Review Article
Cone-beam Computed Tomography for Detecting
Vertical Root Fractures in Endodontically Treated Teeth:
A Systematic Review
Edwin Chang, DDS, MSc, FRCD(C),* Ernest Lam, DMD, MSc, PhD, FRCD(C),*
Prakesh Shah, MD, DCH, MRCP, FRCP(C),†‡ and Amir Azarpazhooh, DDS, MSc, PhD, FRCD(C)‡§k
Abstract
Introduction: A vertical root fracture (VRF), commonly
found in teeth with endodontic treatment, is challenging
to diagnose and has poor treatment outcomes. Cone-
beam computed tomography (CBCT) has become an
increasingly popular imaging modality in endodontics,
but image artifacts arising from root-filling materials
may hinder VRF detection. The aim of this investigation
was to conduct a systematic review to assess the diag-
nostic ability of CBCT for detecting VRFs in endodonti-
cally treated teeth. Methods: A systematic review of
in vivo clinical diagnostic literature (initial search
December 2014, updated August 2015) was conducted.
Assessment of methodological quality was performed
by using the modified Quality Assessment of Diagnostic
Accuracy Studies tool. Results: Four studies with a total
of 130 patients were included. The reported ranges of
values were 40%–90% for VRF prevalence, 84%
(0.64–0.95) to 100% (0.83–1.00) for sensitivity, 64%
(0.35–0.87) to 100% (0.03–1.00) for specificity, 71%
(0.51–0.87) to 100% (0.63–1.00) for positive predictive
value, and 50% (0.01–0.99) to 100% (0.84–1.00) for
negative predictive value. All 4 studies revealed multiple
items at high risk or unclear risk of bias. Conclusions:
Because of the significant imprecision in the range of
reported estimates and the biases observed in the
included studies, there is currently insufficient evidence
to suggest that CBCT is a reliable test in detecting
VRFs in endodontically treated teeth. (J Endod
2016;42:177–185)
Key Words
Cone beam computed tomography, diagnosis, review
literature as topic, tooth fractures
From the *Discipline of Oral and Maxillofacial Radiology,
Faculty of Dentistry, †Department of Paediatrics, Faculty of
Medicine, ‡Institute of Health Policy, Management and Evalua-
tion, Faculty of Medicine, §Discipline of Dental Public Health,
Faculty of Dentistry, and kDiscipline of Endodontics, Faculty
of Dentistry, University of Toronto, Toronto, Ontario, Canada.
Address requests for reprints to Dr Amir Azarpazhooh, Fac-
ulty of Dentistry, University of Toronto, Room 515-C, 124 Ed-
ward Street, Toronto, PM M5G 1G6, Canada. E-mail address:
amir.azarpazhooh@dentistry.utoronto.ca
0099-2399/$ - see front matter
Copyright ª 2016 American Association of Endodontists.
http://dx.doi.org/10.1016/j.joen.2015.10.005
JOE — Volume 42, Number 2, February 2016
A vertical root fracture (VRF) is defined as a longitudinally oriented fracture plane
that is confined to the tooth root. The prevalence of VRFs is not well-established, but
it is thought that they are more commonly found in teeth that have undergone endodon-
tic treatment (1). Clinical studies of endodontically treated teeth that were extracted
with the presumption of VRF suggest a prevalence of 2%–5% (2–6), whereas studies
that observed VRFs in extracted endodontically treated teeth report a prevalence of
between 11% and 20% (7, 8). Traditional methods used to diagnose VRFs include
transillumination (9), projection radiography (10), bite testing (11), periodontal
probing (3), sinus tract detection (3), and direct visual examination (11). Unfortu-
nately, all of these methods have limited diagnostic reliability because most signs and
symptoms are nonspecific for VRF. Hence, distinguishing VRF from pulpal necrosis,
failed root canal treatment, and/or periodontal disease is often challenging (12). Partic-
ularly in relation to projection radiography, it has been shown that to visualize a VRF, the
primary beam needs to be within 4 of the fracture plane (13). Moreover, the super-
imposition of surrounding anatomic structures makes visualization of a VRF even more
difficult (14). Therefore, it is not surprising that a fracture plane is visualized in only
approximately one third of VRF cases on periapical radiographs (15). Once a diagnosis
is established, the prognosis of a tooth with VRF is poor; there are currently no reliable
methods of treating VRF, and the affected tooth is usually extracted (1). In this context, a
reliable diagnosis of VRF is of utmost importance to prevent unnecessary extraction of
an otherwise treatable tooth.
Cone-beam computed tomography (CBCT) is an imaging modality that uses a
revolving cone-shaped x-ray source projected onto a reciprocating digital flat-panel
receptor. The acquired projectional ‘‘basis’’ images are then used to construct a
three-dimensional volumetric data set that can then be used to reconstruct cross-
sectional images in any plane (16). Compared with conventional multidetector
CT, CBCT units are typically smaller and less expensive and offer higher-
resolution images with lower effective doses to the patient (17). Because of the
limitations of conventional radiography, the application of high-resolution CBCT im-
aging in detecting VRFs has generated considerable interest. Although there are a
number of ex vivo studies that have attempted to assess the ability of CBCT in
detecting artificially induced VRFs in endodontically treated teeth (18–30), these
findings do not incorporate the associated periradicular changes in the osseous
structures adjacent to the fracture plane, which can often aid in the
interpretation of a VRF (31). In our experience, ruling out VRF in a previously
endodontically treated tooth is a common reason for CBCT imaging, yet it is one
of the most difficult to interpret because of the presence of imaging artifacts that
may obscure the putative fracture plane (32). Consequently, the ability of CBCT
to detect VRFs in endodontically treated teeth is an important clinical question to
address. The purpose of this systematic review was to investigate the diagnostic
ability of CBCT in detecting VRFs in endodontically treated teeth. The review format
and methodology strictly adhere to those recommended by the Cochrane Collabora-
tion for diagnostic tests of accuracy as well as users’ guide to studies of diagnosis
(33–35).
CBCT for Detecting VRFs 177
Review Article
Methods
Criteria for Considering Studies for This Review
Types of Studies. Prospective or retrospective clinical diagnostic
studies with a consecutive period of patient enrollment were considered
for inclusion.
Participants. Studies reporting on human subjects with at least 1
endodontically treated permanent tooth suspected of having VRF on
the basis of existing clinical signs and symptoms (ie, sensitivity to per-
cussion, pain on biting and release, and/or localized periradicular bone
loss and deep probing depths) were included. To eliminate VRF cases
with grossly distracted root fragments that do not pose a sufficient diag-
nostic dilemma (33–35), the fracture plane must not have been
visualized on periapical radiographs.
Index Test. The index test was CBCT imaging, regardless of the gen-
eration of development of the instrument, field-of-view size, voxel size,
and exposure parameters such as milliamperes, time, and kilovolt peak.
Target Condition. The target condition was VRFs involving any per-
manent endodontically treated tooth. Fractures that run approximately
horizontal to the long axis of the tooth root were excluded.
Reference Standards. The reference standard was direct obser-
vation of the tooth root surface via orthograde retreatment, exploratory
open-flap surgery, or visual inspection of extracted teeth without or with
the aid of stains. Studies with incomplete reference testing, whereby only
