Original Contributions

Impact of visual inspection and radiographs

for caries detection in children through a
2-year randomized clinical trial
The Caries Detection in Childrene1 study
Laura R.A. Pontes, DDS, PhD; Tatiane F. Novaes, MS, PhD; Juan S. Lara, MS, PhD;
Thais Gimenez, PhD; Bruna L.P. Moro, MS; Lucila B. Camargo, MS, PhD;
Edgard Michel-Crosato, MS, PhD; Claudio M. Pannuti, MS, PhD; Daniela P. Raggio, MS, PhD;
Mariana M. Braga, PhD; Fausto M. Mendes, MS, PhD

ABSTRACT

Background. Visual inspection (VIS) with radiographic examination (RAD) is the recommended
diagnostic strategy for detecting caries in children; however, this recommendation is based on
accuracy studies. The authors conducted a clinical trial to compare the detection and subsequent
treatment of carious lesions in primary molars performed with VIS alone and with RAD.
Methods. Children (3-6 years old) were randomly assigned to 2 groups according to the diagnostic
strategy used for caries detection on primary molars: VIS or RAD. Participants were diagnosed and treated
according to the management plan related to the allocated group. The primary outcome was the number of
new operative interventions during the 2-year follow-up period. Other secondary outcomes were also
compared. Comparisons were performed with Mann-Whitney test using an intention-to-treat approach.
Results. Of the 252 children included and randomized, 216 were followed-up after 2 years (14.3%
attrition rate). There was no difference between the groups for the primary outcome (P ¼ .476). For
the secondary outcomes, the RAD group had more restoration replacements (P ¼ .038) and more
restorations performed since the beginning of the study (P ¼ .038) compared with the VIS group. In
addition, the RAD group had a higher number of false-positive results than the VIS group (P < .001).
Conclusions. Simultaneous use of VIS and RAD for caries diagnosis in primary molars of children
who seek dental treatment does not provide additional benefits compared with VIS alone.
Practical Implications. Dentists should perform VIS only, not RAD, for detecting carious lesions
in preschool-aged children.
ClinicalTrials.gov: NCT02078453.
Key Words. Clinical trial; dental caries; radiography; primary teeth; diagnosis; children.
JADA 2020:151(6):407-415
https://doi.org/10.1016/j.adaj.2020.02.008

T
he best diagnostic strategy for detection of carious lesions in children in daily clinical practice
remains controversial. Although visual inspection (VIS) is an essential, inexpensive, and
simple method,1 clinical guidelines have recommended obtaining bite-wing radiographs
as an adjunct method for caries detection in all children.2-5 This relies on the belief that VIS
overlooks some lesions,6 and that radiographic examination7 (RAD) in conjunction with VIS8
This article has an
provides higher sensitivity. However, this increase in sensitivity occurs at the cost of decreasing
accompanying online
specificity.8,9 continuing education activity
For caries, false-positive results are undesirable because the prevalence of nonevident carious available at:
lesions in primary molars is low.8,10 In addition, false-positive results can lead to unnecessary http://jada.ada.org/ce/home.
treatment, and nondetected carious lesions could be detected later on, causing no harm. Although
Copyright ª 2020
researchers from previous accuracy studies have observed these trends,8-10 they only consider right
American Dental
and wrong results in relation to a reference standard, and the long-term benefits for patients’ oral Association. All rights
health are not considered in this type of study. reserved.

JADA 151(6) n http://jada.ada.org n June 2020 407

 A diagnostic strategy should be appraised on the basis of patient health outcomes subsequent to
diagnostic results and related treatments.11-13 Randomized clinical trials (RCTs) are the most
appropriate design because they randomly allocate participants to 2 or more diagnostic strategies
and, subsequently, treatments are chosen on the basis of the results obtained for each strategy.
Afterward, relevant outcomes for the patients’ health can be evaluated.11,12 Many RCTs evaluating
diagnostic strategies for several medical disorders have been conducted,14 but, to our knowledge, no
studies related to dental conditions have been carried out.
In the absence of this type of study on dental conditions and facing the challenge of choosing the
best strategy for detection of carious lesion in children, we designed this RCT to compare detection
strategies of, and subsequent treatments for, carious lesions in primary molars performed by means of
VIS alone and in conjunction with RAD in children seeking dental treatment. Outcomes related to
children’s oral health after 2 years of follow-up were considered. This is the first of a series of
RCTsdthe Caries Detection in Children trialsdthat are the result of a pioneering initiative that
intends to test different diagnostic strategies related to caries in children.

