ORIGINAL CONTRIBUTIONS
Do glass ionomer cements prevent
caries lesions in margins of
restorations in primary teeth?
A systematic review and meta-analysis
Daniela Prcida Raggio, DDS, MSc, PhD; Tamara Kerber
Tedesco, DDS, MSc, PhD; Ana Flvia Bissoto Calvo, DDS,
MSc, PhD; Mariana Minatel Braga, DDS, MSc, PhD
N
ew caries lesions in restoration margins are a
frequent concern in dentistry, especially when
there is no patient compliance. Thus, this
problem has been seen as the main reason for
failure and replacement of restorations in primary teeth,1
with reports showing approximately 8.0% of restoration
failures even within 5 years when these caries lesions are
lled with polyacid-modied resin composite (PMRC),
resin composite (RC), or amalgam.2,3
These restorative materials, in the same way as glass
ionomer cements (GICs), have shown satisfactory per-
formance in restorations of primary teeth.4,5 However,
conventional GICa low-viscosity restorative material
has a shorter longevity than do the other materials.4
Results of a previous systematic review showed that there
is a higher number of failed restorations with the
atraumatic restorative treatment (ART) technique when
it was performed with conventional GIC, whereas the
longevity of ART restorations performed with high-
viscosity GIC (HVGIC) is higher.6 HVGIC is also a
material for which setting is an acid-based reaction;
however, HVGIC performed similarly to the other ma-
terials in both occlusal and occlusoproximal restora-
tions.7,8 Conversely, resin-modied GIC (RMGIC)a
GIC with addition of hydroxyethylmethacrylate, similar
to HVGICalso can be considered an alternative to
restore dentinal caries lesions.5
Fluoride interferes with the processes of demineral-
ized and remineralization of caries lesions, and some
authors suggest that the uoride released from GICs is
capable of preventing caries.9,10 Investigators in previous
Copyright 2016 American Dental Association. All rights reserved.
ABSTRACT
Background. Fluoride released from glass ionomer ce-
ments (GICs) is capable of preventing caries lesions.
However, the preventive effect in margins of occlusal and
occlusoproximal restorations have not been proved. The
aim of this study was to evaluate the ability of GIC to
prevent caries lesions in margins of occlusal and occluso-
proximal restorations in primary teeth compared with that
of other restorative materials.
Types of Studies Reviewed. The authors conducted a
literature search in PubMed and MEDLINE to verify the
clinical trials available on the outcome of caries lesions. The
inclusion criteria were that the subject related to the scope
of this systematic review, the study had a follow-up, and the
study was not performed in specic groups. The authors
performed all meta-analyses by considering the secondary
caries rates for the restorations in clinical trials.
Results. The search strategy identied 450 potentially
relevant studies, and the authors included 8 of them in the
review. The main reasons for exclusion were that the
studies were not related to the scope of this review or were
not longitudinal trials. The secondary caries rate of the
occlusal restorations was not different among the restor-
ative materials (odds ratio, 1.2; 95% condence interval,
0.5-3.1). For occlusoproximal analysis, GIC was associated
signicantly with better ability to prevent caries lesions
(odds ratio, 1.7; 95% condence interval, 1.2-2.5).
Conclusions and Practical Implications. Because
new caries lesions in the margins of restorations are the
main reason for failure and replacement of restorations in
primary teeth, it is important to know whether there is a
benet in using GICs in both occlusal and occlusoproximal
cavities.
Key Words. Dental caries; glass ionomer cements;
uoride.
JADA 2016:147(3):177-185
http://dx.doi.org/10.1016/j.adaj.2015.09.016
JADA 147(3) http://jada.ada.org March 2016 177
 ORIGINAL CONTRIBUTIONS
studies reported that GIC showed a higher reduction of
demineralization in adjacent teeth.11,12 However, GICs
preventive effect has not been proved when compared
with all other available materials considered as denitive
restorative materials, especially for the occlusal and
occlusoproximal surfaces of the primary teeth.
This is important because most caries lesions
occur on the proximal surface of the primary teeth,13
and the caries progression in this area seems to be
faster than on occlusal surfaces.14 In this sense, the
survival rate of restorations could be different between
these surfaces. Furthermore, primary teeth have a
higher tubule density and lower concentration of
phosphate and calcium in peritubular and intertubular
dentin than do permanent teeth,15,16 which could
interfere with the performance of restorative materials.
However, to the best of our knowledge, this is the rst