index test positive cases were subjected to reference testing, were
excluded (33–35).
Search Methods for Identification of Studies
We searched through a comprehensive list of electronic databases
and indices as well as other resources to identify potentially relevant
published or unpublished studies, with no language or publication re-
strictions (Table 1). The reference lists of relevant review articles and
included studies were also hand-searched.
Data Collection and Analysis
Two review authors (E.C., A.A.) independently reviewed and
selected relevant studies from the search results and extracted data
TABLE 1. Summary Table of Search Strategy
Search strategy
Electronic databases/indices
MEDLINE (via OVID)
EMBASE (via OVID)
AMED (via OVID)
Ovid Healthstar (via OVID)
Science Citation Index Expanded (via Web of Science)
BIOSIS Citation Index (via Web of Science)
BIOSIS Previews (via Web of Science)
National Technology Information Service (via ProQuest)
LILACS
MEDION
Chinese Biomedical Literature Database (via CNKI)
Turning Research Into Practice (TRIP)
Scopus
Grey literature search
ProQuest Theses and Dissertations (via ProQuest)
System for Information on Grey Literature in Europe (SIGLE)
Conference Proceedings Citation Index (via Web of Science)
COS Conference Papers Index (via ProQuest)
Google Scholar (first 100 hits by using the search terms ‘‘cone beam’’ and ‘‘root fracture’’)
Annual meeting proceedings of the American Association of Endodontists
Annual meeting proceedings of the American Academy of Oral and Maxillofacial Radiology
Annual meeting proceedings of the International Association of Dento-Maxillo-Facial Radiology
Annual meeting proceedings of the International Association of Dental Research
178 Chang et al.
by using a piloted data collection form (Table 2). The modified Quality
Assessment of Diagnostic Accuracy Studies checklist, as recommended
by the Cochrane Handbook for Diagnostic Tests of Accuracy (40), was
applied by the 2 review authors (E.C., A.A.) for the independent assess-
ment of risk of bias in the included studies. Any disagreements at these
steps were resolved through discussion and consensus or by consulting
the other review authors (E.L., P.S.). The software Meta-DiSc 1.4 (41)
was used to calculate descriptive statistics of sensitivity, specificity, and
positive and negative predictive values (PPV and NPV) and to plot data
on the receiver operating characteristic (ROC) plane, which is a plot of
sensitivity as a function of 1 – specificity.
Results
Results of the Search
In total, we identified 2360 records through electronic database
searches. After screening for abstracts, 2337 records were removed.
We retrieved the full-text versions of the remaining 23 records and
excluded 19 articles that did not meet our inclusion criteria
(Table 3). The remaining 4 records were analyzed in the review
(Fig. 1). These studies were clinically significantly heterogeneous, so
a priori decision was made to conduct only systematic review and
not meta-analyses (50).
Findings
The sample sizes of the 4 included studies ranged from 10 (39) to
49 (38). The patients were selected from a wide age range and varied
geographic locations, including the United States (36), Europe (37),
China (38), and Iran (39). The calculated prevalence of VRF (ie, prob-
ability of VRF before CBCT) ranged from 40% (38) to 90% (39). The
index test consisted of CBCT machines of varying models and imaging
parameters, and the interpreters consisted of oral radiologists and/or
endodontists. There was a large variation in the reported agreement
values among the interpreters; 2 of the studies did not report any
intra-rater or inter-rater scores (36, 39), 1 study reported a wide
range of intra-rater and inter-rater agreement scores ranging from
51% to 100% and 25% to 79%, respectively (37), and 1 study reported
very good inter-rater agreement (94%) (38). The reference tests
Time period
1946 to November 2014
1980 to week 47 2014
1985 to November 2014
1966 to October 2014
Inception to December 1, 2014
Inception to December 1, 2014
Inception to December 1, 2014
Inception to December 2, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
Inception to December 8, 2014
December 8, 2014
Inception to 2014
Inception to 2014
Inception to 2014
Inception to 2014
JOE — Volume 42, Number 2, February 2016
 TABLE 2. Summary of Characteristics of the Included Studies
JOE — Volume 42, Number 2, February 2016
Study
Patient characteristics
Endodontically treated
teeth
Patient demographics
Geographic location
VRF prevalence
Study characteristics
Study design
Study setting
Inclusion criteria
Exclusion criteria
Index test
CBCT model (voxel size)
Imaging parameters
Interpreters
Interpreter(s)
Pre-study calibrations?
Blinding?
CBCT for Detecting VRFs
Agreement
179
Edlund et al, 2011 (36) Metska et al, 2012 (37) Wang et al, 2011 (38) Kajan et al, 2012 (39)
32 39 49 10
29 adults (ages 20–70 y) 39 patients (unknown age range) 128 adults (68 men, 60 women, 10 adults (1 man, 9 women, ages
ages 22–82 y, median age, 45) 21–70 y)
Florida Amsterdam, the Netherlands People’s Republic of China Rasht, Iran
75% 64% 40% 90%
Prospective cohort Prospective cohort Retrospective cohort Prospective cohort
Academic institution Academic institution Academic institution Academic institution
Signs and symptoms of VRF (pain on Signs and symptoms of VRF (pain on Signs and symptoms of VRF (pain Signs and symptoms of VRF (pain
percussion, localized bone loss percussion, deep isolated on mastication or percussion, on mastication or percussion,
and deep periodontal pocket) periodontal pocket, sinus tracts, halo periodontal pockets, sinus apical swelling, repetitive
or J-shaped radiolucency on tracts, inconsistent working fistulas, halo-like
periapical radiograph) length determination) radiolucency, apical or lateral
radiolucency, angular crestal
bone loss)
VRFs evident on periapical None declared Cases related to acute trauma VRFs evident on periapical
radiographs or those medically radiographs
unfit for endodontic surgery Cases with teeth for which
surgical intervention was not
attempted
iCAT 3G (125 mm) 3D Accuitomo NewTom 3G (200 mm) 3D Accuitomo 3DX Accuitomo (125 mm) NewTom VG (150 mm)
(80 mm) (80 mm)
0.125 mm voxels 0.08 mm voxels 3.9–5.6 mA, 110 kV, 5 mA, 90 kV, 17.5 s, 5 mA, 80 kV, 0.125 mm 18–20 s, 5.5 mA, 110 kV, 0.15 mm
0.2 mm voxels 0.08 mm voxels voxels
voxels
2 board-certified oral and 3 endodontists with CBCT 2 oral radiologists with >5 years 1 maxillofacial radiologist and 1
maxillofacial radiologists training of clinical experience with endodontist, each with
CBCT >10 years of professional work
experience
Yes Yes Yes, using 10% of the cases. No
These cases were re-used in
the final study.
Interpreters blinded to patient’s Interpreters blinded to patient’s signs Not specified Radiologist performed initial
signs and symptoms and symptoms blinded interpretation,
endodontist (aware of
patient’s symptoms) aided
radiologist in final consensus
interpretation
Not specified Intraobserver: Kappa coefficient between the 2 N/A
NewTom 3G: 0.51–0.83 observers: 0.94
Review Article
3D Accuitomo 170: 0.81–1.00 Between observer A and
Interclass correlation coefficients consensus interpretation: 0.95
(first interpretation, second Between observer B and
interpretation): consensus interpretation: 0.95
NewTom 3G: 0.25, 0.26
3D Accuitomo 170: 0.79, 0.71
(Continued )
Review Article
consisted of surgical exploration, orthograde retreatment, and/or tooth
extraction (Table 2).
The sensitivity values varied from 84% (0.64–0.95) to 100%
Kajan et al, 2012 (39)