METHODS
Trial design
This article was written in accordance with the Consolidated Standards of Reporting Trials statement.15
This study was designed as a 2-arm, randomized, triple-blind, parallel-design trial with 2 years of follow-
up. The study protocol was approved by the Research Ethics Committee of the School of Dentistry,
University of São Paulo (Certificate of Presentation for Ethical Consideration 02952612.4.0000.0075),
registered with ClinicalTrials.gov (NCT02078453), and published previously.16

Participants
Eligible participants were 3- to 6-year-old children who sought treatment at our dental school.
Children with challenging behavior during their first appointment or whose guardians did not
consent to their participation were excluded.
Once a participant was included, an initial clinical examination was performed to evaluate
present teeth and caries experience.17 A pair of bite-wing and periapical radiographs (when
necessary) were obtained in all participants before randomization. A researcher (L.R.A.P.) who did
not take part in other phases of the study was responsible for both the initial examinations and
obtaining radiographs. Children were classified according to age (3-4 or 5-6 years old), caries
experience, and decayed, missed, and filled primary surfaces ( 3 or > 3). All caries diagnostic
procedures and treatments were conducted in a dental office setting.

Interventions
Tested interventions were strictly related to the caries detection strategy in primary molars corre-
sponding to the allocated group. Children were randomized to 1 of the groups, which was revealed
to examiners only before performing the caries detection procedures. One of 2 trained and cali-
brated examiners (T.F.N., J.S.L.) performed the dental examinations in accordance with the
following:
n VIS group: Children were assessed using VIS alone. Caries assessment was performed according to
the International Caries Detection and Assessment System (ICDAS)18 associated with caries
activity assessment.19 Subsequently, the treatment plan was created. No specific treatment was
determined for primary molar surfaces classified as sound (ICDAS score 0). The same decision was
ABBREVIATION KEY applied for inactive carious lesions. For active carious lesions with ICDAS scores of 1 through 3, a
dmf-s: Decayed, missing, or decision for nonoperative treatment was reached. Operative treatment was indicated for carious
filled surfaces in lesions with an ICDAS score of 4 through 6.
primary teeth index. n RAD group: Children were examined using VIS and RAD. Examiners conducted VIS as
ICDAS: International Caries
Detection and
described, and they could obtain bite-wing radiographs. A treatment plan was then created
Assessment System. considering these methods used jointly. Because a simultaneous strategy was used, a positive result
NA: Not applicable. from either method would be sufficient to classify the surface as carious. No treatment was
RAD: Radiographic indicated when there were no radiolucencies on radiographic images. Dental surfaces with
examination.
radiolucency restricted to enamel were indicated for nonoperative treatment, and surfaces with
RCT: Randomized clinical
trial. radiolucency reaching the dentin were indicated for operative treatment. In case of discordance
VIS: Visual inspection. between methods, the most severe classification was considered for treatment decision.

408 JADA 151(6) n http://jada.ada.org n June 2020

 Treatment plans designed according to the allocated group were stored in sealed envelopes. At
treatment appointments, dental practitioners, unaware of the child’s group, received the treatment
plan with no access to clinical examination records or bite-wing radiographs. Clinicians were pe-
diatric dentistry residents from our school who were trained to perform dental procedures in chil-
dren. Each clinician performed all treatments in the same child, who could be from both groups.
Treatments were performed strictly according to the treatment plan in the envelopes. Nonop-
erative and operative procedures were carried out following protocols described previously for both
groups.16

Follow-up examinations and outcomes

After completing the treatment phase, participants were scheduled for a follow-up examination
every 6 months until 24 months after their last baseline treatment appointment. Children’s
guardians were instructed to contact the research team if they noticed any treatment need.
At recall visits, children received oral hygiene orientation advising the use of fluoride dentifrice
(1,000-1,500 parts per million flouride) and dietary habits counseling. Fluoride varnish was applied
in children with active carious lesions. A trained and calibrated outcome appraiser (D.P.R.) who
was blinded to the allocated groups performed examinations to assess outcome variables.
The primary outcome was the number of primary molar surfaces with operative treatment need
during follow-up: a composite end point including the number of primary molars surfaces with new
carious lesions requiring restorative treatment (carious lesions with evident dentin involvement);
restorations performed at baseline that needed replacement (large failures, caries around restora-
tions, and complete loss of material); and teeth (5 surfaces) with need for endodontic treatment; or
extraction. The outcomes were assessed by means of VIS and periapical radiographs were obtained
only to confirm the need for endodontic treatment or extraction.
The number of primary molar surfaces with new carious lesions and the number of restored
surfaces needing replacement were considered as secondary outcomes. Other secondary outcomes
were the need for small repairs on restorations, number of cavitated carious lesions on permanent
first molars, and number of primary molars submitted for endodontic treatment or extraction. Re-
ports of pain episodes were also recorded. All outcomes were specified previously in the published
study protocol.16
A secondary outcome not considered in the initial protocol was included posteriorly for analysis.
This was the total number of restorative procedures on primary molars, including restorations placed
at baseline and during follow-up, as well as replacement restorations.
Additional secondary outcomes related to oral healtherelated quality of life and economic
analysis proposed in the initial protocol will be explored further owing to their particular nature.

