Banding Versus

8/10/2019 Banding Versus

1/9

SCIENTIFICSECTION

Banding versus bonding of firstpermanent molars: a multi-centrerandomized controlled trialMariyah Nazir, Tanya Walsh and Nicky A. MandallUniversity of Manchester, UK

Susie MatthewThe Abbey Orthodontic Practice, Dalton-in-Furness, UK

Dee FoxSt Annes Orthodontic Clinic, Lytham St Annes, UK

Objective: To assess the effectiveness of banding versus bonding of first permanent molars during fixed appliance treatment; interms of attachment failure, patient discomfort and post-treatment enamel demineralization.

Design: Multi-centre randomized clinical trial.

Setting: One District General Hospital Orthodontic Department and two Specialist Orthodontic Practices.Participants: Orthodontic patients aged between 10 and 18 years old, randomly allocated to either receive molar bands ( n5 40)or molar bonds ( n5 40).

Method: Bands were cemented with a conventional glass ionomer cement and tubes were bonded with light-cured composite toall four first permanent molar teeth for each subject. Attachments were reviewed at each recall appointment to assess looseningor loss. The clinical end point of the trial was the day of appliance debond. Enamel demineralization at debond was assessedusing the modified International Caries Assessment and Detection System (ICDAS).

Results: The first time failure rate for molar bonds was 18.4% and 2.6% for molar bands ( P 5 0.0002). Survival analysisdemonstrated molar bonds were more likely to fail compared with molar bands. First permanent molars with bonded tubesexperienced more demineralization than those with cemented bands ( P 5 0.027). There was no statistically significant differencein discomfort experienced by patients after banding or bonding first permanent molars ( P . 0.05).

Conclusion: This study shows that as part of fixed appliance therapy, American Orthodontics photoetched first permanent

molar bands cemented with 3M ESPE Ketac-Cem perform better than American Orthodontics low profile photo-etched andmesh-based first permanent molar tubes bonded with 3M Unitek Transbond XT in terms of failure behaviour and molarenamel demineralization.

Key words: Molar attachment failure, molar tubes, molar bands, randomized controlled trial, orthodontic enameldecalcification

Received 13th September 2010; accepted 19th February 2011

Introduction

The terminal attachments of xed appliances are placedon molar teeth; most commonly the rst permanentmolars. These attachments can take the form of acemented molar band or a bonded molar tube. Theactive phase of xed orthodontic appliance treatmenttakes an average of two years to complete. To reduce thelikelihood of emergency visits, improve patient experi-ence and avoid lengthy treatment times it is important

that these attachments have low failure rates. Indeed, it

has been suggested that Loose attachments lower morale,reduce protability and wreak havoc with scheduling 1 andwhilst a zero per cent attachment failure is not a realisticgoal, failure rates should be less than 5%. 1

Prior to the advent of enamel bonding techniques, theuse of orthodontic bands on rst permanent molar teethwas universal. Many orthodontists continue to favourmolar bands due to beliefs regarding lower failure ratesand reliability. 2,3 With improvements in band design

Journal of Orthodontics , Vol. 38, 2011, 8189

Address for correspondence: Mariyah Nazir, OrthodonticDepartment, University Dental Hospital Manchester, HigherCambridge Street, Manchester, M15 6FH, UK.Email: [email protected]#

2011 British Orthodontic Society DOI 10.1179/14653121141308


2/9

(micro-etching, innovative mechanical retention fea-tures) further failure rate reductions have followed. 4,5

Simultaneously, bonded molar tubes have becomeincreasingly popular as advances in attachment designand materials science have led to improved survivalbehaviour. 611 Advocates of molar tubes claim these

attachments are more efcient and convenient, allow foreasier maintenance of oral hygiene and reduce deminer-alization. However, there is little prospective clinicalliterature evaluating the success of bands compared withbonds on molar teeth using contemporary modernmaterials. 12 The majority of studies in this eld areretrospective or non-comparative. 1315 Although muchlaboratory based work has been conducted on molarattachment failure behaviour, there are problematicissues with applying in vitro ndings to the clinicalenvironment. 16 Flaws in study design and data analysisin the eld of bond and band failure research have beenhighlighted by others. 17

