagnosis and monitoring of Helicobacter pylori
infection in children. J Pediatr Gastroenterol Nutr 2004;39:51923.
32 Bazzoli F, Cecchini L, Corvaglia L, et al. Validation of the 13C-
urea breath test for the diagnosis of Helicobacter pylori infection
in children: a multicenter study. Am J Gastroenterol
2000;95:64650.
33 Rowland M, Lambert I, Gormally S, et al. Carbon 13-labeled
urea breath test for the diagnosis of Helicobacter pylori infection
in children. J Pediatr 1997;131:81520.
34 Kato S, Ozawa K, Konno M, et al. Diagnostic accuracy of the
13C-urea breath test for childhood Helicobacter pylori infection: a
multicenter Japanese study. Am J Gastroenterol 2002;97:166873.
35 Machado RS, Patricio FR, Kawakami E. 13C-urea breath test to
diagnose Helicobacter pylori infection in children aged up to
6 years. Helicobacter 2004;9:3945.
36 Frenck R Jr, Fathy H, Sherif M, et al. Sensitivity and specificity
of various tests for the diagnosis of Helicobacter pylori in
Egyptian children. Pediatrics 2006;118:e1195202.
37 Rutigliano V, Ierardi E, Francavilla R, et al. Helicobacter pylori
and nonulcer dyspepsia in childhood: clinical pattern,
diagnostic techniques, and bacterial strains. J Pediatr Gastro-
enterol Nutr 1999;28:296300.
38 Ni Y, Lin J, Huang S, Yang J, Chang M. Accurate diagnosis of
Helicobacter pylori infection by stool antigen test and 6 other
currently available tests in children. J Pediatr 2000;136:8237.
39 Vincent P, Michaud L, Martin de L, Benon B, Turck D,
Gottrand F. 13C-urea breath test and gastric mucosal
colonization by Helicobacter pylori in children: quantitative
relation and usefulness for diagnosis of infection. Helicobacter
1999;4:2337.
40 Konstantopoulos N, Russmann H, Tasch C, et al. Evaluation of
the Helicobacter pylori stool antigen test (HpSA) for detection of
Helicobacter pylori infection in children. Am J Gastroenterol
2001;96:67783.
41 Eltumi M, Brueton M, Francis N. Diagnosis of Helicobacter pylori
gastritis in children using the 13C urea breath test. J Clin
Gastroenterol 1999;28:23840.
42 Parejo C, Olivares M, Escobar C, Jimenez A, de Rafael N,
Camarero S. [A comparative analysis of diagnostic methods for
Helicobacter pylori infection in children]. An Esp Pediatr
1998;49:25763.
43 Gonzalez-Cuevas A, Juncosa T, Jene M, et al. [Helicobacter pylori
infections: antigen detection in stool samples]. Enferm Infecc
Microbiol Clin 2001;19:4952.
336
Leal et al.
44 Megraud F. Comparison of non-invasive tests to detect Helico-
bacter pylori infection in children and adolescents: results of a
multicenter European study. J Pediatr 2005;146:198203.
45 Dondi E, Rapa A, Boldorini R, Fonio P, Zanetta S, Oderda G.
High accuracy of noninvasive tests to diagnose Helicobacter pylori
infection in very young children. J Pediatr 2006;149:81721.
46 Canete A, Abunaji Y, varez-Calatayud G, et al. Breath test
using a single 50-mg dose of 13C-urea to detect Helicobacter
pylori infection in children. J Pediatr Gastroenterol Nutr
2003;36:10511.
47 Cadranel S, Corvaglia L, Bontems P, et al. Detection of Heli-
cobacter pylori infection in children with a standardized and
simplified 13C-urea breath test. J Pediatr Gastroenterol Nutr
1998;27:27580.
48 Kalach N, Briet F, Raymond J, et al. The 13carbon urea breath
test for the noninvasive detection of Helicobacter pylori in
children: comparison with culture and determination of
minimum analysis requirements. J Pediatr Gastroenterol Nutr
1998;26:2916.
49 Yanez P, la Garza A, Perez G, Cabrera L, Munoz O, Torres J.
Comparison of invasive and noninvasive methods for the diag-
nosis and evaluation of eradication of Helicobacter pylori infec-
tion in children. Arch Med Res 2000;31:41521.
50 Corvaglia L, Bontems P, Devaster J, et al. Accuracy of serology
and 13C-urea breath test for detection of Helicobacter pylori in
children. Pediatr Infect Dis J 1999;18:9769.
51 Kindermann A, Demmelmair H, Koletzko B, Krauss-Etschmann
S, Wiebecke B, Koletzko S. Influence of age on 13C-urea
breath test results in children. J Pediatr Gastroenterol Nutr
2000;30:8591.