Sample size
Sample size calculation was based on the primary outcome. For this, a mean (standard deviation
[SD]) of 19 (15) surfaces with treatment need was estimated for the group receiving VIS alone, and
a difference (SD) of 5 (10) surfaces as the minimal clinically important difference for the RAD
group. Considering a 2-tailed test, 5% level of significance, and 80% power and anticipating an
attrition rate of 20%, the final sample size was 250 children (125 in each group). Details of the
sample size calculation can be found elsewhere.16

Randomization
Participants were randomly assigned to VIS or VIS and RAD groups with 1:1 allocation rate. The
random allocation sequence was generated using the Web site www.sealedenvelopes.com.
Randomization was stratified by means of age and caries experience in blocks of 8.
The generated sequences were put into opaque envelopes numbered sequentially according to
children’s strata. Envelopes were opened sequentially and once each child was seated in the dental
chair and had received dental prophylaxis before dental examination. Bite-wing radiographs were
only provided to the examiner for children allocated to the RAD group.

Blinding
This was a triple-blind study. As bite-wing radiographs were obtained for all participants before
randomization, participants and their parents were blinded to the allocated group. Oral health care

JADA 151(6) n http://jada.ada.org n June 2020 409

 providers performing dental treatments were also blinded, as they did not know which diagnostic
strategy had been used for treatment planning. The outcome assessor was also blinded to the
allocation group. The researcher responsible for enrolling participants in the study and the exam-
iners who conducted the baseline dental examinations and treatment plans were not blinded.

Statistical methods
All quantitative variables were submitted to D’Agostino-Pearson and Levene tests to check
normality and homogeneity of variances, respectively. As these assumptions were not reached,
statistical comparisons were conducted using nonparametric tests, and data were presented as mean
(SD) and median (interquartile range).
Intention-to-treat analyses were used. Missing data from participants who dropped out were
handled by means of conditional multiple imputation considering children’s age, caries experience,
data of the specific trial group, and data collected in previous follow-up recall visits (when avail-
able). Therefore, all analyses considered all participants who were randomized for the trial groups.
Mann-Whitney tests were conducted comparing groups in relation to variables collected after the
end of the initial treatment as follows: number of primary molar surfaces that did not receive any
type of treatment, surfaces submitted for nonoperative treatment, surfaces that were restored at the
beginning of the study, number of false-positive results, number of primary molars submitted for
endodontic treatment, or extracted molars. False-positive results were considered for surfaces
assigned to operative treatment but that did not have dentin carious tissue after opening.
The primary outcome assessed after 2 years of follow-up was compared between the groups using
Mann-Whitney test. The same statistical procedure was used for secondary outcomes. The pro-
portion of children in each group reporting pain within the follow-up period was compared using c2
test. A secondary analysis, stratified by means of children’s caries experience and considering
proximal surfaces only, was conducted.
We used Stata, Version 13.0 (Stata) for multiple imputation procedures and MedCalc, Version
18.6 (Medcalc Software) for other analyses. The level of significance was set at 5%.

RESULTS
Of 252 children randomized, 216 were followed until the end of the study (follow-up rate, 85.7%);
106 from the VIS group (82.8%) and 110 from the RAD group (88.7%), with no difference be-
tween groups (Table 1). Three children did not receive interventions because they did not attend
subsequent appointments. No children crossed over to the other group during the trial. For the
subsequent analyses, an intention-to-treat approach was used, and missing data were imputed. The
study flowchart is presented in the figure.
Children were included from March 2014 through November 2015, and all treatments were
completed by February 2016. The recall visit phase was completed in January 2018. Participants
from both groups were similar for all variables at baseline (Table 1).
Regarding initial treatments performed according to both diagnostic strategies, children allocated
to the VIS group had significantly more primary molar surfaces that did not receive any type of
treatment (Table 2). Conversely, children allocated to the RAD group had 10 times more surfaces
with false-positive results compared with children submitted for VIS alone. Similarities between
groups were observed for other variables (Table 2).
Regarding the primary outcome, although mean values were similar for both groups, the median
value indicated that children allocated to VIS alone tended to have fewer new interventions;
however, no statistically significant difference was observed (Table 3).
No differences between groups were found for number of surfaces with new carious lesions,
number of surfaces with restoration repairs, carious lesions in the permanent first molars, teeth with
endodontic treatment, or extractions within the follow-up period (Table 3). In contrast, children
allocated to the RAD group had a higher need for restoration replacement and significantly more
restored surfaces since the beginning of the study and during the follow-up period than children
diagnosed by means of VIS alone. These differences, however, were small, varying from 1 through 2
surfaces (Table 3).
No significant differences were found after subgroup analysis stratified by means of children’s
caries experience. The exception was that children in the RAD group with higher caries experience
had more restored teeth since the beginning of the study than those from the VIS group (eTable 1,