Enamel demineralization is a recognized iatrogenicrisk factor associated with xed appliance orthodontictreatment. 18 It has been reported that 50% of patientsundergoing xed appliance treatment develop at least onewhite spot lesion. 19 The aesthetic and dental health im-pact of such demineralization is recognized 20 and manyresearchers are attempting to identify effective interven-tions to reduce demineralization during orthodontictreatment. 21 Recent systematic reviews have recommendedbonding and banding studies should measure deminer-alization as a secondary outcome where possible in orderto improve the reporting quality of clinical trials. 22,23

The aim of this study was to assess whether there wereany differences between rst permanent molar bandsand rst permanent molar bonds in terms of:

1. Failure rate;2. Enamel demineralization (at debond);3. Patient discomfort at first recall.

The null hypothesis tested was that there is no differencebetween rst permanent molar bands and rst perma-nent molar bonds in terms of failure rate during xedappliance treatment, patient discomfort at rst recall orenamel demineralization at debond.

Subjects andmethods

Ethical approval was obtained from the NHS Multi-Centre Research Ethics Committee. All eligible patientswere given written information describing the trial, withthe opportunity to ask further questions. Writtenconsent was taken from patients recruited to the trial.The recruitment period extended from July 2005 to July

2007 and patients were followed until the end of treatment (appliance debond).

The study sample was taken from consecutive patientsattending three centres for xed appliance treatment:

N The Abbey Orthodontic Practice (Dalton-in-Furness) 05/MRE05/27;

N St Annes Orthodontic Clinic (Lytham St Annes) 05/MRE05/27;

N Tameside General Hospital Orthodontic Department(Ashton-under-Lyne) 05/MRE05/27.

An a priori sample size calculation determined that asample size for each group of 38 would be sufcient todetect a failure rate difference of 30% (odds ratio of 6.00)between the molar bands group (with a predicted failurerate of 10%) 24 and a molar bond group (with a predictedfailure rate of 40%) using a two group continuity correctedchi-square test with a 0.05 two-sided signicance level andpower of 80%. A 40% bond failure was based on clinicalopinion as there was no supporting literature. A totalsample size of 76 patients was required and to allow fordropouts, the decision was made to recruit 80 patients.

The appliances were placed by one operator at eachcentre.

Inclusion criteria

N Patients aged between 10 and 18 years old;N Patient starting orthodontic treatment with upper and

lower fixed appliances (pre-adjusted edgewise);N Patient and parent informed and written consent.

Exclusion criteria

N Lack of consent;N Absence of or planned extraction of first permanent

molars;N First permanent molars with evidence of deminerali-

zation or hypoplastic enamel;N Occlusion likely to debond bonded attachments;N Patients requiring orthognathic surgery;N Occlusions that require extra-oral or intra-oral

anchorage reinforcement (headgear, palatal arch,lingual arch) or precluded the use of bonds, e.g. useof a quad-helix appliance.

Random allocation and concealment

Randomization was carried out using random numbertables; even numbers were assigned to the molar bandgroup and odd numbers to the molar bond group. Arestricted randomization method in blocks of 10 was

82 Nazir et al. Scientific Section JO June 2011


3/9

employed to ensure equal numbers of patients wereallocated to each of the treatment groups. The randomallocation sequence was generated by one clinician.Participants were enrolled by individual operators at eachsite. The allocations were concealed in envelopes markedwith each subjects identication number and held in acentral place to which the operators did not have access.The operator and patient remained blind to the attachment

type until after the consent and registration procedures.In line with recent recommendations 21,23 a parallel

method rather than a split-mouth study design wasemployed to reduce any bias that may be introduced bycross-over effects. Therefore if a patient was allocated tothe bonding group for example, molar bonds wereplaced on all four rst permanent molars.