systematic review and meta-analysis that compares the
preventive effect of all restorative materials available on
caries lesions in the margins of restorations in both the
occlusal and occlusoproximal surfaces of primary teeth.
Thus, the aim of this study was to evaluate systemati-
cally and quantitatively the ability of GIC to prevent
caries lesions in the margins of occlusal and occluso-
proximal restorations in primary teeth compared with
that of other restorative materials.
METHODS
We conducted and reported this study according to the
Preferred Reporting Items for Systematic Reviews and
Meta-Analyses statement.17 We registered it on the
PROSPERO register under protocol number
CRD42013006497.
Search strategy and selection criteria. We performed
a comprehensive literature search through PubMed and
MEDLINE to identify articles up to August 19, 2014, in
which the investigators evaluated the prevention of caries
lesions in the margins of occlusal and occlusoproximal
GIC restorations in primary teeth. To retrieve all relevant
articles, 2 authors (T.K.T. and A.F.B.C.) screened refer-
ence lists of included articles and related reviews. We
used the following search strategy: ((((((((((((amalgam)
OR resin*) OR composite*) OR composite resin*) OR
resin composite*) OR compomer*) OR polyacid modi-
ed composite resin*) OR polyacid-modied composite
resin*)) AND (((((((((demineralization) OR tooth
demineralization) OR teeth demineralization) OR caries)
OR carious) OR tooth decay) OR teeth decay) OR dental
caries) OR caries susceptibility)) AND (((((dental resto-
ration*) OR restoration) OR dental restoration, perma-
nent) OR tooth restoration) OR teeth restoration)) AND
(((((glass ionomer cement*) OR glass-ionomer cement*)
OR GIC) OR ART) OR atraumatic restorative proce-
dure*)) AND ((clinical[Title/Abstract] AND trial[Title/
Abstract]) OR clinical trials[MeSH Terms] OR clinical
trial[Publication Type] OR random*[Title/Abstract] OR
178 JADA 147(3) http://jada.ada.org March 2016
random allocation[MeSH Terms] OR therapeutic use
[MeSH Subheading]).
Initially, both reviewers independently assessed the
identied publications, which we selected by title and
abstract on the basis of the inclusion criteria: to investi-
gate occlusal and occlusoproximal GIC restorations and
to be a longitudinal study with a follow-up of at least
12 months. We did not include studies performed in
specic groups (for example, irradiation, special patients,
and teeth with amelogenesis imperfecta). The reviewers
were trained and calibrated for article selection (k
0.929) by a experienced researcher in studies about caries
lesions in margins of restorations (D.P.R.). We resolved
any discrepancies through a third reviewer (D.P.R.). We
made a nal decision about inclusion on the basis of the
full-text articles of the potentially relevant studies in
accordance with the exclusion criteria: having a dropout
rate of more than 30%, not being a randomized or qua-
sirandomized clinical trial, not having a control group
(amalgam, PMRC, or RC), not evaluating GIC as a
denitive restoration (HVGIC or RMGIC), not being
performed in primary teeth, and not evaluating caries
lesions in margins of restorations as the outcome. In the
case of studies reporting the same sample, we included
those that presented more information.
Data extraction. The 2 reviewers independently
collected the data of the eligible studies. For each
article, they systematically extracted the following data:
publication details (title, authors, and year), sample
characteristics (age of participants, caries experience,
number of participants, number of restorations for
each material), study methodology (study design,
restorative materials, type of restored cavity), and
outcome information (survival of restorations,
follow-up, and dropout).
Afterwards, we assessed the risk of bias in the
included studies (k 0.945) by using specic study
designrelated risk-of-bias assessment forms.18
We divided the criteria into 7 main domains related to
randomization, masking, outcome data, and character-
istics of the sample at baseline. We evaluated the studies
by rating each of the study criteria as yes (low risk of
bias), no (high risk of bias), or unclear (no information
or uncertainty about the potential for bias). For the nal
classication of risk of bias, we resolved disagreements
between the reviewers through discussion.
Statistical methods for the meta-analysis. We per-
formed all meta-analyses by using statistical software
(MedCalc Version 12.5.0.0; Microsoft Partner). We
ABBREVIATION KEY. ART: Atraumatic restorative treat-
ment. GIC: Glass ionomer cement. HVGIC: High-viscosity
glass ionomer cement. PMRC: Polyacid-modied resin com-
posite. RC: Resin composite. RMGIC: Resin-modied glass
ionomer cement. USPHS: US Public Health Services.
 ORIGINAL CONTRIBUTIONS