410 JADA 151(6) n http://jada.ada.org n June 2020

Table 1. Baseline demographic and clinical characteristics of participants randomized to study groups related to the
diagnostic strategy used for detection of caries in primary molars and considering participants who were followed to the
end of the study and those who dropped out.*

VIS† GROUP, RAD‡ GROUP, STAYED DROPPED OUT,

CHARACTERISTIC NO. (%) NO. (%) IN, NO. NO.
Categorical Variable

Total 128 (50.8) 124 (49.2) 216 36

§
VIS group NA NA 106 22

RAD group NA NA 110 14

Sex
Male 60 (47.6) 66 (52.4) 108 18

Female 68 (54.0) 58 (46.0) 108 18

Age, y
3-4 62 (49.6) 63 (50.4) 108 17

5-6 66 (52.0) 61 (48.0) 108 19

Caries experience
dmf-s{ score 0-3 57 (50.4) 56 (49.6) 97 16

dmf-s score > 3 71 (51.1) 68 (48.9) 119 20

VIS GROUP, RAD GROUP, STAYED IN, DROPPED OUT,

CHARACTERISTIC MEAN (SD#) MEAN (SD) MEAN (SD) MEAN (SD)
Quantitative Variable

Age, y 4.89 (0.99) 4.92 (0.94) 4.90 (0.97) 4.91 (0.95)

Carious surfaces, no. 7.29 (10.52) 6.92 (10.59) 7.06 (10.29) 7.39 (12.10)

Surfaces missed due to caries, 0.12 (1.24) 0.54 (3.18) 0.29 (2.26) 0.53 (3.17)
no.
Restored surfaces, no. 1.08 (2.38) 0.84 (1.76) 0.97 (2.15) 0.92 (1.81)

DMFS score 8.45 (11.11) 8.31 (11.74) 8.30 (11.28) 8.83 (12.26)
* No differences were observed between groups considering participants who stayed in and who dropped out (P ¼ .182, calculated
by c2 test). † VIS: Visual inspection. ‡ RAD: Radiographic examination. § NA: Not applicable. { dmf-s: Decayed, missed, and
filled surfaces in primary teeth (considering all dental surfaces). # SD: Standard deviation.

available online at the end of this article). The same trend was observed after analyzing the in-
terventions performed in the proximal surfaces only (eTable 2, available online at the end of this
article).
No harms or unintended effects specifically related to diagnostic strategy were reported. Twenty-
three children (18.8%) from the VIS group or their caregivers and 30 children or their caregivers
(24.2%) from the RAD group reported pain episodes (P ¼ .342). The causes of pain, however,
varied. From the pain reports, it was determined that most were related to the exfoliation process of
primary teeth, and only 9 could be associated with caries in the primary molars (4 from the VIS
group, 5 from the RAD group). No other adverse episodes were reported.

DISCUSSION
The recommended diagnostic strategy for detection of carious lesions in daily clinical practice is the
simultaneous use of VIS and RAD for all patients of different ages.2-5 This recommendation has
been based on accuracy studies, and assessing the true clinical value of diagnostic tests should not be
limited to their accuracy. The impact on patients’ health outcomes must also be investigated
through RCTs.11-13 To our knowledge, our study is the first RCT comparing 2 diagnostic strategies
for detection of carious lesions in primary teeth and is probably the first one with this design,
considering the whole field of dentistry.

JADA 151(6) n http://jada.ada.org n June 2020 411

 Enrollment Assessed for eligibility (n = 261)

Excluded (n = 9)
• Did not meet inclusion criteria/
behavioral problems (n = 8)
• Declined to participate (n = 1)

Randomized (n = 252)

Allocation

Allocated to visual inspection group (n = 128) Allocated to radiographic method group (n = 124)
• Received allocated intervention (n = 126) • Received allocated intervention (n = 123)
• Did not attend subsequent • Did not attend subsequent appointments (n = 1)
appointments (n = 2)

Follow-up, 24 Months

Followed-up (n = 106) Followed-up (n = 110)

Lost to follow-up (n = 22) Lost to follow-up (n = 14)
• Family moved to another city (n = 1) • Family moved to another city (n = 2)
• Did not attend recall visits (n = 21) • Did not attend recall visits (n = 12)
• Discontinued intervention (n = 0) • Discontinued intervention (n = 0)

Analysis

Analyzed (n = 128) Analyzed (n = 124)