Clinical sequence

A standardized procedure was followed for each patientas follows:

N

Consent and registration;N A letter was sent to the patients GDP to inform them

of their participation in the study;N Baseline variables recorded including patient age,

gender, intervention group and malocclusion type;N Patients allocated to receive bands had separators placed

for 1 week followed by cementation of AmericanOrthodontics H (Sheboygan, WI, USA) photo-etchedmolar bands with 3M ESPE Ketac-Cem H (St. Paul,

MN, USA) mixed according to the manufacturersinstructions;

N Where patients were allocated to the bonding group,molars were isolated with cotton wool rolls andsuction; Kerr H (Orange County, CA, USA) 37%phosphoric acid enamel etchant was applied to thebuccal surfaces for 30 s. The teeth were then washedwith water from a three-in-one tip for at least 15 sbefore drying. 3M Unitek Transbond XT primer andlight cured composite were used to bond AmericanOrthodontics low profile molar tubes with photo-etched mesh bases to the molar teeth;

N The remainder of the upper and lower fixed appliances(Roth prescription) was then placed in the usual way andeach clinician followed their routine orthodontic treat-ment mechanics. Distal-end cutters were not used in themouth as this was likely to debond bonded attachments;

N All patients were instructed to avoid hard, sticky or sweetfoods. They were requested to use a 0.05% sodiumfluoride mouthwash daily for the duration of their

treatment.

Stopping criteria

In line with routine clinical practice the followingstopping rules and subsequent appliance removal wereapplied:

N Repeated failure to attend appointments;

N Sound tooth surface: Code 0; N First visual change in enamel upon drying (when seen wet there is no evidence of any change): Code 1;N Distinct visual change in enamel when viewed wet: Code 2; N Localized shallow enamel breakdown due to caries with no visible dentine:Code 3; N Localized deeper enamel breakdown due to caries with no visible dentine:Code 4; N Distinct cavity with visible dentine: Code 5

Figure 1 International Caries Detection and Assessment System (ICDAS) criteria

JO June 2011 Scientific Section Banding and bonding molars: a multi-centre RCT 83


4/9

N Repeated breakages or general lack of co-operation;N Poor oral hygiene;N Presence or development of clinically significant

demineralization, root resorption and/or caries;N Patient/parental request to stop treatment for any reason.

Clinical outcome measures

The primary outcome measure was molar band or molarbond failure. At review appointments, outright bond or

cement failure resulting in frank loss of the attachmentor attachment loosening due to partial bond or cementfailure, were both recorded as failure. When a failureepisode occurred the same type of attachment wasreplaced to allow further analysis of failure behaviour.

Post-treatment demineralization was assessed at debondby direct visual examination. 19,2528 Conventionally,clinical assessment of enamel demineralization related toorthodontic treatment has been evaluated by the absence/presence of white spot lesions. 19,27,29 With time, the

Figure 2 Trial prole for comparison of rst permanent molar bands and bonds



5/9

majority of such white spot lesions improve 30 and are notalways indicative of true demineralization. 31 Moreextensive enamel lesions are of greater concern as thereis less potential for remineralization. There is a lack of consistency among the contemporary demineralizationcriteria systems which limits the comparability of

measured outcomes. The International Caries Detectionand Assessment System (ICDAS) criteria have beendeveloped by an international team of caries researchersto produce one standard system for caries detection andassessment histological evaluation of extracted teeth hasdemonstrated that as ICDAS codes increase in severity,there is an increased likelihood of demineralization. 32 Themodied ICDAS was used at the chairside to record theextent of enamel demineralization on the buccal surfacesof the rst permanent molars at debond (Figure 1). Inaddition to being practical, the system has been proven tohave content, correlational and discriminatory validity. 33

Patient based outcomes

Patient discomfort was measured on a 7 point Likertscale at the rst review appointment after attachmentplacement, where 1 5 no discomfort and 7 5 extremediscomfort. This was a requirement of the local ethicscommittee, as they were concerned that one or othertreatment may signicantly increase patient discomfort.