(0.83–1.00), and the specificity values varied from 64% (0.35–0.87)

to 100% (0.03–1.00) (Fig. 2). The PPV ranged from 71% (0.51–
0.87) to 100% (0.63–1.00), and NPV ranged from 50% (0.01–
Tooth extraction

0.99) to 100% (0.84–1.00) (Fig. 3). Particularly, the NPV of Kajan

et al (39) was equivocal. By plotting these values in the ROC plane,
it can be observed that the 4 included studies are loosely clustered
in the upper left-hand corner and above the reference line (Figs. 4).
5
1
0
4

Methodological Quality of Included Studies

The risk of bias assessment of the included studies, using QUADAS
checklist, is presented in Supplemental Table S1 (Supplemental Table
Surgical exploration or tooth
Wang et al, 2011 (38)

S1 is available online at www.jendodon.com). Overall, the included

studies demonstrated a high risk of bias, with each study having 3 or
more items at high or unclear risk of bias (Table 4). The most common
items of concern were the following:
1. An unspecified period of time between index testing and refer-
extraction

ence testing: It is uncertain whether the delay between the tests may
have altered the outcome or detection of disease (36, 37, 39).
2. A differential verification bias: It is uncertain whether the index
20
21

test result may have influenced which reference test the patient
8
0

received, potentially leading to a bias in disease confirmation

(37–39).
Orthograde retreatment, endodontic

3. A lack of reporting of test transparency: It is uncertain whether

microsurgery, or tooth extraction

those performing the reference test were blinded to the index test
Metska et al, 2012 (37)

results (36–39) because having prior knowledge of the CBCT

interpretation may have altered the clinician’s ability to surgically
detect a VRF.
4. Lack of sufficient clinical data available to the interpreters: In-
terpreters in 2 studies (36, 37) were blinded to the patient’s signs
and symptoms and preexisting clinical records, which in practice
should be available to the reporting clinician (40).
Despite having the largest sample size (n = 49), the study by Wang
et al (38) demonstrated the highest risk of bias, with 3 items at high risk
21

of bias and 2 items with an unclear risk of bias. Furthermore, because

9
5
4

some patients who did not undergo surgical intervention were excluded
from this study, the disease prevalence may have been inflated. In addi-
tion, 10% of cases that were initially used for calibration in this study
Edlund et al, 2011 (36)

were later reused in the final analysis; this may have led to an overesti-
mation of diagnostic ability.
Surgical exploration

Discussion
This systematic review explored the diagnostic ability of CBCT in de-
tecting VRFs in endodontically treated teeth. Four studies were considered
suitable for inclusion, yet they all demonstrated a high risk of bias. We
initially planned for a meta-analysis by using a multivariate hierarchical
summary ROC curve model as recommended by the Cochrane Collabora-
21
6
2
3

tion (51) to derive pooled estimates of diagnostic accuracy measures.