52 Yang H, Seo J. Diagnostic accuracy of the C-urea breath test in
children: adjustment of the cut-off value according to age.
J Gastroenterol Hepatol 2005;20:2649.
53 Yang H, Ko J, Seo J. Does the diagnostic accuracy of the 13C-
urea breath test vary with age even after the application of
urea hydrolysis rate? Helicobacter 2008;13:23944.
54 Elitsur Y, Tolia V, Gilger M, et al. Urea breath test in children:
the United States prospective, multicenter study. Helicobacter
2009;14:13440.
55 Delvin E, Brazier J, Deslandres C, Alvarez F, Russo P, Seidman
E. Accuracy of the [13C]-urea breath test in diagnosing Helicob-
acter pylori gastritis in pediatric patients. J Pediatr Gastroenterol
Nutr 1999;28:5962.
56 Levine A, Shevah O, Miloh T, et al. Validation of a novel real
time 13C urea breath test for rapid evaluation of Helicobacter
pylori in children and adolescents. J Pediatr 2004;145:1124.
57 Herold R, Becker M. 13C-urea breath test threshold calculation
and evaluation for the detection of Helicobacter pylori infection
in children. BMC Gastroenterol 2002;2:127.
58 Aas J, Paster B, Stokes L, Olsen I, Dewhirst F. Defining the
normal bacterial flora of the oral cavity. J Clin Microbiol
2005;43:572132.
59 Bik E, Eckburg P, Gill S, et al. Molecular analysis of the
bacterial microbiota in the human stomach. Proc Natl Acad Sci
USA 2006;103:7327.
60 Peng N, Lai K, Liu R, et al. Clinical significance of oral urease
in diagnosis of Helicobacter pylori infection by [13C] urea breath
test. Dig Dis Sci 2001;46:17728.
61 Vaira D, Gatta L, Ricci C, Di M, Lanzini A. Accuracy of urea
breath tests tablets after 10 minutes compared with standard
30 minutes to diagnose and monitoring Helicobacter pylori
infection: a randomized controlled trial. J Clin Gastroenterol
2009;43:6934.
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337

Leal et al.
62 Klein P, Malaty H, Czinn S, Emmons S, Martin R, Graham D.
Normalizing results of 13C-urea breath testing for CO2 produc-
tion rates in children. J Pediatr Gastroenterol Nutr 1999;29:297
301.
63 Rektorschek M, Weeks D, Sachs G, Melchers K. Influence of
pH on metabolism and urease activity of Helicobacter pylori.
Gastroenterology 1998;115:62841.
64 Graham D, Opekun A, Jogi M, et al. False negative urea breath
tests with H2-receptor antagonists: interactions between Heli-
cobacter pylori density and pH. Helicobacter 2004;9:1727.
65 Epple H, Kirstein F, Bojarski C, et al. 13C-urea breath test in
Helicobacter pylori diagnosis and eradication. Correlation to
histology, origin of false results, and influence of food intake.
Scand J Gastroenterol 1997;32:30814.
66 Vakil N. Review article: the cost of diagnosing Helicobacter pylori
infection. Aliment Pharmacol Ther 2001;15:105.
67 Perri F, Pastore M, Clemente R, et al. Helicobacter pylori infec-
tion may undergo spontaneous eradication in children: a 2-year
follow-up study. J Pediatr Gastroenterol Nutr 1998;27:1813.
68 Phillips CV, Goodman KJ. Interpreting data in the face of com-
peting explanations: assessing the hypothesis that observed
2011 Blackwell Publishing Ltd, Helicobacter 16: 327337
13
C-UBT H. pylori Diagnosis in Children
spontaneous clearance of Helicobacter pylori was all measure-
ment error. Int J Epidemiol 2009;38:11107.
69 Leal-Herrera Y, Torres J, Monath T, et al. High rates of recur-
rence and of transient reinfections of Helicobacter pylori in a
population with high prevalence of infection. Am J Gastroenterol
2003;98:2395402.
70 Goodman KJ, ORourke K, Day S, et al. Dynamics of Helicobact-
er pylori infection in a US-Mexico cohort during the first two
years of life. Int J Epidemiol 2005;34:134855.
71 Nurgalieva Z, Opekun A, Graham D. Problem of distinguishing
false-positive tests from acute or transient Helicobacter pylori
infections. Helicobacter 2006;11:6974.
72 Leal YA, Cedillo-Rivera R, Simon JA, Velazquez JR, Flores LL,
Torres J. Utility of stool sample-based tests for the diagnosis of
Helicobacter pylori infection in children. J Pediatr Gastroenterol
Nutr 2011;52:71828.
73 Rutjes A, Reitsma J, Di N, Smidt N, van Rijn J, Bossuyt P.
Evidence of bias and variation in diagnostic accuracy studies.
CMAJ 2006;174:46976.
337