Statistical data analysis

The data were checked for normality and to allowmeaningful comparisons with previous studies, only rsttime failures were analyzed. Subsequent failures wereevaluated, but not included in the survival analysis.

Where molar bonds were replaced with bands (repeatedfailures, anchorage requirements) an intention-to-treatanalysis was applied and all patients were analyzedaccording to the group to which they had been originallyrandomized.

To assess ICDAS reliability, 27 images were assessedand re-assessed a week apart by all operators. Theresults were then analyzed using a weighted kappastatistic.

Clustering of data within patients was accounted forby appropriate statistical analysis: the cluster adjustedchi-square test was used to analyze attachment failure atthe tooth level. The Mann-Whitney U test was used to

compare median levels of patient discomfort betweenthe two treatment groups and to analyze demineraliza-tion data. Survival analysis was used to estimate therelative risk of failure from the point of attachmentplacement until the end-point of the study, usingSTATA statistical software version 9.0.

Results

A CONSORT diagram illustrating the ow of patientsthrough each stage of the trial is shown (Figure 2).Baseline demographic and clinical characteristics arepresented in Table 1.

Agreement and repeatability

All operators showed at least 95% intra-rater agreementwith the gold standard ICDAS scores, demonstratinghigh levels of agreement. Additionally, weighted kappascores for intra-rater reliability between initial testing

Table 1 Distribution of patient characteristics at baseline for each

treatment group.

Molar bands group Molar bonds group

Patient variables n (%) n (%)

Total 38 (50.0) 38 (50.0)GenderMale 16 (42.1%) 14 (36.8%)Female 22 (57.9%) 24 (66.7%)Age1012 12 (31.6%) 12 (31.6%)1315 24 (63.2%) 20 (52.6%)1618 2 (5.3%) 6 (15.7%)

MalocclusionClass I 8 (21.1%) 13 (34.2%)Class II/I 22 (57.9%) 16 (42.1%)Class II/II 5 (13.2%) 6 (15.8%)Class III 3 (7.9%) 3 (7.9%)CentreAbbey 12 (31.6%) 11 (28.9%)St Annes 9 (23.7%) 11 (28.9%)TGH 17 (44.7%) 16 (42.1%)

Table 2 Cluster adjusted chi-square test to compare first timefailures between molar bands and bonds

Failure at tooth level

Failures No failures Totaln (%) n (%) n (%)

Bands 4 (2.6) 148 (97.4) 152 (100)Bonds 28 (18.4) 124 (81.6) 152 (100)

Inter-cluster correlation 5 0.0860.ICC 95% confidence intervals 5 [0.000, 0.0192].Pooled adjustment chi-square statistic 5 14.295 and P 5 0.0002.Group adjustment chi-square statistic 5 14.295 and P 5 0.0002.



6/9

and one week re-testing ranged from 0.95 to 0.96 (SE0.19200.1924), demonstrating good repeatability.

Failure outcomes

There was a total of 32 (10.5%) rst time attachment

failures and these occurred in 23 (60.5%) patients. Theproportion of rst time attachment failures was greaterfor molar bonds (18.4%, n5 28) than for molar bands(2.6%, n5 4). This nding was statistically signicant(P 5 0.002) using the cluster adjusted chi-squared test(group adjustment) (Table 2). Patients who had molarband failures only experienced one failure episode ( n5 4;10.5%), whereas multiple attachment failure episodeswere more common in the molar bonds group ( n5 10;26.3%). Analysis using the chi-square test showed nostatistically signicant differences in the number of patients experiencing attachment failures between trialcentres ( P 5 0.27); age groups ( P 5 0.90); genders(P 5 0.665) or between upper and lower arches ( P 5 0.453).

Discomfort outcomes

There was no statistically signicant difference in thediscomfort levels recorded at rst review followingattachment placement between the molar band andmolar bond groups in terms of upper arch ( P 5 0.754) orlower arch ( P 5 0.583) discomfort (Table 3).