Articles identified through Additional articles identified

Identification

database searching through other sources

(n = 448) (n = 2)

Potential articles identified and screened for inclusion

(n = 450)
Not included (n = 397)
Not related to occlusal or occlusoproximal
restorations (n = 85)
Screening

Other restorative materials (n = 30)

Not a longitudinal study (n = 149)
Outcomes not related to presence of
caries in margins at restorations (n = 72)
Related to evaluation of restoration
technique (n = 49)
Outcome related to adjacent teeth (n = 4)
Specific groups (n = 8)
Articles retrieved for
full-text evaluation
(n = 53)
Excluded (n = 45)
Eligibility

Dropout rate more than 30% (n = 7)

Other restorative materials (n = 13)
Not a longitudinal study (n = 9)
Without comparison groups (n = 6)
Related to permanent teeth (n = 2)
Outcome not related to secondary
caries (n = 3)
Study with the same sample (n = 3)
Not accessible (n = 2)
Articles included in
qualitative synthesis
(n = 8)
Included

Articles included in
quantitative synthesis
(n = 7)

Figure 1. Flow diagram of trial selection.

considered the secondary caries rate of occlusal and
occlusoproximal restorations that were shown as the
reason for failure reported in the clinical trials. For both
types of restored cavity, we performed the meta-
analysis by using the longest follow-up of each study.
We used random-effects models for all calculations. For
the pooled studies, we used an Egger test to aid the
analysis of publication bias. It was not possible to
perform a meta-analysis of HVGIC versus other ma-
terials and of RMGIC versus other restorative materials
in different follow-up periods because there was
insufcient information to be included in the pooling.
Thus, we analyzed these data descriptively.
RESULTS
Study selection. The search strategy identied 448
potentially relevant records. We identied another 2
studies from the reference lists of related reviews. After
screening titles and abstracts, we retrieved 53 full-text
articles for more detailed information. Most of the
studies not included were not related to the scope of this
review (240; 60.5%) or were not longitudinal clinical

JADA 147(3) http://jada.ada.org March 2016 179

 ORIGINAL CONTRIBUTIONS

TABLE 1
Main characteristics of data sets from randomized studies.
STUDY AGE OF PARTICIPANTS IN CARIES EXPERIENCE STUDY DESIGN NO. OF NO. OF
YEARS, MEAN (STANDARD PARTICIPANTS RESTORATIONS
DEVIATION)*
Donly and Colleagues,19 8 (1.2) At least 2 class II caries Split mouth 40 60 (RMGIC, 60;
1999 lesions amalgam, 60)
Dutta and Colleagues,20 6.5 Two to 6 caries lesions in Parallel group 120 480 (RMGIC, 360;
2001 primary molars amalgam, 120)
Taifour and 6.5 Decayed, missing, or lled Parallel group 835 1,891 (HVGIC, 1,086;
Colleagues,21 2002 primary teeth index 4.4 amalgam, 805)

Andersson-Wenckert 8 At least 2 proximal caries Split mouth 57 132 (RMGIC, 66;

and Sunnegardh- lesions RC,# 66)
Grnberg,22 2006
Ersin and Colleagues,23 8.07 (1.51) Decayed or lled primary Split mouth 219 419 (HVGIC, 215;
2006 teeth index 5.08 (2.66)** RC, 204)

Daou and Colleagues,24 7 (0.5) At least 2 dentin lesions in Split mouth 45 149 (RMGIC, 37;
2008 primary teeth (high caries HVGIC, 35; amalgam,
risk activity) 38; PMRC, 39)

dos Santos and 5 (0.75) At least 2 primary caries Split mouth 48 141 (RMGIC, 46;
Colleagues,25 2009 lesions RC, 44; PMRC, 51)

Hilgert and 6.8 (0.40) At least 2 caries lesions in Parallel group 280 750 (HVGIC, 386;
Colleagues,26 2014 dentin in primary teeth amalgam, 364)

* Standard deviation is not available for all studies, as some studies mentioned the age range of the included children and not the mean
(standard deviation).
RMGIC: Resin-modied glass ionomer cement.
USPHS: US Public Health Service criteria.
HVGIC: High-viscosity glass ionomer cement.
ART: Atraumatic restorative treatment.
# RC: Resin composite.
** Index is mean (standard deviation).
PMRC: Polyacid-modied resin composite.

