Journal of Orthodontics, Vol. 33. 2006. 198-204

SCIENTIFICSECTION

A clinical comparison of bracket bondfailures in association with direct andindirect bondingS. ThiyagarajahDepartment of Orl hod unties. School of Dentistry, Birmingham, UK

D. J. SparyQueens Hospital, Burion upon Trent. UK

W. p. RockDcpartmunl of Orlhodonlics. School of Dentistry, Birmingham, UK

Objective: To compare bond failure rates between direct and indirect techniques for bonditig orthodontic brackets.

Design: A two-centre single blinded prospective randomized controlled clinical trial.

Materials and methods: This study was undertaken at the Birmingham Dental Hospital and Good Hope Hospital, SuttonColdfield. Thirty-three subjects meeting the inclusion criteria were selected from orthodontic waiting lists and assigned toeither of two study groups according to a split-mouth study design. The number and site of bracket failures between toothtypes was recorded over I year. Statistical analysis was carried out using chi-square tests.

Results: Brackets were lost from 14 of the 553 teeth bonded, giving an overall bond failure rate of 2.5%. There were nosignificant differences in bond failures between direct and indirect bonding or in the tooth types of the failures.

Conclusions: There was no significant difference in the bond failure rates between direct and indirect bonding.

Key words: Bracket bonding, randomized clinical trial

Received 25th April 2005: accepted 28th Mtireh 2006

Introduction

Accurate bracket placement is essential to the properfunctioning of a pre-adjusted appliance and this may beaided by the indirect bonding technique, which involvesplacement of brackets in optimal positions on plastermodels of the patient's dentition, and then transferringthem to the mouth via a tray so that they can then bebonded to the teeth in positions predetermined in thelaboratory.

Silverman and Cohen' first described the method,using a methylmethacrylate adhesive in combinationwith light-cured bis-GMA resin. The later Thomastechnique"""' advocated placement of resin paste onto thebracket bases as part of the laboratory procedure. Atransfer tray was made from flexible material thatpreserved the bracket positions on the model teeth andthe set composite was bonded to the teeth using a two-part unfilled resin.

However, a disadvantage of chemically-cured resins isthat the uneven rate of polymerization produced by

Address for correspondence: Dr W. P, Rock, School of Dentistry,St, Chad's Queensway. Birmingham. B4 6NN. UK.Email: w.p.rockfabham.ac.uk<D 2006 British Orthodontic Sodely

loading the bracket bases at different times may producean increase in air inclusions within the adhesive."* Theuse of opaque trays also meant that only self-curedcomposites could be used and improper seating of thetray was not revealed until after tray removal. Thedevelopment of transparent trays'* ** made possible theuse of light-cured composites, which are more easilyremoved from around the brackets after setting.

The next stage of deveiopment was the use of adhe-sive pre-coated brackets, which made efficient useof laboratory time and kept contamination to aminimum.^ ''

A heat-cured fluoride-releasing indirect bonding sys-tem has also been described,'^' although several chn-icians who used this technique have reported problemswith bracket float while heating the resin, since themodels have to be heated to 350 C for 30 minutes inorder to cure the resin. Furthermore, as ceramicbrackets cannot be exposed to such heat, they couldnot be placed at the same time as metal brackets.^Special indirect bonding adhesives are now available for

DOl 10.1179/146531205225021615

3O September 2006 Scientific SetnonBracket bond failures and direct and indirect bonding 199

final positioning of brackets on the teeth. Theseadhesives are chemically-cured and have short workingtimes, the argument being that light-cured compositesare not needed at this stage as an unlimited workingtime is not necessary.