Demineralization outcomes

Analysis at the patient level demonstrated that 44 (60%)patients developed some degree of demineralization(ICDAS code 1 and above). Sixteen (36%) bandedpatients experienced demineralization compared with 28(64%) bonded patients ( P 5 0.038) (Table 4). However,the majority of patients with demineralization onlydeveloped ICDAS code 1 lesions (rst visual change inenamel upon drying when seen wet there was noevidence of any change: n5 30; 68%). No patients

developed frank carious lesions (ICDAS codes 3, 4and 5).

Analysis at the tooth level demonstrated 98 (33%) rstpermanent molars developed some degree of deminer-alization (ICDAS code 1 and above). Thirty-six (24%)banded molars developed demineralization compared

with 62 (43%) bonded molars demineralization(P 5 0.027) (Table 5). The majority of teeth affected bydemineralization had ICDAS code 1 lesions n5 78 (80%),the remaining teeth had ICDAS code 2 lesions. No teethdeveloped frank carious lesions (ICDAS codes 3, 4 and 5).

Survival analysis

The mean total treatment time was 19 months. Theshortest course of xed appliance treatment was9 months and the longest course was 3 years and1 month. The minimum time to attachment failure was8 days and the maximum time was 310 days (10 months

and 10 days). The assumption of the Cox proportionalhazard was met (i.e. the hazard ratio is constant;P 5 0.888). Survival functions for the two attachmenttypes are illustrated by the Cox proportional hazardssurvival analysis (Figure 3). This demonstrates molarbonds were more likely to fail during xed appliancetreatment. The relative risk of failure for eitherattachment is low at any time point as demonstratedby the high survival function values on the y-axis.Although the hazard ratio was positive and thus, theprognosis worse for the molar bonds group, the ndingsshould be interpreted with caution as there were only asmall number of failures ( n5 36) despite a large numberof observations ( n5 304). In survival analysis, it is thenumber of failures which gives power to the test and inthis study that number was small.

Discussion

The main nding of this study was that during xedorthodontic appliance treatment bonds placed on rstpermanent molars are more likely to fail than bands.The relatively low attachment failure rates reported mayTable 3 Mann-Whitney U test to compare molar bands and bonds

in terms of arch discomfort.

n MedianMaximumpain score

Minimumpain score

UpperArch

Band 38 2 4 1Bond 38 2 6 1

z52 0.313, P 5 0.754LowerArch

Band 38 2 5 1Bond 38 2 6 1

z52 0.549, P 5 0.583

Table 4 Mann-Whitney U test to compare molar bands and bondsin terms of ICDAS score (patient level).

n Median

MaximumICDASscore

MinimumICDASscore

Band 38 0 2 0Bond 36 1 2 0

z52 2.078, P 5 0.038.



7/9

be due the application of stringent inclusion andexclusion criteria, the use of systematic standardizedclinical techniques or as a result of improvements inattachment design.

Strengths of the study

This was a real world study evaluating the effectivenessof banding and bonding rst permanent molars as partof xed appliance treatment. Conducting an effective-ness study allows data to be generalized with regard tothe rest of the orthodontic population where the sameattachments and adhesives are used. All patients werefollowed until the end of treatment as recommended byCochrane systematic reviews of banding and bondingstudies. It has been reported that the majority of attachment failures occur within the rst 6 months of treatment. 34 The current study demonstrated all failuresoccurred within 13 months of commencing treatment.Attachment failure rates do increase with time 11,12 andreporting of failure rates early in treatment mayunderestimate true failure rates. Each patient wasrandomly allocated to either molar bands or molarbonds at the outset of treatment. Other workers haveused split-mouth studies when investigating attachmentfailures. The split mouth design allows patients to act astheir own control, which reduces variability; however

disadvantages of this methodology are possible cross-over effects and one adhesive or cement contaminatingthe other materials performance. This may confoundthe results and limit the robustness of any ndings.