• Excluded from analysis (n = 0) • Excluded from analysis (n = 0)

Figure. Consolidated Standards of Reporting Trials study flowchart of participants enrolled, followed, and analyzed.15

Results of previous accuracy studies for detection of caries in children showed that the combi-
nation of methods did not have advantages.8-10 In our study, we compared children allocated to the
strategy recommended by means of clinical guidelines with VIS alone, considering the occurrence of
new operative interventions after 2 years of follow-up. Our basic premise guiding this choice was
that if the failure rate of restorations performed at baseline is lower than new carious lesion oc-
currences during follow-up, a more sensitive diagnostic strategy would be preferable. However, if the
occurrences of restoration failure are higher, a method with higher specificity would be more ad-
vantageous. We considered that, after dental treatments, patients would like to remain healthy,
with no new treatments needed for as long as possible.
Results of the primary outcome analysis showed no differences between strategies, although
children from the RAD group had more new interventions than those in the VIS only group. The
absence of significant differences between groups might be due to the procedure used for sample size
calculation. In the original protocol, the occurrence of new intervention needs was overestimated,
probably because we gathered data from observational studies and all dental surfaces were consid-
ered.15 In our study, however, we only analyzed new operative interventions in primary molars,
which can be considered as a limitation of our study.
Even considering the groups’ similarities, we can assert that simultaneous use of VIS with RAD as
a diagnostic strategy for caries detection in children did not lead to fewer new interventions than

412 JADA 151(6) n http://jada.ada.org n June 2020

Table 2. Number of primary molar surfaces of children allocated according to the initial treatment performed, decided
by respective diagnostic strategy used for detection of caries in primary molars (trial groups).
VARIABLE RADIOGRAPHIC P
VISUAL INSPECTION GROUP EXAMINATION GROUP VALUE‡
(N [ 128) (N [ 124)

Mean (SD*) Median (IQR†) Mean (SD) Median (IQR)

Surfaces With No Initial 20.3 (10.1) 22.0 (11.5-29.0) 17.8 (10.2) 18.0 (9.5-26.5) .046
Treatment Done

Surfaces With 9.9 (6.1) 9.5 (6.0-13.5) 10.0 (6.2) 10.0 (6.0-15.0) .692
Nonoperative Treatment

Surfaces Restored at 3.3 (4.7) 1.0 (0.0-5.0) 4.1 (5.0) 2.0 (0.0-6.0) .065
the Beginning of
the Study

False-Positive Results§ 0.04 (0.26) 0.0 (0.0-0.0) 0.48 (1.17) 0.0 (0.0-1.0) < .001

Teeth Submitted for 0.30 (0.81) 0.0 (0.0-0.0) 0.28 (0.70) 0.0 (0.0-0.0) .852
Endodontic Treatment

Teeth Extracted 0.23 (0.62) 0.0 (0.0-0.0) 0.22 (0.61) 0.0 (0.0-0.0) .929

* SD: Standard deviation. † IQR: Interquartile range. ‡ Derived by Mann-Whitney test. § Obtained from restored surfaces.

Table 3. Intention-to treat analyses with all randomized children (N ¼ 252) considering number of surfaces that needed
new interventions during the follow-up (primary outcome) and other secondary outcomes according to groups related
to the diagnostic strategy used for detection of caries in primary molars.
VARIABLE RADIOGRAPHIC P
VISUAL INSPECTION EXAMINATION GROUP VALUE‡
GROUP (N [ 128) (N [ 124)

Mean Median Mean Median

(SD*) (IQR†) (SD) (IQR)
Primary Outcome

Surfaces with new operative interventions 3.4 (5.5) 1.0 (0.0-5.0) 3.2 (4.1) 2.0 (0.0-5.0) .476

Secondary Outcomes

Surfaces with new carious lesions 0.8 (1.6) 0.0 (0.0-1.0) 0.7 (1.2) 0.0 (0.0-1.0) .858

Surfaces with replacement of restorations 1.2 (3.6) 0.0 (0.0-1.0) 1.3 (2.2) 0.0 (0.0-2.0) .038

Surfaces with repair of restorations 1.7 (3.2) 0.0 (0.0-2.0) 2.0 (3.8) 1.0 (0.0-2.0) .412

Surfaces with restorative procedures since 5.3 (8.1) 3.0 (0.0-7.0) 6.1 (6.3) 5.0 (1.5-9.0) .038
the beginning of the study

Carious lesions in the permanent first molars 0.16 (0.51) 0.0 (0.0-0.0) 0.17 (0.54) 0.0 (0.0-0.0) .939

Teeth with new endodontic treatments 0.05 (0.25) 0.0 (0.0-0.0) 0.03 (0.18) 0.0 (0.0-0.0) .775

Teeth extracted during follow-up 0.16 (0.51) 0.0 (0.0-0.0) 0.19 (0.49) 0.0 (0.0-0.0) .258

* SD: Standard deviation. † IQR: Interquartile range. ‡ Calculated by Mann-Whitney test.