However, because of the small number of included studies and marked
clinical heterogeneity, a meta-analysis could not be performed.
We identified a significant imprecision in the reported ranges of
TABLE 2. (Continued )

diagnostic ability. This imprecision is a direct reflection of the limited

False negatives
True negatives
Reference test

False positives
Study

True positives

sample sizes used in these clinical studies, the great variability in how
Reference test

Study results

the studies were conducted, the different CBCT models and imaging pa-
rameters used, the different types of reference testing used, and the
potentially different populations under study.
The results of this review should be interpreted with caution.
Because of the relatively subjective nature of radiologic interpretation,

180 Chang et al. JOE — Volume 42, Number 2, February 2016


TABLE 3. Table of Excluded Studies and Reasons for Exclusion
Study
Chen et al, Chinese J Stomatol 2014;49:513–6
Ding et al (42), 2012
Du et al (43), 2010
Liu et al (44), 2013
Liu et al (45), 2010
Ning et al (46), 2011
Qin et al (47), 2011
Xie et al, J Clin Stomatol 2013;29:27–30
Xue et al (48), 2011
Yan et al (49), 2011
Yin et al, Hebei Med 2014;20:95–7
Zhang et al, Chinese J Aesthetic Med 2013;22:22–4
Zhou et al, Stomatology 2013;33:810–2
Youssefzadeh et al, Radiology 1999;210:545–9
Chavda et al, J Endod 2014;40:1524–9
Komatsu et al, Dentomaxillofac Radiol 2014;
43:20140256
Bernardes et al (14), 2009
Bornstein et al, Dent Traumatol 2009;25:571–7
Wang et al, Chin J Dent Res 2010;13:31–5
Zhu et al, J Oral Sci Res 2013;29:620–6
Figure 1. Study flow diagram showing the process of selection of records and
studies for the systematic review. The diagram represents the initial search in
December 2014. An updated search in August 2015 at the time of this manu-
script’s submission did not reveal any additional studies relevant for inclusion.
JOE — Volume 42, Number 2, February 2016
Review Article
Reason for exclusion
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
Did not specify which teeth (if any) were endodontically treated.
Did not specify which teeth (if any) were endodontically treated.
Did not specify which teeth (if any) were endodontically treated.
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
Did not specify which teeth (if any) were endodontically treated.
Did not specify which teeth (if any) were endodontically treated. Also, only
cases tested positive by CBCT were subjected to reference testing.
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
Did not specify which teeth (if any) were endodontically treated.
Did not specify which teeth (if any) were endodontically treated. Also, only
cases tested positive by CBCT were subjected to reference testing.
Collectively examined both endodontically treated and non-
endodontically treated teeth without separate analyses for each type.
The index test was multi-detector CT, not CBCT.
Observers were instructed to ignore changes in the periradicular osseous
structures and only examine the tooth roots.
Observers only analyzed changes to the periradicular osseous structures,
without examining the tooth roots.
Used clinical signs and symptoms as a reference standard, which is not valid.
Did not confirm imaging findings with a reference standard; only compared
detection rate of projectional imaging versus CBCT. Also, analysis
included only horizontal root fractures.
Only cases tested positive by CBCT were subjected to reference testing.
Only cases tested positive by CBCT were subjected to reference testing.
another source of variability is the possibility of positivity threshold ef-
fects that can directly affect the final sensitivity and specificity values
(51). For instance, a more conservative observer who interprets root
fractures as present only if they clearly see a distracted fracture plane
centered within a region of periradicular bone loss may produce higher
specificity and lower sensitivity values as compared with a less conser-
vative observer who records fracture presence despite seeing similarly
appearing imaging artifacts nearby. Because of the range of observers’
training (ie, oral radiologists versus endodontists) and experience, this
may be another significant source of variability. Furthermore, these
findings may not be reflective of other clinicians interpreting these im-
ages, including general practitioners. Because of the limited number of
included studies, an objective analysis of the magnitude of threshold ef-
fects cannot be performed. Moreover, the results presented from the
included studies were based on interpretations from oral radiologists
and/or endodontists with CBCT training. In fact, the reproducibility of
the 4 included studies is also questionable; there was dramatic variation
in the reported agreement scores, and 2 studies did not report any
agreement scores (36, 39).
As a result of multiple sources of bias, the validity of these findings
may also be questionable, with the largest sample size study having the
greatest bias (38). Also, the generalizability of these findings is unclear
because of the different CBCT systems and imaging parameters used be-
tween the different studies. Although there is no clinical evidence to sug-
gest any significant differences in diagnostic ability between these
variables, there is limited in vitro evidence that suggests that different
voxel sizes (100 mm versus 300 mm) may result in differences in VRF
detection, particularly among inexperienced interpreters (52). Last,
because the patient populations of the included studies spanned multi-
ple different geographic regions from around the world, it should be
CBCT for Detecting VRFs 181
Review Article
Figure 2. Summary plots showing the different ranges of reported sensitivity
and specificity values of studies that used CBCT for detecting VRFs in endodon-
tically treated teeth. CI, confidence interval.
cautioned that local differences in diet, chewing habits, and methods of