The use of antisialogogues has been recommended toreduce moisture contamination when using the indirectbonding technique.""^ Moisture control is also improvedif the transfer tray is correctly trimmed so that it doesnot extend further on the model than the gingivalmargins of the teeth."*

Two clinical trials have compared bond failure ratesusing indirect and direct bonding techniques. In onestudy, 2.5% of directly bonded brackets were lost, while14% of indirectly bonded brackets failed," The indirecttechnique was considered inferior due to the greaternumber of brackets lost, and also because of theincreased time required ibr bracket placement andremoval of excess adhesive flash around the bracketbases. The higher failure rate was thought to be due tothe chemically-cured composite adhesive used and totechnique variations. Composite was placed ontothe bracket bases in the transfer tray immediatelybefore this was seated in the mouth so that pooradaptation or uneven pressure may have producedan uneven thickness of adhesive, resulting indecreased bond strength and. therefore, an increasedfailure rate."*

A second study used a chemically activated bondingsystem and assessed bracket failure after 3 months.Fewer brackets were lost than in the previous study,failure rates being 4.5% for the indirect technique and5.3% for the direct technique.'"^

A laboratory study using the Thomas techniquecompared bond strengths for the two bonding methodsusing 41 extracted human premolar teeth in combina-tion with a self-cured composite.'"* Although 65% of theindirectly bonded teeth had marginal voids, there wereno significant differences in bond strengths between thetwo groups.

In another study, an overall bracket loss rate of 6.5%was found over a period of 30 months when 407brackets were placed indirectly using a light-curedadhesive."* These results are comparable with thoseobtained in previous trials, which compared light-curedmaterials with chemically activated composites andfound similar overall bond failure rates.'"*''

Finally, a clinical comparison of two chemically-curedadhesives with the indirect bonding technique resulted inan overall failure rate of 5,6%i.'̂

In view of such variations in previous studies, the aimsand objectives of this study were therefore to:

test for differences in bond failure rates between directand indirect bonding techniques;test for differences in the tooth type o{ bond failuresbetween the two techniques.

Materials and methods

Study design and sample selection

This was a two centre prospective randomized con-trolled trial in which one clinician bonded all bracketsfor 33 consecutive subjects aged between 12 and 15 yearswith a variety of malocclusions. Subjects were selectedfrom the fixed appliance waiting lists at the BirminghamDental Hospital and Good Hope Hospital. SuttonColdfield and treatments began between April 2002and March 2003. No potential subject refused consent.Subjects were included if they required orthodontictreatment with full upper and lower pre-adjusted edge-wise appliances and the teeth to be bonded showed nosigns of caries, large restorations, fluorosis, hypoplasiaor abnormalities of crown morphology, which may haveaffected bracket bonding.

Sample size was based on the number of teeth neededto demonstrate statistically significant differencesbetween direct and indirect bond failures and wasdetermined using a sample calculation software pack-age. nQucry®. Using data from two previous studies o'(similar design'"^ ''̂ the proportions of bracket failures indirectly and indirectly bonded groups, respectively, wereestimated to be 0,033 and 0.107. It is acknowledged,however, that ultimately analysis was based on quad-rants (dependent units) within individuals (Figure 1),but as noted by Mandall et al.^'^ there is little useful dataavailable as only three trials were identified which metall the criteria with which to compare. Based on thedifference in these proportions (odds ratio of 3,511), atwo group continuity corrected chi-square test suggesteda sample size of 271 teeth per group at the P<0.05significance level and a power of 90%).

Subjects who fulfilled the inclusion criteria wereenrolled into the study and consecutively allocated anumber at the time of record collection. A CONSORTdiagram showing the flow of participants through eachstage of the study is shown as Figure 2, Subjects wererandomly allocated into one of two split mouth designsusing a randomization table, as shown in Figure l.'^

Allocation to a group was made randomly to reducethe possible effect of variability in cooperation andaccess in individual subjects as well as any operator bias(e.g. a right-handed operator may find it easier to bondthe right hand side of the mouth).

200 S. Thiyagarajah el al. Scientific Section JO September 2006

Group One

Right

Group Two

Left Right

Upper Indirect

Lower Direct

Direct

Indirect

Figure 1 Split-mouth design

The randomization table was also used to decide theorder in which quadrants were bonded in order to avoidbias that may have arisen from using the same techniquefirst in every subject.

Data analysis

Between-group differences were examined using chi-square. When analyzing the data we had a large numberin each group and the statistician advised that acontinuity correction would not have an impact. Itwas therefore not used as it had been during sample sizecalculation when numbers were unknown.