trials (149; 37.5%). Finally, 8 articles met the eligibility
criteria, and we included them in the review. The ow
diagram summarizes the study selection process and the
reasons for exclusions (Figure 1).
Study characteristics. Table 119-26 shows the main
characteristics of the included articles. All studies were
performed in participants aged 5 to 8 years. Investigators
in most studies (62.5%) compared RMGIC with others
materials: RC,22,23,25 amalgam,19-21,24,26 or PMRC.23,25
However, investigators in 4 studies used HVGIC.21,23,24,26
Considering the type of restored cavity, investigators in
all trials evaluated occlusoproximal restorations, whereas
investigators in only 5 studies evaluated occlusal cav-
ities.21,23-26 However, Daou and colleagues24 presented
the survival data of the occlusal and occlusoproximal
restorations together.
The main criterion of the outcome evaluation was
based on US Public Health Services criteria,27 which were
used in 75% of the studies. This criterion considers
marginal adaptation, discoloration, secondary caries,
anatomic form, color match, and surface texture.
Furthermore, most of the studies followed up the
patients for 12 and 24 months.19,22,25,26
Risk of bias. Table 219-26 shows the nal assessment
of risk of bias in the included studies. Investigators in
5 articles reported masking of outcome assessment.21-25
We found there was a lack of information about alloca-
tion concealment and masking of participants and
personnel in the studies. Investigators in only 1 study
reported sample characteristics at baseline,21 and in-
vestigators in 2 other studies reported other sources
of uoride.21,26

180 JADA 147(3) http://jada.ada.org March 2016

 ORIGINAL CONTRIBUTIONS

TABLE 1 (CONTINUED)

TYPE OF RESTORED INDEX FOLLOWUP TIME, DROPPED OUT, % PERCENTAGE OF SECONDARY CARIES
CAVITY MONTHS ACCORDING TO ALL FAILURES

Occlusoproximal USPHS 6, 12, 24, and 36 30 Occlusoproximal: RMGIC: 23.08

Amalgam: 29.41
Occlusoproximal USPHS 4, 8, and 12 6.9 Occlusoproximal: RMGIC: 4.7
Amalgam: 12.5
Occlusal, ART restoration 36 22.1 Occlusal: HVGIC: 7.4
occlusoproximal criteria Amalgam: 14.1

Occlusoproximal: HVGIC: 3.2

Amalgam: 5.5
Occlusoproximal USPHS 12 and 24 9.1 Occlusoproximal: RMGIC: 3.1
RC: 11.3

Occlusal, USPHS 24 17.8 Occlusal: HVGIC: 15.6

occlusoproximal RC: 11.1

Occlusoproximal: HVGIC: 28.1

RC: 24.6
Occlusal, USPHS 6 and 12 2.6 Occlusal/Occlusoproximal: RMGIC: 3
occlusoproximal HVGIC: 12
Amalgam: 2.8
PMRC: 8.3
Occlusal, USPHS; visible 12, 18, and 24 12.5 Occlusal: RMGIC: 6
occlusoproximal plaque index RC: 3.3
PMRC: 5.5

Occlusoproximal: RMGIC: 23.7

RC: 21.4
PMRC: 40
Occlusal, ART restoration 6, 12, 24, and 36 24.3 Occlusal: HVGIC: 20
occlusoproximal criteria Amalgam: 20

Occlusoproximal: HVGIC: 4.9

Amalgam: 2.6

TABLE 2
Methodological quality and risk of bias: review authors assessment.*
STUDY RANDOM ALLOCATION MASKING OF MASKING OF NO NO OTHER
SEQUENCE CONCEALMENT PARTICIPANTS OUTCOME INCOMPLETE BASELINE SOURCES OF
GENERATION AND PERSONNEL ASSESSMENT OUTCOME DATA IMBALANCE FLUORIDE
EXPOSURE
Donly and Yes Unclear Unclear Unclear Yes Unclear Unclear
Colleagues,19 1999
Dutta and Yes Unclear Unclear Unclear Yes Unclear Unclear
Colleagues,20 2001
Taifour and Yes Unclear Unclear Yes Yes Yes Yes
Colleagues,21 2002
Andersson-Wenckert Yes Unclear Unclear Yes Yes Unclear Unclear
and Sunnegardh-
Grnberg,22 2006
Ersin and Yes Unclear Unclear Yes Yes Unclear Unclear
Colleagues,23 2006
Daou and Yes Unclear Unclear Yes Yes Unclear Unclear
Colleagues,24 2008
dos Santos and Yes Unclear Unclear Yes Yes Unclear Unclear
Colleagues,25 2009
Hilgert and Yes Unclear Unclear Unclear Yes No Yes
Colleagues,26 2014
* Studies were evaluated by rating each of the study criteria as yes (low risk of bias), no (high risk of bias), or unclear (no information or uncertainty
about the potential for bias).