endodontic/restorative treatment may predispose certain populations
to a VRF (53–56).
Our findings are not in agreement with a 2013 meta-analysis (57)
that evaluated the diagnostic ability of CBCT in detecting any type of root
fracture (vertical and horizontal) in both endodontically and non-
endodontically treated teeth. They reported pooled estimates of sensi-
tivity of 92% (0.89–0.94) and of specificity of 85% (0.75–0.92) and
suggested that CBCT is a highly discriminative test, but its sensitivity
was decreased in the presence of endodontic treatment. It should be
appreciated that although meta-analysis can be a powerful and informa-
tive tool, its results may also be misleading if care is not taken to ensure
that the pooled information is reflective of the clinical question at hand
(58). First, the authors used a univariate model of weighted averaging, a
meta-analysis technique for which a necessary prerequisite is relatively
homogeneous studies with minimum between-study variation and
threshold effects (41). Unfortunately, such an assumption is rarely
true in diagnostic accuracy studies owing to the wide array of different
study designs, variations in index and reference testing, and different
positivity thresholds (50), as we have shown. More statistically rigorous
multivariate models that can accurately account for between-study het-
erogeneity such as threshold effects (50, 59) were not used in the
analysis, a fact that may have ultimately compromised the statistical
validity of their results. Second, the unusually high prevalence values
reported in the Chinese literature may be reflective of differences in
study design and reporting practices, healthcare delivery systems and
referral patterns, and cultures and practices inherent to the Chinese
population. Interestingly, there are a number of published case
reports and case series that suggest that the Chinese may be at a
particularly high risk of developing a VRF in non-endodontically treated
teeth, a phenomenon not reported elsewhere (55, 56, 60–62). This
is thought to be related to specific diets (56) and/or chewing patterns
Figure 3. Summary plots showing the different ranges of reported PPVs and
NPVs of studies that used CBCT for detecting VRFs in endodontically treated
teeth. CI, confidence interval.
182 Chang et al.
Figure 4. The ROC plot of studies that used CBCT for detecting VRFs in
endodontically treated teeth. The graph displays the distribution of the true-
positive rate (sensitivity) as a function of the false-positive rate (1 – specificity)
of the included studies. The dashed diagonal line represents the reference
line along which the diagnostic test becomes non-discriminative.
(55, 62) unique to the Chinese. Because all 8 Chinese language studies
included in the meta-analysis by Long et al (57) examined patients with
non-endodontically treated teeth, the potential difference in reported
root fracture prevalence between the Chinese language publications
(90%–100% reported in 8 studies [42–49]) and the English
language publications (40%–90% reported in the remaining 4
studies [36–39]) may have introduced further heterogeneity in the
meta-analysis that was unaccounted for. Taken together, although the
pooled results were derived from 12 studies, these studies differed
significantly in their patient spectra (Chinese patients versus non-
Chinese patients), target conditions (horizontal versus vertical root
fractures), and covariates that may significantly alter diagnostic ability
(endodontically treated versus non-treated teeth). Therefore, the
appropriateness and clinical significance of statistically pooling data
from such a heterogeneous set of studies are questionable.
We are in agreement with 2 recent systematic reviews. Corbella
et al (63) attempted to assess diagnostic ability of CBCT in VRF detection
in both endodontically treated and non-endodontically treated teeth by
separately analyzing both in vivo and ex vivo studies and concluded
that because of a very limited number of studies and significant hetero-
geneity in study characteristics and reported outcomes, there is
currently no evidence to suggest that CBCT testing can provide any addi-
tional diagnostic benefit to VRF detection in teeth with endodontic treat-
ment. Rosen et al (64) examined the diagnostic efficacy hierarchy of
CBCT in endodontics and concluded that there is currently insufficient
evidence to support that CBCT usage is diagnostically efficacious, mean-
ing that its potential impact in improving patients’ outcomes and
reducing the cost-benefit ratio to society is questionable.
Within the past decade, CBCT has quickly evolved to become an
imaging modality that is increasingly used in the clinical practice of
dentistry (16). Particularly within the specialty of endodontics, signifi-
cant increases in CBCT usage have been reported on the basis of a
Web-based survey of active members of the American Association of
Endodontists in the United States and Canada; 34.2% of respondents
said they used CBCT, with the diagnosis of pathosis and preoperative
JOE — Volume 42, Number 2, February 2016
 Review Article
TABLE 4. Risk of Bias Assessment Summary Table
Modified Quality Assessment of Diagnostic Accuracy Studies checklist
Edlund et al, Metska et al, Wang et al, Kajan et al,
2011 (36) 2012 (37) 2011 (38) 2012 (39)
Representative spectrum? Yes Yes No Yes
Acceptable reference standard? Yes Yes Yes Yes
Acceptable delay between tests? Unclear Unclear Yes Unclear
Partial verification avoided? Yes Yes Yes Yes
Differential verification avoided? Yes Unclear No No
Incorporation bias avoided? Yes Yes Yes Yes
Index test results blinded during reference testing? Unclear Unclear Unclear Unclear
Reference standard results blinded during index testing? Yes Yes No Yes
Relevant clinical information provided? No No Unclear Yes
Uninterpretable results reported? Yes Yes Yes Yes
Withdrawals explained? Yes Yes Yes Yes

