Self-Ligating Brackets in Orthodontics: A Systematic Review

Self-ligating brackets (SLBs) are not new conceptually, having been pioneered in the 1930s. They have undergone a revival over the past 30 years with a variety of new appliances being developed. A host of advantages over conventional appliance systems have been claimed typically relating to reduced frictional resistance.1–4 

The most compelling potential advantages attributed to SLBs are a reduction in overall treatment time5,6 and less associated subjective discomfort.7 Other purported improvements include more efficient chairside manipulation8 and promotion of periodontal health due to poorer biohostability. Preliminary retrospective research has pointed to a definite advantage, with a reduction in overall treatment time of 4 to 7 months and a similar decrease in required appointments.5,6 Consequently, the use of SLBs has increased exponentially; over 42% of American practitioners surveyed reported using at least one system in 2008.9 This figure was just 8.7% in 2002.10 

Retrospective research may be confounded by a variety of factors including operator enthusiasm, different appointment intervals and archwire sequences, and multiple operators. However, prospective research relating to SLBs has emerged in recent years.

The purpose of this systematic review is to evaluate the clinically significant effects of SLBs on orthodontic treatment with respect to the quality of scientific evidence and the methodology of those reports. An understanding of clinical evidence on the impact of SLBs on orthodontic treatment would inform the orthodontist's decisions in relation to their choice of fixed appliance system.

To be included in the review, trials had to meet the following selection criteria:

• Study design: Randomized and controlled clinical trials.

• Participants: Patients with full arch, fixed orthodontic appliance(s) treated with SLBs or conventional brackets (CBs).

• Interventions: Fixed appliance orthodontic treatment involving SLBs or CBs.

• Outcome measures: The outcome measures were alignment efficiency, pain experience, arch dimensional changes, rate of orthodontic space closure, bond failure rate, and periodontal effects related to both SLB and CB systems.

The efficiency of arch alignment, subjective pain experience, arch dimensional changes, rate of orthodontic space closure, and periodontal effects related to both appliances were recorded. Dichotomous data on the attachment failure rate related to each appliance were also noted.

The following electronic databases were searched: MEDLINE via OVID (1950 to April 2009; see  Appendix), EMBASE (1980 to April 2009), and Cochrane Central Register of Controlled Trials (The Cochrane Library, 2009). Language restrictions were not applied. Unpublished or “gray” literature was searched using ClinicalTrials.gov (www.clinicaltrials.gov) and the National Research Register (www.controlled-trials.com) using the term, “orthodontic and bracket.” In addition, Pro-Quest Dissertation Abstracts and Thesis database was searched (www.lib.umi.com/dissertations) using “orthodontic*” and “ligat*.” Conference proceedings and abstracts were also searched. Authors were contacted to identify unpublished or ongoing clinical trials and to clarify data as required. Reference lists of the included studies were screened for relevant research.

Assessment of research for inclusion in the review and assessment of validity and extraction of data were performed independently and in duplicate by two authors who were not blinded to the authors or the results of the research. Disagreements were resolved by discussion.

Six key methodological criteria were assessed: sample size calculation, random sequence generation, allocation concealment, reporting of withdrawals, blinding of measurement assessment, and the use of intention to treat analysis. An overall assessment of risk of bias (high, medium, low) was undertaken for each included trial using Cochrane Collaboration criteria. When five or more quality items were met, studies were considered to have a low risk of bias; three or more had medium risk; studies fulfilling less than three criteria were deemed to have high risk of bias. Only those at low to medium risk of bias were to be considered for meta-analysis.

A data extraction form was used to tabulate data on the outcomes of interest. Pain intensity using a visual analog scale (VAS) was obtained at all available time intervals. Pain scores assessed by means other than a zero-to-100 VAS were to be equated with this scale by multiplying the original scale employed by an appropriate factor.