JADA 147(3) http://jada.ada.org March 2016 181

 ORIGINAL CONTRIBUTIONS

materialsGIC (HVGIC
and RMGIC) versus
amalgam and RC
Taifour and Colleagues,21 2002
(odds ratio, 1.2:, 95%
Ersin and Colleagues,23 2006
condence interval, 0.5-
3.1) (Figure 2).21,23,25,26 For
dos Santos and Colleagues,25 2009
the longest follow-up of
all included studies for
Hilgert and Colleagues,26 2014
occlusoproximal analysis,
GIC (HVGIC and
RMGIC) was associated
Total (fixed-effects model)
signicantly with better
ability than that of
Total (random-effects model)
amalgam and RC to pre-
vent caries lesions in the
margins of occlusoprox-
imal restorations (odds
Heterogeneity: Q = 7.6837, P = .0530 0.01 0.1 1 10 100
Egger test: P = .23670 ODDS RATIO ratio, 1.7; 95% condence
interval, 1.2-2.5)
(Figure 3).19-23,25,26
Figure 2. Random-effects model meta-analysis of the ability of restorative materials to prevent caries in margins Furthermore, in-
of occlusal restorations, with the longest follow-up of each included study: resin-modied glass ionomer cement vestigators in 1 of the
and high-viscosity glass ionomer cement versus resin composite and amalgam.
included studies pre-
sented the secondary
caries rate of occlusal
Donly and Colleagues,19 1999 and occlusoproximal
restorations together.24
Dutta and Colleagues,20 2001 Investigators in only
Taifour and Colleagues,21 2002
1 study evaluated
separated data for
Andersson-Wenckert and PMRC25 and those
Sunnegardh-Grnberg,22 2006 in 4 studies evaluated
Ersin and Colleagues,23 2006 separate data for
HVGIC.21,23,24,26 How-
dos Santos and Colleagues,25 2009 ever, the last 4 trials
Hilgert and Colleagues,26 2014
were conducted with
different follow-ups.
Total (fixed-effects model) Similarly, investigators
in only 5 studies evalu-
Total (random-effects model)
ated RMGIC,19,20,22,24,25
also with different
Heterogeneity: Q = 6.2034, P = .4008 0.1 1 10 100 follow-ups. For that
Egger test: P = .26067 ODDS RATIO reason, we did not
conduct meta-analysis
to evaluate these factors
Figure 3. Random-effects model meta-analysis of the ability of restorative materials to prevent caries in margins individually, and we
of occlusoproximal restorations, with the longest follow-up of each included study: resin-modied glass ionomer
cement and high-viscosity glass ionomer cement versus resin composite and amalgam. did not assess statistical
heterogeneity further.
Survival rate and meta-analysis results. Figures 221,23,25,26
and 319-23,25,26 present the main results of the meta-analyses. DISCUSSION
We observed no heterogeneity in all meta-analyses. Results Caries lesions in the margins of restorations remain a
from the Egger test showed no publication bias in all meta- major reason for the replacement of restorative materials
analyses. worldwide.28 To decrease this failure rate, investigators in
When we considered the secondary caries rate of the previous studies have targeted the effect of uoride release
longest follow-up of all included studies for occlusal by GIC in the prevention of caries lesions.29 In our sys-
analysis, it was veried that no differences between the tematic review, we evaluated the available evidence of the