Ethical approval

Ethical approval was obtained by North and SouthBirmingham local ethics committees (LREC 655.02 and

Upper Direct

Lower Indirect

Indirect

Direct

LREC 0835). Parents were given an information leafletand written consent for entry into the trial was alsoobtained.

Record collection

A split-mouth design was randomly allocated at the timeworking records were taken and subjects were treatedconsecutively.

The indirect bonding technique (laboratory stage)

Models were cast on the same day as impressiontaking to ensure accurate fit of the transfer traysand trimmed so thai they were no higher than2 cm. to allow easy use of the vacuum formingapparatus.

Assessed for eligibility

Randomized to Group

Excluded (n = 0)Not meeting inclusion criteria (n -0)

Refused to participate (n ^ 0)Other reasons (n = 0)

or 2

Group I (n=l7) Group 2 (n=16)

Discontinued TrT (n Discontinued TrT (n=0)

Analyzed (n=1 Analyzed (n=16)

Figure 2 A CONSORT diagram showing ihe flow of p;trlicipants ihrough each stage oi' the trial

JO September 2006 Scientific S^ff/o/iBracket bond fitilures and direct and indirect bonding 201

Figure 3 A tra> blank adapted to a model Figure 4 A completed trii>

Quadrants to be indirectly bonded were marked withvertical and horizontal pencil lines on each tooth toidentify the LA point."^

The appropriate pre adjusted edgewise bracket(MBT Versatile + Bracket System) was selected foreaeh tooth and a small amount of 3M Unitek laboratoryadhesive was placed onto the base. Each bracket wasthen positioned on its tooth and the adhesive wasallowed to dry for at least I hour before the next step.

Trays were made using a 0.45 mm thick blank ofDrufolen W^'^ transparent tray material. The trans-parency of the material allowed the use of light curing,which gave better control of working time. A circularblank was draped over a dry model and brackets. Theblank was first heated and then closely adapted to themodel by means of negative pressure using a vacuumforming apparatus (Drufomat^'^; Figure 3). After theDrufolen had cooled it was trimmed with a hotinstrument and removed from the model along withthe brackets that were contained within it. Finally, thetray was trimmed close to the gingival margins of theteeth and two vertical slits were made from the edge ofthe tray to the mesial and distal gingival wings of eachbracket in order to facilitate removal from the mouth(Figure 4).

Preparation for bonding

A similar method was employed to prepare the teethfor bonding whether a direct or indirect technique wasto be used. Each quadrant of teeth was prepared andbonded separately to minimize the risk of moisturecontamination. Teeth were polished for 5 s each using abristle brush in a slow speed hand piece with a slurry ofpumice and water. The teeth were then rinsed with anair/water spray until all traces of pumice had been

removed. A cheek retractor and a flexible saliva ejectorwere used for moisture control, and cotton wool rollswere placed in the buccal and lingual sulci to improveisolation. The teeth were then dried with oil-freecompressed air for 5 s each and etched for 20 s withDeTrey® Conditioner 36 containing 36'^. phosphoricacid, in accordance with the tnanufacturer's instruc-tions. Each tooth was then rinsed thoroughly for 15 suntil all traces of the blue etching gel were removedbefore they were dried again with oil-free compressed airuntil they exhibited a frosty white appearance with notraces of moisture.

The indirect bonding technique {clinical stage)

Following the steps above a thin layer of Transbond "^XT primer was applied to the bracket bases and to theteeth in the quadrant to be indirectly bonded. A smallamount of Transbond XT light cure orthodonticadhesive was placed onto the base of each bracket andthe tray was seated with even pressure to allow goodadaptation of the brackets to the teeth and an eventhickness of composite resin (Figure 5). Molar bandswere fitted in all four quadrants only after bracketplacement, to ensure that accurate seating of the traywas not prevented.**

Care was taken to place a minimum amount ofcomposite resin onto each bracket base to avoidexcessive adhesive flash. Each bracket was cured usinga standard light source for 20 s, 10 s on the mesial and10 s on the distal aspect. Brackets were cured startingwith the most posterior tooth, then moving forwardsand the tray was then carefully removed using a flatplastic instrument (Figure 6). Excessive adhesive flashwas removed using a Mitchell's trimmer and rotaryinstruments if necessary.