Heterogeneity of the clinical studies was gauged by referring to each study assessing treatment protocol, timing of data collection, and measurement technique. Statistical heterogeneity was to be assessed by inspecting a graphical display of the estimated treatment effects from the trials along with their 95% confidence intervals. Chi-squared and I-squared tests for homogeneity were undertaken prior to each meta-analysis. Meta-analyses would also be possible only on studies reporting the same outcome measures at similar time intervals. Mean differences, standard deviations, and 95% confidence intervals were to be calculated for individual trials and combined using a random-effects model. Where necessary, sensitivity analyses were to be done with regard to the individual quality criteria, risk of bias, and publication status.

Forty-three trials were initially deemed potentially relevant to the review, 42 being derived from MEDLINE via OVID and 1 study from the National Research Register11 (www.controlled-trials.com). Following detailed assessment, 13 satisfied the inclusion criteria. One of these was subsequently omitted following retrieval of the full-text article; the remaining 30 studies were also excluded. However, after we contacted the authors of published trials, a further five studies were included. Of the 17 papers selected, 6 were randomized controlled trials (Table 111–27).

Outcomes assessed include alleviation of irregularity using Little's irregularity index, subjective pain experience recorded using VASs, rate of orthodontic space closure, dimensional changes during orthodontic alignment, plaque retention, extent of root resorption developing during treatment, and attachment debond rate related to either appliance system.

The methodological quality of the trials considered in the review is presented in Table 2.11–27 A priori sample-size calculations were undertaken in just six of the studies.11,12,14,18,21,24 Generation of the random-number sequence was considered adequate in six trials using computer-generated random allocation.11,12,14,17,18,24 In many of the studies, allocation was performed using a quasi-random method, with consecutive subjects being alternated between appliances. Six trials had acceptable allocation concealment.11,12,14,17,18,24 Group allocation was not concealed in the split-mouth studies.13,19,20 

Outcome assessment was blind in five studies.12,14,20–22 There were no dropouts in six studies16,19,21-23,27; in studies with dropout, those lost to follow-up were reported on. However, statistical analysis was invariably per protocol with dropouts excluded from analysis. Overall, six studies were deemed to be at low risk of bias.11,12,14,17,18,24 

Five trials considered the efficiency of initial orthodontic alignment.12–16 One study used a three-dimensional measuring technique, making comparison unfeasible.12 The remaining studies used two-dimensional measurement13–16; one of these trials incorporated a split-mouth design allowing consideration of just four mandibular contact points.13 Alignment efficiency was assessed in the mandibular arch in all cases, with four studies confined to the lower anterior region and one study considering the arch from first molar to first molar.12 

Miles et al,13 Scott et al,14 and Miles15 followed similar treatment protocols with alignment efficiency assessed using Little's irregularity index in the mandibular arch recorded at similar intervals. Scott et al14 assessed changes in the irregularity index 8 weeks after appliance placement; Miles15 and Miles et al13 both assessed residual irregularity 10 weeks and 20 weeks after placement of appliances. However, two of the studies13,15 failed to include standard deviations and were at high risk of bias, precluding meta-analysis. Instead of measuring the amount of irregularity relieved in a given time frame, Pandis et al16 calculated the time taken for alignment of the lower anteriors to occur.

Four trials investigated subjective pain experience after initial placement of the appliances.11,13,17,18 Of these, one split-mouth study considered pain reports after both the first and second visits, with patients indicating which system was associated with the greatest discomfort.13 Data in three of the trials are presented as continuous pain scores from 0 to 100 on a 100-mm VAS.11,17,18 One trial reported pain scores at 15 time intervals11; two trials used four time points: 4 hours, 24 hours, 3 days, and 7 days after appliance placement. The findings from these studies conflicted slightly with one study reporting a tendency to less pain experience with Damon 3 SLBs, although this finding did not reach statistical significance.11 Reported pain peaked within 24 hours11,17,18 before subsiding to near baseline levels 7 days after appliance placement. Three studies11,17,18 were regarded as being at low risk of bias, and they reported similar outcomes permitting statistical comparison; pain scores at four analogous time intervals were extracted from each study to facilitate this.11 Pain intensity over the first 7 days was reported in three studies involving 160 patients, with 83 in the SLB group and 77 in the CB group. Patients in the SLB group reported a mean difference in pain intensity of 0.99 to 5.66 points lower than in the CB group, the greatest difference being reported 3 days after appliance placement (Figures 1–4). However, differences were not of statistical significance.