assessment being cited as the most common reasons for imaging (65).
Because of the increased uptake of CBCT, the possibility of indiscrimi-
nant prescription and unjustified reliance on a test of unclear diagnostic
ability is concerning. As with all other imaging modalities involving
ionizing radiation, the principles of as low as reasonably achievable
must be upheld when ordering CBCT images, because it is currently
accepted that any exposure to ionizing radiation could lead to poten-
tially deleterious stochastic effects (66).
Moreover, the financial burden to the patient and/or the health-
care system must also be considered. To our knowledge, a small
(<8-cm diameter) field-of-view CBCT scan in Ontario, Canada currently
costs at least $125 CAD, whereas the suggested fee for conventional
periapical radiographs according to the Ontario Dental Association is
significantly less ($20 CAD) (67). Considering that CBCT is more costly
and potentially more harmful as compared with periapical radiography
(63), the decision to order a CBCT, as jointly outlined by the American
Academy of Oral and Maxillofacial Radiology and the American Associ-
ation of Endodontists, must only be considered ‘‘if clinical examination
and 2D intraoral radiography are inconclusive in the detection of ver-
tical root fracture’’ (68). Moreover, as per European guidelines out-
lined by the SEDENTEXCT project, clinicians must keep in mind that
even if CBCT imaging is undertaken, ‘‘the weight of evidence suggests
that root fillings and posts [will] limit diagnostic accuracy’’ (69).
Implications for Practice
For a patient who has undergone CBCT testing, the most clin-
ically relevant question is ‘‘what is the probability that the interpre-
tation of presence or absence of VRF is correct (ie, PPV or NPV,
respectively)?’’ To answer this, the clinician must first understand
the relationship between disease prevalence (ie, pre-test probability)
and the PPV (ie, post-test probability). When disease prevalence is
high, PPV is also high. But in a population with low prevalence, the
PPV also decreases correspondingly. In fact, the lower the sensitivity
and specificity of a test, the stronger the relationship between prev-
alence and predictive values (70). Therefore, disease prevalence in
a population must first be elucidated, and this value is highly depen-
dent on the clinical pathway the patient had undergone before
testing. To maximize the chances of detecting a VRF by using
CBCT, the clinician’s best tools still consist of what is done before
the CBCT (ie, a thorough clinical examination and recognizing the
signs and symptoms that are suggestive of a VRF). As such, the cur-
rent literature must be interpreted with caution because they could
possibly mislead practitioners toward a non–evidence-based uptake
of CBCT for purposes of VRF detection. To further illustrate the
complexity of diagnostic accuracy testing and its strong dependence
on disease prevalence, refer to the clinical scenario defined
in Figure 5.