202 S. Thiyagarajah ei ai Scientific Section JO September 2006

Figure 5 Placement of a tray in the mouthTray removal

The direct bonding technique

The teeth were prepared as before and TransbondXT^'^ primer was then painted onto each tooth andbracket base. A small amount of Transbond XT^'^composite was applied to each bracket base and thebracket was then positioned onto the LA point of thetooth. All brackets in the quadrant were positioned andexcess composite was removed before the curing lightwas applied. Each bracket was cured for 20 s, 10 s onthe mesial and 10 s on the distal aspect.

To minimize variation in the magnitude of orthodonticforces applied to the teeth, a similar initial 0.014-inchnickel titanium archwire was used in each case. At eachvisit, a record was kept of the tooth type, date andcircumstances of bracket bond failures. Only first timebond failures were recorded since it has been recom-mended that clinical studies evaluating bond failure ratesshould either only record first time failures or analyzemultiple failures at the same site in a different category."^'All subjects were observed over a period of I year.

Results

Thirty-three subjects entered into the trial with a meanage of 13 years and 7 months. One subject discontinuedtreatment, leaving 32 to complete the study. Five-hundred-and-sixty brackets were bonded. Seven of thebrackets placed indirectly required rebonding at the timeof placement and these were not included in the results.Omission ofthese immediate bracket failures left a totalof 553 brackets of which 14 were lost over the year, anoverall failure rate of 2.5% (Table 1). Bond failureoccurred for six indirectly bonded brackets andeight direct bonds, the difference was not significant,chi-square=O.33l, 7^=0.565. Inclusion of the seven

Figure 6 Placement of a tray in the mouthTray removal

early bond failures would have altered the statisticalsignificance of the between group difference only slightlychi-square= 1.025, P=O.31I)

Due to the small number of bracket failures, data weresubdivided into incisors, canines and premolars, ratherthan individual teeth. Comparisons were made betweenthe upper and lower arches, and the right and left sidesof the mouth.

Overall, there were eight bond failures on incisors andsix bond failures on premolars (Table 2). There wereeight bracket failures in the upper arch and six failuresin the lower. Premolar bracket failure was equal in botharches (three) and there were no canine bracket failures.There were five incisor bracket failures in the upper archand three in the lower. There were six failures on theright side of the mouth and eight failures on the left.

Four directly bonded brackets failed in the upper archand four failed in the lower. Three of the direct bondfailures were on the right side of the mouth and five onthe left. Five of the direct bond failures were on incisorteeth and three on premolars (Table 2).

After indirect bonding, four brackets were lost in theupper arch and two from the lower. Three brackets werelost from each side of the mouth and three brackets werelost from both incisors and premolars (Table 2).

Sixty-six per cent of the failures following indirectbonding occurred in the first 6 months, while with the

Table 1 Bond failures with direct and indirect bonding

Indirect bondingDirect bondingTotal

Bonds

No.

273266

survived

%

97.897.1

Bonds

No.

68

14

failed

/n

2.22.9

Key:

JO September 2006 Scientific S^cf(o«Bracket bond failures and direct and indirect bonding 203

direct method, 43% were lost in this same time period.Overall. 50"/> of the bond failures occurred in the first6 months after placement (Table 3). The remainderoccurred later, thus suggesting a relatively constant hazard.

Discussion

This clinical trial found no significant difference in thenumber of bracket failures that followed direct andindirect bracket placement, respectively. The study useda split-mouth design in order to remove differences thatmay have existed between the subjects from comparisonof the effectiveness of direct and indirect bonding.'*^ Apotential disadvantage of the split mouth design is thattreatments applied to one side of the mouth may havecarry-across effects on contralateral teeth (this situationis avoided if the patient is randomized to a treatment-type^"). In the present study, the transfer trays, whichcarried the brackets for indirect bonding containedbrackets on one side only. It impossible to say whetherthis facilitated or complicated tray placement, althoughthe results suggest that there was no effect, since resultswere similar on the right and left sides of the mouth.