VIS alone. Given this, and based on the principle of parsimony, detection of carious lesions in
preschool-aged children as part of the caries diagnostic process should be based on VIS alone.
Analyses of secondary outcomes reinforce this recommendation. We hypothesized that children
submitted to a combination of VIS with subsequent RAD for detection of carious lesions would
receive more restorations and, consequently, more restoration failures would be found at follow-up.
However, children who received a diagnosis by means of VIS only would have a higher number of
missed carious lesions and, therefore, new treatments would be necessary at follow-up. Because we
observed more restorations during the study and more restoration replacements in the RAD group,
our first hypothesis was confirmed, although on the basis of the analyses of the secondary outcomes.
However, children undergoing VIS alone did not have a substantially higher number of new res-
torations at follow-up, rejecting our second hypothesis. In fact, RAD had a low impact on changes
in the treatment decision made by means of VIS alone in a before-and-after study published

JADA 151(6) n http://jada.ada.org n June 2020 413

 previously by our group,20 probably owing to the low prevalence of nonevident carious lesions
requiring operative treatment in primary molars, as observed in other studies.8,10
The same trends were observed in the secondary analyses stratified by means of caries experience
or considering only proximal surfaces. Investigators of several studies have speculated that under-
going radiography could be more useful in children with higher caries experience10,19 or for lesion
detection on proximal surfaces.19 However, our findings did not support these previous conjectures,
even though subgroup analyses should be interpreted with caution.
Authors who advocate use of radiography as a protocol for caries detection have stated that
obtaining bite-wing radiographs would allow detection of carious lesions before operative inter-
vention is needed.3-5 Our results contradict that observation because the number of surfaces
indicated for nonoperative treatment was similar between groups. No differences were observed in
the occurrences of new lesions at follow-up. Another important finding is the higher number of
false-positive results in children in the RAD group, which is in accordance with accuracy studies.6,7
We observed that the management of carious lesions based on the simultaneous use of VIS with
RAD as a detection strategy is a more invasive approach than VIS only, providing no benefits for
children. Recommendations for clinical practice resulting from this investigation focus primarily on
the use of VIS alone as the strategy of choice for detection of carious lesions and treatment
planning. Obtaining bite-wing radiographs, however, could be considered, in a sequential associ-
ation, helpful for choosing the best treatment approach (that is, nonoperative versus operative
treatment) for some lesions detected by means of VIS. This association aligns best with the min-
imum intervention dentistry approach, but it should be tested further through an RCT.

CONCLUSIONS
The simultaneous use of VIS and RAD for detection of carious lesions in primary molars does not lead to
avoidance of new operative treatments compared with VIS alone. Consequently, VIS must always be used
for detection of carious lesions and for making decisions about appropriate treatment management in
children. n

SUPPLEMENTAL DATA
Supplemental data related to this article can be found at: http://doi.org/10.1016/j.adaj.2020.02.008.

Dr. Pontes is a PhD student, School of Dentistry, University of São Paulo,
São Paulo, Brazil.
Dr. Novaes is a lecturer, Dental Institute, Cruzeiro do Sul University, São
Paulo, Brazil.
Dr. Lara is a postdoctoral research fellow, Department of Cariology,
Operative Dentistry and Dental Public Health, Indiana University School of
Dentistry, Indianapolis, IN.
Dr. Gimenez is a lecturer, School of Dentistry, Ibirapuera University, São
Paulo, Brazil.
Ms. Moro is a PhD student, School of Dentistry, University of São Paulo,
São Paulo, Brazil.
Dr. Camargo is a lecturer, Department of Pediatric Dentistry, Paulista
University, São Paulo, Brazil.
Dr. Michel-Crosato is an associate professor, Department of Community
Dentistry, School of Dentistry, University of São Paulo, São Paulo, Brazil.
Dr. Pannuti is an associate professor, Division of Periodontics, School of
Dentistry, University of São Paulo, São Paulo, Brazil.
Dr. Daniela P. Raggio is an associate professor, Department of Pediatric
Dentistry, School of Dentistry, University of São Paulo, São Paulo, Brazil.
Dr. Braga is an associate professor, Department of Pediatric Dentistry,
School of Dentistry, University of São Paulo, São Paulo, Brazil.
Dr. Mendes is an associate professor, Department of Pediatric
Dentistry, School of Dentistry, University of São Paulo, 2227 Lineu Prestes
Ave, São Paulo, SP, Brazil 05508-000, e-mail fmmendes@usp.br. Address
correspondence to Dr. Mendes.
Disclosure. None of the authors reported any disclosures.
This trial was supported by grant 2012/24243-7 from the Fundação de
Amparo à Pesquisa do Estado de São Paulo (São Paulo Research Founda-
tion), grants 471817/2012-0 and 306304/2015-5 from National Council
for Scientific and Technological Development, and Coordination for
Improvement of Higher Education Personnel, Brazilian funding agencies.
National Council for Scientific and Technological Development provides
scholarship awards for research productivity in Brazil to Dr. Mendes, Dr.
Pannuti, Dr. Raggio, and Dr. Braga.
The authors would like to thank all members of the Caries Detection in
Children collaborative group, who were important for promoting, orga-
nizing, and conducting all procedures related to this study and the entire
clinical trial. A list of these members and their respective roles can be
accessed at https://is.gd/FrqlYo. The authors also wish to thank the par-
ticipants of the pediatric dentistry seminar graduate program from the
University of São Paulo, São Paulo, Brazil, for their critical comments. The
Center for Research and Learning at Indiana University, Indianapolis,
Indiana, proofread this article.