182 JADA 147(3) http://jada.ada.org March 2016

 ORIGINAL CONTRIBUTIONS
ability of GICs to prevent adjacent caries lesions in occlusal
and occlusoproximal restorations in primary teeth
compared with that of other restorative materials.
We observed that all restorative materials performed
similarly on occlusal surfaces. Conversely, GIC was
better able to prevent caries lesions in the margins of
occlusoproximal restorations. We can hypothesize that
this difference is related to the higher risk of this surface
developing caries lesions because of the presence of
contact areas, which results in areas that are difcult to
reach by toothbrush and are accessible only by ossing.30
In this sense, investigators in a previous study reported a
lower survival rate of the restorations on proximal sur-
faces associated with poor oral hygiene.31 Because biolm
results in greater demineralization in the approximal re-
gion, the presence of uoride from GIC in this region
appears to be more important; this hypothesis is a
possible explanation for our results. The hypothesis of the
benet of GIC in preventing caries lesions in the margins
of occlusoproximal surfaces is related to the uoride-
releasing ability of this material when compared with
RC and amalgam. Investigators conducted previous
studies to verify this material property, which has been
conrmed in both in vitro and in situ studies.32,33 Thus,
although the other restorative materials can have satis-
factory performance in bond strength34 and retention,22-25
these factors are different outcomes compared with sec-
ondary caries lesions. This condition must be considered
in the choice of restorative material because a fractured
restoration could be repaired. However, in the presence of
caries lesions, restorations must be replaced.35
In this systematic review, we observed that the
longer the follow-up period, the better RMGIC tends
to perform compared with other nonuoride-releasing
restorative materials. This nding is expected because
there is an increase of this preventive effect because of
the uoride-releasing ability and uptake.36 However,
the small number of studies does not allow us to
conduct subgroup analyses considering different
follow-ups to conrm this hypothesis.
We did not perform quantitative synthesis on HVGIC
separately because there were not enough studies to do
so. It is important to highlight that the HVGIC resto-
rations, which we included in our systematic review,
were placed as ART. This technique is based on partial
caries removal, different from the conventional restor-
ative treatments, which in most cases are conducted with
total caries removal. However, the differences between
the techniques seem not to be related to secondary caries
formation.37 Moreover, the longevity of HVGIC occlu-
soproximal restorations already has been reported in a
previous meta-analysis, in which the investigators sug-
gested using this GIC because there are no differences
between it and the other available materials.8 Further-
more, GIC is easier to work because it does not need light
curing, and a shorter time is required for the application
protocol, which is a fundamental characteristic of
pediatric dentistry. In this sense, we encourage further
clinical trials to verify the preventive effect of this ma-
terial, given that the use of HVGIC seems to result in
greater cost-effectiveness than does RMGIC.
In the same way, it was not possible to perform a
meta-analysis comparing GIC with PMRC because of the
number of trials screened. However, investigators in
most included studies observed that there was a trend
of RMGIC performing better than PMRC. A possible
explanation for this result could be related to the
composition of this material that behaves more like an
RC and releases uoride at a lower rate than does GIC.33
Otherwise, the risk of bias assessment showed that
investigators in most of the studies did not report
important data for clinical trials. Thus, this lack of in-
formation could interfere in the quality analyses of these
studies. Likewise, there is no information about the
sample size calculated, which may not be representative
of the population, limiting the extrapolation of results.
Moreover, these evaluated studies had up to 36 months
of follow-up, which seems to be a brief time to evaluate
caries lesions as an outcome. Investigators have per-
formed some research with longer follow-ups; however,
most did not fulll the elegibility criteria, because even
if the patients did not return for the evaluations because
of exfoliation of primary teeth, they were considered to
have dropped out. Although tooth longevity appears to
be our goal as a clinical outcome, in our systematic re-
view we focused on development of secondary caries.
Thus, tooth longevity could not be considered a success
because of the impossibility of evaluating secondary
caries.
We also observed that there are no records about
other sources of uoride exposure because only 1 study
showed these data. This information is important
because the presence of uoride can mask the preventive