Bond failure rates of 2.2%i for the indirect and 2.9% forthe direct technique are lower than found in previousstudies of indirect bonding, which reported an overallfailure rate of 5.6'VI. for two chemically-cured compositebonding resins."' The low numbers of bond failuresrecorded with each bonding system in the present trialmay be due to the careful bonding technique employed.

Table 2 Bond failures according to site and bonding meihod

Bonds survived Bonds failed

Indirect Direct Indirect Direct

Site of bond failures No. % No. % No. % No. %

Since the numbers of bracket failures were low. only simplestatistical analyses have been used in the results section.

Our results are comparable with those of Aguirre.'" inthat there was no statistically significant differencebetween the number of bond failures following directand indirect bonding, respectively, although they differfrom the finding of Zachrisson and Brobakken whoreported a failure rate of \4% for indirect bonding and2.5% for the direct method."

However, it is difficult to make direct comparisonssince this last study used four different combinations ofbonding techniques, adhesives and bracket bases foreach patient.

Overall bond failure rates for light-cured compositesused with a conventional two-stage bonding system havebeen reported to be between 2.9 and 23% in randomizedcontrolled trials.'''•'^•^'""^ However, again it is difficultto make direct comparisons of bracket failure ratesbetween different studies due to variations in materials,research design and trial duration.

An observation period of 12 months following bracketplacement should give a reasonable estimate of the long-term performance of a bonding system, since other workhas shown that most failures occur within the first6 months.'"^

Indirect bonding technique

When using indirect bonding, it is essential that thecorrect amount of adhesive is placed on the bracketbases before seating the tray, since subsequentremoval of excessive set adhesive f̂ ash can provedifficult, especially with chemically-cured composites."''Adhesive flash became less of a problem as the operator(ST) became more proficient in the technique.

Care must be taken to seat the tray properly and toapply even pressure over brackets when light curing.Otherwise, there is a danger that an uneven thickness ofcomposite on a bracket base may weaken the bond and

Upper arch

Incisors

Canines

PremolarsTotalLower arch

IncisorsCanines

PremolarsTotalRighttleft

RightLeft

Key: No. —number

6128

48137

623149

142

141138

96.70

95.897.1

98.40

9898.6

97.997.8

5828

47133

6131

49141

133141

94.80

97.997

96.70

95.997.2

97.796.5

20

24

10

12

33

3.30

4.22.9

1.60

2.01.4

2.12.2

3

0

14

20

24

35

5.2

0

2.13.0

3.30

4.12.8

2.33.5

lead to bond failure at the t

Table 3 Bond failures in relation

Time following bonding (days)

0-9091-180

181-270271-362

Key: No. = number

ime of tray removal.

lo time inlerval

Bond failures

Indirect

No. %

3 1.11 0.4

2 0.70

following

Direct

No.

2I

4

1

bonding

%

0.70.4

1.5

0.4

204 S. Thiyagarajah el iil. Scientific Section JO Septetnber 2006

It has been suggested that an advantage of indirectbonding is its ability to isolate teeth from moisture

contamination. " This is attributed to the coverageafforded by the close-fitting transfer tray, whichimproves moisture isolation in the posterior segments.

It has been widely recognized for many years thataccurate bracket positioning is of critical importance torealizing the full potential of a pre-adjusted edgewiseappliance."^ Indirect bonding allows more accuratebracket placement''' with less placement ^̂

than is possible when using the direct system.

Conclusions

• There is no difference in bond failure rates between

direct and indirect bonding.

• The site of bond failure with regards to tooth type

does not vary between the two techniques.

A poster describing this project was awarded the GunterRussell Prize at the British Orthodontic Conference of2004.

Contributors

David Spary developed the clinical technique andPeter Rock designed the study. All clinical work wascarried out by Shanthi Thiyagarajah. who collected andanalyzed the data. Materials were provided by 3MUnitek. Peter Rock is the guarantor.

Acknowledgement

We are grateful to Dr M.valuable statistical advice.

S. Haque who provided

References

1. Silverman E, Cohen M. A universal direct bonding systemfor both metal and plastic brackets. Am J Orihod 1972; 62:236-44.