Figure 5. Clinical scenarios. NPV, negative predictive value, the probability of a negative test result being truly negative; PPV, positive predictive value, the prob-
ability of a positive test result being truly positive; VRF, vertical root fracture. *Assuming that CBCT examination is performed and interpreted by an experienced
clinician. †Estimated from the Edlund et al study (41). ‡The most conservative estimates, that is, the lowest end of the reported ranges identified in our review.
sensitivityprevalence specificityð1prevalenceÞ
§
PPV ¼ sensitivityprevalence þ ð1specificityÞð1prevalenceÞ: NPV ¼ ð1sensitivityÞprevalence þ specificityð1prevalenceÞ: Reported in (1–8).
¶ #

JOE — Volume 42, Number 2, February 2016 CBCT for Detecting VRFs 183
Review Article
Implications for Research
On the basis of the findings of this review, more clinical in vivo
studies of higher methodological quality and better reporting standards
are needed before a more definitive assessment of the diagnostic ability
of CBCT in VRF detection can be made. Potential investigators are
advised to carefully review the criteria used to assess methodological
quality by the Cochrane Collaboration (35) and other users’ guides
to diagnosis (33, 34) before conducting a new study to avoid
methodological flaws that may compromise the validity of their
results. Also, to ensure that the study methodology and findings can
be rigorously evaluated, investigators are further encouraged to
adhere to the stringent reporting criteria as laid out by the STARD
initiative (71).
Conclusion
With only 4 included studies and the relative imprecision of the
estimates in each study, it was difficult to draw any definitive conclusions
regarding diagnostic ability. Because VRFs are the most difficult to treat
and also the most challenging to diagnose in the presence of root-filling
materials, the most clinically relevant question, which is how accurate is
CBCT in detecting VRFs in endodontically treated teeth, remained unan-
swered. Until more evidence is presented to suggest that CBCT is both
diagnostically accurate and efficacious, the prudent clinician should
carefully consider its potential risks and harms before its prescription.
Acknowledgments
The authors thank Dr Joseph Beyene for his advice on the
appropriateness of statistical meta-analysis.
The authors deny any conflicts of interest related to this study.
Supplementary Material
Supplementary material associated with this article can be
found in the online version at www.jendedon.com (http://dx.doi.
org/10.1016/j.joen.2015.10.005).
References
1. Tamse A. Vertical root fractures in endodontically treated teeth: diagnostic signs and
clinical management. Endod Topics 2006;13:84–94.
2. Torbj€orner A, Karlsson S, Odman PA. Survival rate and failure characteristics for two
post designs. J Prosthet Dent 1995;73:439–44.
3. Testori T, Badino M, Castagnola M. Vertical root fractures in endodontically treated
teeth: a clinical survey of 36 cases. J Endod 1993;19:87–91.
4. Moule AJ, Kahler B. Diagnosis and management of teeth with vertical root fractures.
Aust Dent J 1999;44:75–87.
5. Morfis AS. Vertical root fractures. Oral Surg Oral Med Oral Pathol 1990;69:631–5.
6. Bergman B, Lundquist P, Sj€ogren U, et al. Restorative and endodontic results after
treatment with cast posts and cores. J Prosthet Dent 1989;61:10–5.
7. Fuss Z, Lustig J, Tamse A. Prevalence of vertical root fractures in extracted endodon-
tically treated teeth. Int Endod J 1999;32:283–6.
8. Coppens CRM, DeMoor RJG. Prevalence of vertical root fractures in extracted
endodontically treated teeth. Int Endod J 2003;36:926.
9. Bender IB, Freedland JB. Adult root fracture. J Am Dent Assoc 1983;107:413–9.
10. Meister F, Lommel TJ, Gerstein H. Diagnosis and possible causes of vertical root
fractures. Oral Surg Oral Med Oral Pathol 1980;49:243–53.
11. Seo DG, Yi YA, Shin SJ, et al. Analysis of factors associated with cracked teeth.
J Endod 2012;38:288–92.
12. Khasnis SA, Kidiyoor KH, Patil AB, et al. Vertical root fractures and their manage-
ment. J Conserv Dent 2014;17:103–10.
13. Kositbowornchai S, Nuansakul R, Sikram S, et al. Root fracture detection: a compar-
ison of direct digital radiography with conventional radiography. Dentomaxillofac
Radiol 2001;30:106–9.
14. Bernardes RA, de Moraes IG, Hungaro Duarte MA, et al. Use of cone-beam volu-
metric tomography in the diagnosis of root fractures. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 2009;108:270–7.
184 Chang et al.
15. Rud J, Omnell KA. Root fractures due to corrosion: diagnostic aspects. Scand J Dent
Res 1970;78:397–403.
16. De Vos W, Casselman J, Swennen GR. Cone-beam computerized tomography (CBCT)
imaging of the oral and maxillofacial region: a systematic review of the literature. Int
J Oral Maxillofac Surg 2009;38:609–25.
17. Angelopoulos C, Scarfe WC, Farman AG. A comparison of maxillofacial CBCT and
medical CT. Atlas Oral Maxillofac Surg Clin North Am 2012;20:1–17.
18. Tsesis I, Kamburoglu K, Katz A, et al. Comparison of digital with conventional radi-
ography in detection of vertical root fractures in endodontically treated maxillary
premolars: an ex vivo study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
2008;106:124–8.
19. Bechara B, McMahan CA, Nasseh I, et al. Number of basis images effect on detection
of root fractures in endodontically treated teeth using a cone beam computed to-
mography machine: an in vitro study. Oral Surg Oral Med Oral Pathol Oral Radiol
2013;115:676–81.
20. Ferreira RI, Bahrami G, Isidor F, et al. Detection of vertical root fractures by cone-
beam computerized tomography in endodontically treated teeth with fiber-resin and
titanium posts: an in vitro study. Oral Surg Oral Med Oral Pathol Oral Radiol 2013;
115:49–57.
21. Bechara B, McMahan C, Noujeim M, et al. Comparison of cone beam CT scans with
enhanced photostimulated phosphor plate images in the detection of root fracture
of endodontically treated teeth. Dentomaxillofac Radiol 2013;42:20120404.
22. Ozer SY. Detection of vertical root fractures of different thicknesses in endodonti-
cally enlarged teeth by cone beam computed tomography versus digital radiography.
J Endod 2010;36:1245–9.
23. Melo SL, Haiter-Neto F, Correa LR, et al. Comparative diagnostic yield of cone beam
CT reconstruction using various software programs on the detection of vertical root
fractures. Dentomaxillofac Radiol 2013;42:20120459.
24. Bechara B, Alex McMahan C, Moore WS, et al. Cone beam CT scans with and without
artefact reduction in root fracture detection of endodontically treated teeth. Dento-
maxillofac Radiol 2013;42:20120245.
25. Da Silveira PF, Vizzotto MB, Liedke GS, et al. Detection of vertical root fractures by
conventional radiographic examination and cone beam computed tomography: an
in vitro analysis. Dent Traumatol 2013;29:41–6.
26. Melo SL, Bortoluzzi EA, Abreu M, et al. Diagnostic ability of a cone-beam computed
tomography scan to assess longitudinal root fractures in prosthetically treated teeth.
J Endod 2010;36:1879–82.
27. Hassan B, Metska ME, Ozok AR, et al. Comparison of five cone beam computed to-
mography systems for the detection of vertical root fractures. J Endod 2010;36:
126–9.
28. Hassan B, Metska ME, Ozok AR, et al. Detection of vertical root fractures in
endodontically treated teeth by a cone beam computed tomography scan.
J Endod 2009;35:719–22.
29. Khedmat S, Rouhi N, Drage N, et al. Evaluation of three imaging techniques for the
detection of vertical root fractures in the absence and presence of gutta-percha root
fillings. Int Endod J 2012;45:1004–9.
30. Junqueira RB, Verner FS, Campos CN, et al. Detection of vertical root fractures in the
presence of intracanal metallic post: a comparison between periapical radiography
and cone-beam computed tomography. J Endod 2013;39:1620–4.
31. Kishen A, Messer HH. Vertical root fractures: radiological diagnosis. In: Basrani B,
ed. Endodontic Radiology, 2nd ed. Ames, IA: John Wiley & Sons; 2012:235–50.
32. Schulze R, Heil U, Gross D, et al. Artefacts in CBCT: a review. Dentomaxillofac Radiol
2011;40:265–73.
33. Guyatt G, Rennie D, Meade MO, et al. Diagnosis. In: Guyatt G, Rennie D, Meade MO,
Cook DJ, eds. Users’ Guides to the Medical Literature: Essentials of Evidence-
Based Clinical Practice, 2nd ed. Chicago: McGraw-Hill Professional; 2008:
397–505.
34. Brignardello-Petersen R, Carrasco-Labra A, Glick M, et al. A practical approach to
evidence-based dentistry: V—how to appraise and use an article about diagnosis.
J Am Dent Assoc 2015;146:184–91.
35. Website of the diagnostic test accuracy working group of the Cochrane
Collaboration. Available at: http://srdta.cochrane.org. Accessed September 22,
2015.
36. Edlund M, Nair MK, Nair UP. Detection of vertical root fractures by using cone-beam
computed tomography: a clinical study. J Endod 2011;37:768–72.
37. Metska ME, Aartman IH, Wesselink PR, et al. Detection of vertical root fractures
in vivo in endodontically treated teeth by cone-beam computed tomography scans.
J Endod 2012;38:1344–7.
38. Wang P, Yan XB, Lui DG, et al. Detection of dental root fractures by using cone-beam
computed tomography. Dentomaxillofac Radiol 2011;40:290–8.
39. Kajan ZD, Taromsari M. Value of cone beam CT in detection of dental root fractures.
Dentomaxillofac Radiol 2012;41:3–10.
40. Reitsma JB, Rutjes AW, Whiting P, et al. Chapter 9: Assessing methodological quality.
In: Deeks JJ, Bossuyt PM, Gatsonis C, eds. Cochrane Handbook for Systematic Re-
views of Diagnostic Test Accuracy, version 1.0. The Cochrane Collaboration;
JOE — Volume 42, Number 2, February 2016