effect of GIC or lead to overestimation of the ability
of other materials in the prevention of caries lesions.
However, for these reasons, we collected this information
for quality analyses. It seems to be more important in the
parallel studies that the benet of other sources of uo-
ride exposure cannot have been considered, because
there were no baseline balance data. However, most of
the studies included in this meta-analysis (62.5%) were
split-mouth design, which may have resulted in little or
no inuence of the presence of other sources of uoride
exposure on the outcome.
A possible limitation of our systematic review is
related to the literature search that we conducted
only in PubMed and MEDLINE. However, although
Embase results in a higher number of studies inden-
tied,38 this difference between databases seems unim-
portant for biomedical studies because the Embase
search shows many studies that did not pass the in-
clusion and exclusion criteria. Hence, this wider search
JADA 147(3) http://jada.ada.org March 2016 183
 ORIGINAL CONTRIBUTIONS
does not always result in higher quality citations.38 In
this sense, the search only in PubMed and MEDLINE
seems a good option for systematic reviews in medical
science.
CONCLUSIONS
Thus, we can conclude that there is moderate strength
of evidence for a positive association between GIC and
the prevention of caries lesions only in the margins of
occlusoproximal restorations of primary teeth. n
Dr. Raggio is an associate professor, Department of Orthodontics and
Pediatric Dentistry, University of So Paulo, Av Lineu Prestes, 2227, Cidade
Universitria, So Paulo SP, Brazil, e-mail danielar@usp.br. Address cor-
respondence to Dr. Raggio.
Dr. Tedesco is a PhD student, Department of Orthodontics and Pediatric
Dentistry, University of So Paulo, So Paulo, Brazil.
Dr. Calvo is a PhD student, Department of Orthodontics and Pediatric
Dentistry, University of So Paulo, So Paulo, Brazil.
Dr. Braga is an associate professor, Department of Orthodontics and
Pediatric Dentistry, University of So Paulo, So Paulo, Brazil.
Disclosure. None of the authors reported any disclosures.
This study was supported by grant 141486/2014-7 from the Brazilian
National Council for Scientic and Technological Development (CNPq),
grant 2012/11846-5 from the So Paulo Research Foundation, and scholarships
303109/2012-2 (D.P.R.) and 308736/2012-5 (M.M.B.) from Research Produc-
tivity ScholarshipCNPq.
The authors thank Coordination for the Improvement of Higher Educa-
tion Personnel and the participants of the Post-Graduation in Pediatric
Dentistry Seminar of the School of Dentistry of the University of So Paulo
for the critical comments put forth.
1. Hickel R, Kaaden C, Paschos E, Buerkle V, Garca-Godoy F, Manhart J.
Longevity of occlusally-stressed restorations in posterior primary teeth.
Am J Dent. 2005;18(3):198-211.
2. Metz I, Rothmaier K, Pitchika V, et al. Risk factors for secondary caries
in direct composite restorations in primary teeth. Int J Paediatr Dent. 2015;
25(6):451-461.
3. Soncini JA, Maserejian NN, Trachtenberg F, et al. The longevity of
amalgam versus compomer/composite restorations in posterior primary
and permanent teeth: ndings from the New England Childrens Amalgam
Trial. JADA. 2007;138(6):763-772.
4. Qvist V, Poulsen A, Teglers PT, Mjr IA. The longevity of different
restorations in primary teeth. Int J Paediatr Dent. 2010;20(1):1-7.
5. Casagrande L, Dalpian DM, Ardenghi TM, et al. Randomized clinical
trial of adhesive restorations in primary molars: 18-month results. Am J
Dent. 2013;26(6):351-355.
6. Frencken JE, vant Hof MA, van Amerongen WE, Holmgren CJ.
Effectiveness of single-surface ART restorations in permanent dentition: a
meta-analysis. J Dent Res. 2004;83(2):120-123.
7. de Amorim RG, Leal SC, Frencken JE. Survival of atraumatic restor-
ative treatment (ART) sealants and restorations: a meta-analysis. Clin Oral
Investig. 2012;16(2):429-441.
8. Raggio DP, Hesse D, Lenzi TL, Guglielmi CAB, Braga MM. Is
atraumatic restorative treatment an option for restoring occlusoproximal
caries lesions in primary teeth? A systematic review and meta-analysis. Int
J Paediatr Dent. 2013;23(6):435-443.
9. Hara AT, Magalhes CS, Serra MC, Rodrigues AL Jr. Cariostatic effect
of uoride-containing restorative systems associated with dentifrices on
root dentin. J Dent. 2002;30(5-6):205-212.
10. Serra MC, Cury JA. The in vitro effect of glass-ionomer cement
restoration on enamel subjected to a demineralization and remineraliza-
tion model. Quintessence Int. 1992;23(2):143-147.
11. Yengopal V, Mickenautsch S. Caries-preventive effect of resin-
modied glass-ionomer cement (RM-GIC) versus composite resin: a
quantitative systematic review. Eur Arch Paediatr Dent. 2011;12(1):5-14.
184 JADA 147(3) http://jada.ada.org March 2016
12. Mickenautsch S, Yengopal V, Leal SC, Oliveira LB, Bezerra AC,
Bonecker M. Absence of carious lesions at margins of glass-ionomer
and amalgam restorations: a meta-analysis. Eur J Paediatr Dent. 2010;10(1):
41-46.
13. Ripa LW, Leske GS. Two years effect on the primary dentition of