2. Thomas RG. Indirect bonding: simplicity in action. J ClinOrthod 1979; 13: 9-V106.

3. Shiau JY, Rasmussen ST, Phelps AE, Enlow DH, WolfGR.Bond strength of aged composites found in brackets placedby an indirect technique. An^le Orihod 1993: 63: 213-20.

4. Read MJF, O'Brien KD. A clinical trial of an indirectbonding technique with a visible light-cured adhesive. Am JOrihod Dcniofacial Orihop 1990: 98: 259-62.

5. Kasrovi PM, Timmins S. Shen A. A new approach toindirect bonding using light-cure composites. Am J OrthodDentofacUil Orihop 1997: HI: 652-56.

6. Read MJF, Pearson Al. A method for light-cured indirectbonding. J Clin Orihod 1998: 32: 502-03.

7. Kalange JT. Ideal appliance placement with APC bracketsand indirect bonding. / Clin Orihod 1999; 33: 516-26.

8. Sondhi A. Efficient and effective indirect bonding. Am JOrihod Dentofacial Orthop 1999: 115: 352-59.

9. Cooper RB. Sorenson NA. Indirect bonding wilh adhesiveprecoated brackets. J Clin Orihod 1993; 27: 164 67.

10. Sinha PK. Nanda RS, Ghosh J. A thermal-cured, fluoride-releasing indirect bonding system. J Clin Orthod 1995: 29:97 100.

11. Zachrisson BU. Brobakken BO. Clinical comparison ofdirect versus indirect bonding with different bracket typesand adhesives. Am J Orthod 1978: 74: 62 78.

12. Aguirre M. King G. Waldron J. Assessment of bracketplacement and bond strength when comparing directbonding to indirect bonding techniques. Am J Orihod1982:82: 269-76.

13. Hocevar RA. Vincent HP. indirect versus direct bonding:bond strength and failure location. Am J Orthod DcniofacialOrthop 1988:94: 367 71.

14. O'Brien KD. Read MJF. Sandison RJ. Roberts CT. Avisible light activated direct bonding material: an in vi\ocomparative study. Am J Orihod 1989: 95: 348 5i.

15. Sunna S. Rock WP. Clinical performance of orthodonticbrackets and adhesive systems: a randomized clinical trial.Br J Orihod 1998; 25: 283-87.

16. Miles PG. Weyant RG. A clinical comparison oftwo chemically-cured adhesives used for indirect bonding../ (9(7/;w/2003: 30: 331-36.

17. Armitage P. Berry G. Stalislical Methods in MedicalResearch. 2nd edn. Oxford: Blackwell. 1987.

18. Andrews LF. The straight-wire appliance. Br J Orihod1979: 6: 125^3.

19. Hujoel PP. DeRouen TA. Validity issues in split mouthtrials. J Clin Periodontol 1992: 19: 625-27.

20. Mandall NA, Millett DT. Mattick CR. Hickman J.Worthington HV. Maclarlanc TV. Orthodontic adhesives:a systematic review. J Orihod 2002: 29: 205-10.

21. Sonis AL, Snell W. An evaluation of a 11 uoride-releasing.visible light activated bonding system for orthodontic bracketplacement. Am J Orihod Denlofacial Orihop 1989; 95: 306 11.

22. De Saeytijd C, Carels CEL, Lesaffre E. An evaluation of alight curing composite for bracket placement. F.ur J Orihod1994; 16: 541^5.

23. Lovius BBJ, Pender N, Hewage S, O'DowIing I, TomkinsA. A clinical trial of a light activated bonding material overan 18 month period. Br J Orthod 1987: 14: 11-20.

24. Read MJF. Indirect bonding using a light-cured adhesive.Br J Orihod 1987; 14: 137-41.

25. Andrews LF. Straight Wire: the concept and appliance. SanDiego: LA Wells. 1989.

26. Hodge TR. Spary DJ. Dhopatkar AA, Rock WP. Arandomised clinical trial comparing the accuracy of directversus indirect bracket placement. J Orthod2004; 31: 132-37.