Weaknesses of the study

Considering a number of patients in the study experi-enced multiple bond failures we were unable to elicit anypredictor variables that could help identify vulnerablepatients at the start of treatment. Multivariate sub-group analysis was not planned at the outset, whichmeans that the study may have been insufcientlypowered to detect any failure predictors that did exist.Other workers have demonstrated the inuence of age,malocclusion and socio-economic background on thenumber of attachment failures. 12,15,24

Context and implications for clinical practice

Failure outcomes. The molar band failure rate in thisstudy (2.6%) compares favourably with the ndings of other workers. 2,3 Additionally, the molar bond failurerate (18.4%) was lower than that reported by othercomparative prospective 12 and retrospective studies. 14

This 16% difference in failure rates would suggest bandsare a more reliable rst permanent molar attachmentthan bonds during xed appliance treatment. A numberof patients experienced numerous molar bond failures;this was not always on the same rst permanent molar.We may have a cohort of patients who have a predilectionfor appliance wrecking and in such circumstances it may

be prudent to revert to bands on all molar teeth as soon assuch wrecking behaviour is observed.

Demineralization

It has been reported that almost 50% 19 of individualsundergoing xed appliance treatment develop enameldemineralization. In this study, 60% of patients and 33%of molar teeth developed enamel demineralization.However, the majority of lesions (80%) were limited toICDAS code 1. No patients experienced frank enamelbreakdown. Although the majority of enamel lesions willremineralize with low dose uoride therapy 21 they can

compromise aesthetics, particularly where anterior teethare affected. Appropriate and careful selection of patientswith good oral hygiene habits from the outset shouldallow for minimal demineralization experience. The lowerdemineralization codes observed in the molar band groupcompared with the molar bond group may be due to theuse of glass-ionomer banding cement, with its resultantbenecial local uoride releasing properties. 7,35 However,a recent systematic review investigating the effectiveness

Figure 3 Survival functions for molar bands and bonds (Coxproportional hazards regression)

Table 5 Mann-Whitney U test to compare molar bands and bondsin terms of ICDAS score (tooth level).

n Median

MaximumICDASscore

MinimumICDASscore

Band 152 0 2 0Bond 154 1 2 0

z52 3.757, P 5 0.027.



8/9

of orthodontic banding adhesives concluded there was alack of suitable evidence to allow specic recommenda-tions regarding the use of any one material overanother. 23 Studies comparing glass ionomer cement withcomposite resin for bonding have demonstrated reducedrates of enamel demineralization 26,27 however; this has

been at the expense of increased attachment failurerates. 17,36,37 Benson et al. 21 in their systematic reviewconcluded the current level of evidence demonstratingglass ionomer based adhesives for bracket bonding maybe more effective at preventing enamel demineralizationthan conventional composite resins is weak. Factorsother than cement/adhesive composition may affectdemineralization experience and site of lesions; such asuoride supplementation in the form of mouthwashesand varnishes as well the role of the right-handed or left-handed dominance when tooth brushing.

Patient discomfort

This study is the rst to report patient-based outcomesassociated with placement of rst permanent molarattachments. It would be tempting to postulate bandingwould be more uncomfortable for patients as theattachment physically surrounds the entire tooth surfaceand placement can involve trauma to the gingivae;however, no differences were demonstrated betweenbands and bonds, low levels of discomfort were reportedand patients tolerated both types of attachment well.

Conclusions

N First permanent molars bonded with 3M UnitekTransbond XT have a higher failure rate than firstpermanent molar bands cemented with 3M ESPEKetac-Cem.

N Bonded first permanent molars demonstrated higherlevels of post-treatment demineralization than bandedfirst permanent molars.

N No differences in discomfort were experienced bypatients when banding or bonding first permanentmolars as part of fixed appliance treatment.

ContributorsNicky Mandall was responsible for the study concept,ethical approval, data interpretation, critical revisionand nal approval of the article. Mariyah Nazir wasresponsible for study design, administration, recruit-ment and treatment, drafting, critical revision and nalapproval of the article. Dee Fox was responsible forrecruitment, treatment, data entry, organization and

accommodation for examiner training and reliabilityexercises. Susie Matthew was responsible for recruit-ment, treatment and data entry. Tanya Walsh wasresponsible for statistical analysis, data interpretation,critical revision and nal approval of the article.Mariyah Nazir is the guarantor.