1. Braga MM, Mendes FM, Ekstrand KR. Detection 2. American Academy on Pediatric Dentistry Ad Hoc Guideline on prescribing dental radiographs for infants,
activity assessment and diagnosis of dental caries lesions. Committee on Pedodontic Radiology, American Acad- children, adolescents, and persons with special health care
Dent Clin North Am. 2010;54(3):479-493. emy on Pediatric Dentistry Council on Clinical Affairs. needs. Pediatr Dent. 2010;30(7 suppl):236-237.

414 JADA 151(6) n http://jada.ada.org n June 2020

 3. Ismail AI, Pitts NB, Tellez M, et al. The Interna-
tional Caries Classification and Management System
(ICCMS): an example of a caries management pathway.
BMC Oral Health. 2015;15(1 Suppl):S9.
4. Kuhnisch J, Ekstrand KR, Pretty I, et al. Best clinical
practice guidance for management of early caries lesions in
children and young adults: an EAPD policy document. Eur
Arch Paediatr Dent. 2016;17(1):3-12.
5. Martignon S, Pitts NB, Goffin G, et al. CariesCare
practice guide: consensus on evidence into practice. Br
Dent J. 2019;227(5):353-362.
6. Gimenez T, Piovesan C, Braga MM, et al. Visual
inspection for caries detection: a systematic review and
meta-analysis. J Dent Res. 2015;94(7):895-904.
7. Schwendicke F, Tzschoppe M, Paris S. Radiographic
caries detection: a systematic review and meta-analysis.
J Dent. 2015;43(8):924-933.
8. Mendes FM, Novaes TF, Matos R, et al. Radiographic
and laser fluorescence methods have no benefits for detecting
caries in primary teeth. Caries Res. 2012;46(6):536-543.
9. Bussaneli DG, Restrepo M, Boldieri T,
Albertoni TH, Santos-Pinto L, Cordeiro RC. Proximal
JADA 151(6) n http://jada.ada.org n June 2020
caries lesion detection in primary teeth: does this justify
the association of diagnostic methods? Lasers Med Sci.
2015;30(9):2239-2244.
10. Moro BLP, Novaes TF, Pontes LRA, et al. The in-
fluence of cognitive bias on caries lesion detection in
preschool children. Caries Res. 2018;52(5):420-428.
11. Sackett DL, Haynes RB. The architecture of diag-
nostic research. BMJ. 2002;324(7336):539-541.
12. Ferrante di Ruffano L, Hyde CJ, McCaffery KJ,
Bossuyt PM, Deeks JJ. Assessing the value of diagnostic
tests: a framework for designing and evaluating trials. BMJ.
2012;344:e686.
13. Glick M. Prevention, screening, and a chance of
rain. JADA. 2015;146(4):217-218.
14. El Dib R, Tikkinen KAO, Akl EA, et al. Systematic
survey of randomized trials evaluating the impact of
alternative diagnostic strategies on patient-important
outcomes. J Clin Epidemiol. 2017;84:61-69.
15. Schulz KF, Altman DG, Moher D; CONSORT
Group. CONSORT 2010 statement: updated guidelines
for reporting parallel group randomised trials. BMJ. 2010;
340:c332.
16. Mendes FM, Pontes LR, Gimenez T, et al;
CARDEC Collaborative Group. Impact of the radio-
graphic examination on diagnosis and treatment deci-
sion of caries lesions in primary teeth: the CARies
DEtection in Children (CARDEC-01) trialdstudy
protocol for a randomized controlled trial. Trials. 2016;
17:69.
17. World Health Organization. Oral Health Surveys:
Basic Methods. 5th ed. Geneva, Switzerland: World Health
Organization; 2013:125.
18. Ismail AI, Sohn W, Tellez M, et al. The International
Caries Detection and Assessment System (ICDAS): an in-
tegrated system for measuring dental caries. Community Dent
Oral Epidemiol. 2007;35(3):170-178.
19. Nyvad B, Machiulskiene V, Baelum V. Reliability of
a new caries diagnostic system differentiating between
active and inactive caries lesions. Caries Res. 1999;33(4):
252-260.
20. Pontes LRA, Novaes TF, Lara JS, et al. Impact of the
radiographic method on treatment decisions related to
dental caries in primary molars: a before-after study. Clin
Oral Investig. 2019;23(11):4075-4081.
415
 eTable 1. Intention-to treat analyses with the participants stratified by caries experience: children with lower caries
experience (decayed, missing, or filled surfaces index < 4) and higher caries experience (decayed, missing, or filled
surfaces index  4) considering the primary outcome and some secondary outcomes.
VARIABLE RADIOGRAPHIC P VALUE‡
VISUAL INSPECTION GROUP EXAMINATION GROUP