2009. Available at: http://dta.cochrane.org/handbook-dta-reviews. Accessed
November 19, 2015.
41. Zamora J, Abraira V, Muriel A, et al. Meta-DiSc: a software for meta-analysis of test
accuracy data. BMC Med Res Methodol 2006;6:31.
42. Ding S. Diagnostic value of cone-beam CT for root fractures. J Mudanjiang Med Univ
2012;33:52–4.
43. Du Y, Tang KL, Yu XJ, et al. Cone-beam CT application in the diagnosis of tooth lon-
gitudinal fractures. Chin J Conserv Dent 2010;20:450–3.
44. Liu J, Zhang H, Wu J, et al. Comparison of diagnostic value of cone beam computed
tomography and orthpantomography for root fracture. Int J Stomatol 2013;40:
117–9.
45. Liu R, Gao Y, Shi X, et al. Application of cone beam computed tomography (CBCT) in
diagnosis and treatment of dental root fracture. Med J Chin PLA 2010;35:1484–5.
46. Ning F, Yu Q, Cao D. Application of cone-beam computed tomography in the
early diagnosis of vertical root fracture. Chin Comput Medicine Imag 2011;17:
506–8.
47. Qin H, Zhang C, Yan B, et al. The ability of cone-beam CT in the diagnosis of vertical
tooth fractures. J Zhengzhou Univ Med Sci 2011;46:933–5.
48. Xue Y, Ni LX, Wang HG, et al. A comparison of cone-beam computed tomography
with periapical radiography in detection of vertical root fracture. Chin J Conserv
Dent 2011;21:233–6.
49. Yan X, Peng H, Liu L, et al. Application of conebeam computed tomography in the
early diagnosis of vertical root fracture. J Pract Stomatol 2011;27:46–9.
50. Rutter C, Gatsonis C. A hierarchical regression approach to meta-analysis of diag-
nostic test accuracy evaluations. Stat Med 2001;2884:2865–84.
51. Macaskill P, Gatsonis C, Deeks JJ, et al. Chapter 10: Analysing and presenting results.
In: Deeks JJ, Bossuyt PM, Gatsonis C, eds. Cochrane Handbook for Systematic Re-
views of Diagnostic Test Accuracy, version 1.0. The Cochrane Collaboration;
2010. Available at: http://dta.cochrane.org/handbook-dta-reviews. Accessed
November 19, 2015.
52. Amintavakoli N. The Effect of Cone Beam CT Voxel Size on the Identification of
Vertical and Horizontal Root Fractures: An In-Vitro Study (MSc thesis). Toronto,
Canada: University of Toronto; 2013.
53. Fuss Z, Lustig J, Katz A, et al. An evaluation of endodontically treated vertical root
fractured teeth: impact of operative procedures. J Endod 2001;27:46–8.
54. Tang W, Wu Y, Smales RJ. Identifying and reducing risks for potential fractures in
endodontically treated teeth. J Endod 2010;36:609–17.
55. Yeh CJ. Fatigue root fracture: a spontaneous root fracture in non-endodontically
treated teeth. Br Dent J 1997;182:261–6.
JOE — Volume 42, Number 2, February 2016
Review Article
56. Yang SF, Rivera EM, Walton RE. Vertical root fracture in nonendodontically treated
teeth. J Endod 1995;21:337–9.
57. Long H, Zhou Y, Ye N, et al. Diagnostic accuracy of CBCT for tooth fractures: a meta-
analysis. J Dent 2014;42:240–8.
58. Da Costa BR, Juni P. Systematic reviews and meta-analyses of randomized trials:
principles and pitfalls. Eur Heart J 2014;35:3336–45.
59. Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity
produces informative summary measures in diagnostic reviews. J Clin Epidemiol
2005;58:982–90.
60. Chan CP, Lin CP, Tseng SC, et al. Vertical root fracture in endodontically versus non-
endodontically treated teeth: a survey of 315 cases in Chinese patients. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1999;87:504–7.
61. Chan CP, Tseng SC, Lin CP, et al. Vertical root fracture in nonendodontically
treated teeth: a clinical report of 64 cases in Chinese patients. J Endod 1998;
24:678–81.
62. Gao YJ, Yin XM, Wu HJ. Relationship between vertical root fracture and the habits of
chewing betel nut. Hunan Yi Ke Da Xue Xue Bao 2001;26:161–2.
63. Corbella S, Del Fabbro M, Tamse A, et al. Cone beam computed tomography for the
diagnosis of vertical root fractures: a systematic review of the literature and meta-
analysis. Oral Surg Oral Med Oral Pathol Oral Radiol 2014;118:593–602.
64. Rosen E, Taschieri S, Del Fabbro M, et al. The diagnostic efficacy of cone-beam
computed tomography in endodontics: a systematic review and analysis by a hier-
archical model of efficacy. J Endod 2015;41:1008–14.
65. Dailey B, Mines P, Anderson A, et al. The use of cone beam computer tomography in
endodontics: results of a questionnaire—2010: AAE Annual Session abstract pre-
sentation PR10. J Endod 2010;36:567.
66. Preston RJ. Update on linear non-threshold dose-response model and implications
for diagnostic radiology procedures. Health Phys 2008;95:541–6.
67. ODA Suggested Fee Guide for General Practitioners. Toronto, Canada: Ontario
Dental Association; 2015.
68. AAE, AAOMR. AAE and AAOMR joint position statement: use of cone beam computed
tomography in endodontics 2015 update. J Endod 2015;41:1393–6.
69. European Commission. Radiation Protection No 172 Cone Beam CT for Dental
and Maxillofacial Radiology (Evidence-based Guidelines). Luxembourg: The
SEDENTEXCT Project; 2012.
70. Akobeng AK. Understanding diagnostic tests 1: sensitivity, specificity and predictive
values. Acta Paediatr 2007;96:338–41.
71. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting
of studies of diagnostic accuracy: the STARD initiative. BMJ 2003;326:41–4.
CBCT for Detecting VRFs 185