mouthrinsing with a 0.2% neutral NaF solution. Community Dent Oral
Epidemiol. 1979;7(3):151-153.
14. Vanderas AP, Manetas C, Koulatzidou M, Papagiannoulis L. Pro-
gression of proximal caries in the mixed dentition: a 4-year prospective
study. Pediatr Dent. 2003;25(3):229-234.
15. Angker L, Nockolds C, Swain MV, Kilpatrick N. Quantitative analysis
of the mineral content of sound and carious primary dentine using BSE
imaging. Arch Oral Biol. 2004;49(2):99-107.
16. Lenzi TL, Guglielmi Cde A, Arana-Chavez VE, Raggio DP. Tubule
density and diameter in coronal dentin from primary and permanent
human teeth. Microsc Microanal. 2013;19(6):1445-1449.
17. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for
reporting systematic reviews and meta-analyses of studies that evaluate
health care interventions: explanation and elaboration. J Clin Epidemiol.
2009;62(10):e1-e34.
18. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews
of Interventions: Version 5.0.1 [updated March 2011]. Available at: www.
cochrane-handbook.org. Accessed October 6, 2015.
19. Donly KJ, Segura A, Kanellis M, Erickson RL. Clinical performance
and caries inhibition of resin-modied glass ionomer cement and amalgam
restorations. JADA. 1999;130(10):1459-1466.
20. Dutta BN, Gauba K, Tewari A, Chawla HS. Silver amalgam
versus resin modied GIC class-II restorations in primary molars:
twelve month clinical evaluation. J Indian Soc Pedod Prev Dent. 2001;
19(3):118-122.
21. Taifour D, Frencken JE, Beiruti N, vant Hof MA, Truin GJ.
Effectiveness of glass-ionomer (ART) and amalgam restorations in
the deciduous dentition: results after 3 years. Caries Res. 2002;36(6):
437-444.
22. Andersson-Wenckert I, Sunnegardh-Grnberg K. Flowable resin
composite as a class II restorative in primary molars: a two-year clinical
evaluation. Acta Odontol Scand. 2006;64(6):334-340.
23. Ersin NK, Candan U, Aykut A, Onag O, Eronat C, Kose T. A clinical
evaluation of resin-based composite and glass ionomer cement restorations
placed in primary teeth using the ART approach: results at 24 months.
JADA. 2006;137(11):1529-1536.
24. Daou MH, Tavernier B, Meyer JM. Clinical evaluation of four
different dental restorative materials: one-year results. Schweiz Monatssch
Zahnmed. 2008;118(4):290-295.
25. dos Santos MP, Passos M, Luiz RR, Maia LC. A randomized
trial of resin-based restorations in Class I and Class II beveled
preparations in primary molars: 24-month results. JADA. 2009;140(2):
156-166.
26. Hilgert LA, de Amorim RG, Leal SC, Mulder J, Creugers NH,
Frencken JE. Is high-viscosity glass-ionomer-cement a successor to
amalgam for treating primary molars? Dent Mater. 2014;30(10):
1172-1178.
27. Cvar J, Ryge G. Criteria for the Clinical Evaluation of Dental
Restorative Materials. San Francisco, CA: U.S. Department of Health,
Education, and Welfare, U.S. Public Health Service 790244; 1971:1-42.
28. Tyas MJ. Placement and replacement of restorations by selected
practitioners. Aust Dent J. 2005;50(3):81-89.
29. Mickenautsch S, Yengopal V. Demineralization of hard tooth tissue
adjacent to resin-modied glass-ionomers and composite resins: a quan-
titative systematic review. J Oral Sci. 2010;52(3):347-357.
30. Alm A, Wendt LK, Koch G, Birkhed D. Prevalence of approximal
caries in posterior teeth in 15-year-old Swedish teenagers in relation
to their caries experience at 3 years of age. Caries Res. 2007;41(5):
392-398.
31. Kemoli AM, Van Amerongen WE. Effects of oral hygiene, residual
caries and cervical Marginal-gaps on the survival of proximal atraumatic
restorative treatment approach restorations. Contemp Clin Dent. 2011;2(4):
318-323.
32. Karantakis P, Helvatjoglou-Antoniades M, Theodoridou-Pahini S,
Papadogiannis Y. Fluoride release from three glass ionomers, a compomer
and a composite resin in water, articial saliva and lactic acid. Oper Dent.
2000;25(1):20-25.
 ORIGINAL CONTRIBUTIONS
33. Vermeersch G, Leloup G, Vrenen J. Fluoride release from glass ion-
omer cements, compomers and resin composites. J Oral Rehabil. 2001;
28(1):26-32.
34. Cruz JB, Lenzi TL, Tedesco TK, Guglielmi Cde A, Raggio DP. Eroded
dentin does not jeopardize the bond strength of adhesive restorative ma-
terials. Braz Oral Res. 2012;26(4):306-312.
35. Pallesen U, van Dijken JW, Halken J, Hallonsten AL, Higaard R.
A prospective 8-year follow-up of posterior resin composite restorations in
permanent teeth of children and adolescents in Public Dental Health
Service: reasons for replacement. Clin Oral Investig. 2014;18(3):819-827.
36. Wiegand A, Buchalla W, Attin T. Review on uoride-releasing
restorative materials: uoride release and uptake characteristics, antibac-
terial activity and inuence on caries formation. Dent Mater. 2007;23(3):
343-362.
37. Ricketts DN, Kidd EA, Innes N, Clarkson J. Complete or ultracon-
servative removal of decayed tissue in unlled teeth. Cochrane Database
Syst Rev. 2006;19(3):CD003808.
38. Wilkins T, Gillies RA, Davies K. EMBASE versus MEDLINE for
family medicine searches: can MEDLINE searches nd the forest or a tree?
Can Fam Physician. 2005;51:848-849.
JADA 147(3) http://jada.ada.org March 2016 185