Mean (SD*) Median (IQR†) Mean (SD) Median (IQR)

{
Children With dmf-s < 4§

Surfaces with new operative 1.3 (3.1) 0.0 (0.0-1.3) 1.2 (2.1) 0.0 (0.0-2.0) .612
interventions (primary outcome)

Surfaces with new carious lesions 0.7 (1.5) 0.0 (0.0-0.0) 0.5 (1.0) 0.0 (0.0-0.5) .663

Surfaces with replacement of 0.2 (0.6) 0.0 (0.0-0.0) 0.5 (0.9) 0.0 (0.0-1.0) .052
restorations

Surfaces with repair of restorations 0.5 (1.9) 0.0 (0.0-0.0) 0.5 (1.3) 0.0 (0.0-1.0) .166

Surfaces with restorative 1.5 (2.6) 0.0 (0.0-2.0) 2.0 (2.4) 1.5 (0.0-3.0) .086
procedures since the beginning of
the study

Children With dmf-s ‡ 4#

Surfaces with new operative 5.2 (6.3) 4.0 (0.0-7.0) 4.9 (4.5) 4.5 (1.0-7.0) .537
interventions (primary outcome)

Surfaces with new carious lesions 1.0 (1.7) 0.0 (0.0-1.0) 0.9 (1.3) 0.0 (0.0-1.0) .985

Surfaces with replacement of 2.0 (4.7) 0.0 (0.0-2.0) 1.9 (2.7) 1.0 (0.0-3.0) .187
restorations

Surfaces with repair of restorations 2.7 (3.6) 2.0 (0.0-4.0) 3.2 (4.7) 2.0 (0.0-4.0) .609

Surfaces with restorative 8.4 (9.6) 6.0 (3.0-11.0) 9.5 (6.6) 8.0 (5.0-12.5) .038
procedures since the beginning of
the study

* SD: Standard deviation. † IQR: Interquartile range. ‡ Calculated by Mann-Whitney test. § dmf-s: Decayed, missing, or filled
surfaces in primary teeth index. { Visual inspection group, n ¼ 57; radiographic examination group, n ¼ 56. # Visual
inspection group, n ¼ 71; radiographic examination group, n ¼ 68.

eTable 2. Intention-to treat analyses with all randomized children (N ¼ 252) considering only intervention in proximal
surfaces of primary molars according to the groups related to the diagnostic strategy used for detection of caries.
VARIABLE RADIOGRAPHIC P VALUE‡
VISUAL INSPECTION GROUP EXAMINATION GROUP
(N [ 128) (N [ 124)

Mean (SD*) Median (IQR†) Mean (SD) Median (IQR)

Primary Outcome

Proximal surfaces with new operative 1.3 (2.6) 0.0 (0.0-1.0) 1.4 (1.8) 0.0 (0.0-2.0) .086
interventions

Secondary Outcomes

Proximal surfaces with new carious lesions 0.5 (1.0) 0.0 (0.0-1.0) 0.4 (0.8) 0.0 (0.0-0.5) .763

Proximal surfaces with replacement of 0.5 (1.5) 0.0 (0.0-0.0) 0.5 (1.0) 0.0 (0.0-1.0) .071
restorations

Proximal surfaces with repair of restorations 0.6 (1.4) 0.0 (0.0-1.0) 0.8 (1.6) 0.0 (0.0-1.0) .108

Proximal surfaces with restorative 1.3 (2.6) 0.0 (0.0-2.0) 1.8 (2.2) 1.0 (0.0-3.0) .001
procedures since the beginning of the study

* SD: Standard deviation. † IQR: Interquartile range. ‡ Calculated by Mann-Whitney test.

415.e1 JADA 151(6) n http://jada.ada.org n June 2020