References

1. Bennett RK. Loose brackets: minor inconvenience or profitvacuum? Ormco Clinical Impressions 2001; 10 : 2228.

2. Gillgrass TJ, Benington PC, Millett DT, Newell J, GilmourWH. Modified composite or conventional glass ionomer forband cementation? A comparative clinical trial. Am J Orthod Dentofacial Orthop 2001; 120 : 4953.

3. Clark JR, Ireland AJ, Sherriff M. An in vivo and ex vivostudy to evaluate the use of a glass polyphosphonate cementin orthodontic banding. Eur J Orthod 2003; 25 : 31923.

4. Millett DT, Duff S, Morrison L, Cummings A, GilmourWH. In vitro comparison of orthodontic band cements. AmJ Orthod Dentofacial Orthop 2003; 123 : 1520.

5. Hodges SJ, Gilthorpe MS, Hunt NP. The effect of micro-etching on the retention of orthodontic molar bands: aclinical trial. Eur J Orthod 2001; 23 : 9197.

6. Fricker JP. A new self-curing resin-modified glass-ionomercement for the direct bonding of orthodontic brackets invivo. Am J Orthod Dentofacial Orthop 1998; 113 : 38486.

7. Gaworski M, Weinstein M, Borislow AJ, Braitman LE.Demineralisation and bond failure: a comparison of a glassionomer and a composite resin bonding system in vivo. AmJ Orthod Dentofacial Orthop 1999; 116 : 51821.

8. Kula K, Schreiner R, Brown J, Glaros A. Clinical bond

failure of pre-coated and operator-coated orthodonticbrackets. Orthod Craniofac Res 2002; 5 : 16165.

9. Sharma-SayalSK,Rossouw PE, Kulkarni GV,Titley KC.Theinfluence of orthodontic bracket base design on shear bondstrength. Am J Orthod Dentofacial Orthop 2003; 124 : 7482.

10. Aljubouri YD, Millett DT, Gilmour WH. Six and twelvemonths evaluation of a self-etching primer versus two-stageetch and prime for orthodontic bonding: a randomizedclinical trial. Eur J Orthod 2004; 26 : 56571.

11. Manning N, Chadwick SM, Plunkett D, Macfarlane TV. Arandomized clinical trial comparing one-step and two-steporthodontic bonding systems. J Orthod 2006; 33 : 27683.

12. Banks P, MacFarlane TV. Bonded versus banded first

permanent molar attachments: a randomized clinical trial. J Orthod 2007; 32 : 12836.

13. Millett DT, Hallgren A, Fornell AC, Robertson M. Bondedmolar tubes: a retrospective evaluation of clinical perfor-mance. Am J Orthod Dentofacial Orthop 1999; 115 : 667 74.

14. Millett DT, Letters S, Roger E, Cummings A, Love J.Bonded molar tubes an in vitro evaluation. Angle Orthod 2001; 71 : 38085.



9/9

15. Pandis N, Christensen L, Eliades T. Long-term clinicalfailure rate of molar tubes bonded with a self-etchingprimer. Angle Orthod 2005; 75 : 100002.

16. Eliades T, Brantley WA. The inappropriateness of conven-tional orthodontic bond strength assessment protocols. EurJ Orthod 2000; 22 : 1323.

17. Miller JR, Mancl L, Arbuckle G, Baldwin J, Phillips RW. Athree-year clinical trial using a glass ionomer cement for thebonding of orthodontic brackets. Angle Orthod 1996; 66 :30912.

18. Ellis PE, Benson PE. Potential hazards of orthodontictreatment what your patient should know. Dent Update2002; 29 : 49296.

19. Gorelick L, Geiger AM, Gwinnett AJ. Incidence of whitespot formation after bonding and banding. Am J Orthod Dentofacial Orthop 1982; 81 : 9398.

20. Geiger AM, Gorelick L, Gwinnett AJ, Benson BJ. Reducingwhite spot lesions in orthodontic populations with fluoriderinsing. Am J Orthod Dentofacial Orthop 1992; 101 : 40307.

21. Benson PE, Shah AA, Millett DT, Dyer F, Parkin N, VineRS. Fluorides, orthodontics and demineralization: a sys-tematic review. J Orthod 2005; 32 : 10214.

22. Mandall NA, Hickmann J, Macfarlane TV, Mattick CR,Millett DT, Worthington D. Adhesives for fixed orthodon-tic brackets. Cochrane Database Syst Rev 2003; (2):CD002282.

23. Millett DT, Glenny AM, Mattick CR, Hickmann J,Mandall NA. Adhesives for fixed orthodontic bands.Cochrane Database Syst Rev 2007; (2): CD004485.

24. Millett DT, Gordon PH. The performance of first molarorthodontic bands cemented with glass ionomer cement Aretrospective analysis. Br J Orthod 1992; 19 : 21520.

25. Twetman S, McWilliam JS, Hallgren A, Oliveby A.

Cariostatic effect of glass ionomer retained orthodonticappliances: an in vivo study. Swed Dent J 1997; 21 : 16975.

26. Marcusson A, Norevall LI, Persson M. White spotreduction when using glass ionomer cement for bondingin orthodontics: a longitudinal and comparative study. EurJ Orthod 1997; 19 : 23342.

27. Chung CK, Millett DT, Creanor SL, Gilmour WH, FoyeRH. Fluoride release and cariostatic ability of a compomer

and a resin-modified glass ionomer cement used fororthodontic bonding. J Dent 1998; 26 : 53338.

28. Millett DT, McCluskey LA, McAuley F, Creanor SL,Newell J, Love J. A comparative clinical trial of acompomer and a resin adhesive for orthodontic bonding.Angle Orthod 2000; 70 : 23340.

29. Ghiz MA, Ngan P, Kao E, Martin C, Gunel E. Effects of sealant and self-etching primer on enamel demineralisation.Part II: an in-vivo study. Am J Orthod Dentofacial Orthop2009; 135 : 20613.

30. Millett DT, Nunn JH, Welbury RR, Gordon PH.Demineralisation in relation to brackets bonded with glassionomer cement or a resin adhesive. Angle Orthod 1999; 69 :6570.

31. Kanthathas K, Wilmott DR, Benson PE. Differentiation of developmental and post-orthodontic white spot lesionsusing image analysis. Eur J Orthod 2005; 27 : 16772.

32. Pitts N. ICDAS an international system for cariesdetection and assessment being developed to facilitate cariesepidemiology, research and appropriate clinical manage-ment. Community Dent Health 2004; 21 : 19398.

33. Ismail A, Sohn W, Tellez M, et al . The International CariesDetection and Assessment System (ICDAS): an integratedsystem for measuring dental caries. Community Dent Oral Epidemiol 2007; 35 : 17078.

34. Hegarty DJ, Macfarlane TV. In Vivo bracket retentioncomparison of a resin-modified glass ionomer cement and aresin-based bracket adhesive system after a year. Am J Orthod Dentofacial Orthop 2002; 121 : 496501.

35. Pascotto RC, Navarro MF, Capelozza-Filho L, Cury JA. Invivo effect of a resin-modified glass ionomer cement onenamel demineralization around orthodontic brackets. AmJ Orthod Dentofacial Orthop 2004; 125 : 3641.

36. Miguel JA, Almeida MA, Chevitarese O. Clinical compar-ison between a glass ionomer cement and a composite fordirect bonding of orthodontic brackets. Am J Orthod Dentofacial Orthop 1995; 107 : 48487.

37. Norevall LI, Marcusson A, Persson M. A clinical evalua-tion of a glass ionomer cement as an orthodontic bondingadhesive compared with an acrylic resin. Eur J Orthod 1996;18 : 37384.


Documents

Banding Versus