Two studies13,18 reported greater pain experience during chairside manipulation of self-ligating appliances. However, as the mechanisms of archwire engagement and disengagement are very different using SmartClip18 and Damon 2,13 it was felt that direct statistical comparison of this research finding would be invalid.

Two studies have considered failure of bonded attachments over 20 weeks13 and 12 months.19 The date used for assessing failure or time taken for failure to occur was not reported, and only first-time failures for each tooth were recorded. No significant differences were noted in the more extensive study.19 

Two trials have compared the impact of SLBs and elastomeric ligation on plaque retention.20,21 A split-mouth design was used in one study assaying plaque specimens harnessed 1 and 5 weeks after bonding.20 Longer term effects of bracket system on periodontal health and accumulation of debris has also been assessed.22 

Pellegrini et al20 investigated the influence of method of archwire ligation on plaque retention using ATP-driven bioluminescence to assess bacterial load. Mean streptococcal and total bacterial levels harvested from tooth surfaces were lower with the SLB (P < .05). A further prospective trial, however, failed to show an association between bracket type and bacterial load.21 This finding may reflect the different measurement technique employed involving estimation of salivary levels of Streptococcus mutans.21 

Furthermore, Pandis et al22 failed to demonstrate a link between bracket type and periodontal health following removal of orthodontic appliances. It appears that, while bracket type might influence bacterial load with appliances in situ, this effect may not be sustained after treatment.

In relation to the mandibular arch, Pandis et al,16 Fleming et al,24 and Pandis et al25 reported identical incisor proclination and intercanine expansion with both appliance systems during arch alignment. Statistically greater intermolar expansion with self-ligating appliances has been shown in the latter studies.24,25 Similar findings were not observed by Scott et al,14 although this study involved assessment after mandibular premolar extraction, precluding direct comparison. There were insufficient trials of low- or medium-bias risk in homogenous groups to allow meta-analysis of this outcome.

Only one study considered the rate of orthodontic space closure26 at intervals of 5 weeks until complete space closure was achieved. This study had an inadequate sample size, with 4 of 18 subjects (22%) failing to complete the study. Posted archwires were used on both sides; therefore, tooth movement on one side may not have been independent of the other.

Pandis et al,27 using panoramic radiographs, reported no mean difference in the amount of apical root resorption of the maxillary incisors with Microarch and Damon 2 systems. Similar results were obtained by Scott et al,14 who assessed changes in root lengths of mandibular incisors on periapical radiographs following arch alignment. The mean amount of resorption was slightly greater with the Damon 3 appliance (2.26 vs 1.21 mm), although the difference failed to reach statistical significance.

Most of the studies included were considered to be at low to medium risk of bias. However, a priori sample size calculations were reported in only six studies, increasing the risk of false negative outcomes. The method of randomization and allocation concealment was often inadequate or incompletely reported. Many studies used alternate allocation, which precluded concealment of the participant to group allocation. Seven studies reported no dropouts; in the remaining studies, dropouts were clearly outlined. A CONSORT flow diagram was included in just four studies.11,14,17,25 

Per-protocol analysis was used in all studies, with dropouts being excluded from statistical analysis. Intention-to-treat analysis would be a more appropriate technique ensuring consideration of all subjects initially randomized, maintaining the benefits of randomization throughout the trial. Further prospective research in this area should be reported in accordance with the CONSORT guidelines28; this will improve the quality of research studies, permitting further meta-analyses, and will make components of research including method of randomization and allocation concealment more transparent.

Meta-analysis of the influence of bracket type on pain experience confirmed that SLBs do not have a clinically significant bearing on subjective pain experience. The three studies included in the meta-analysis had discordant findings; one favored SLBs11 and the other two studies demonstrated little difference between appliance systems. We can only speculate as to why this discrepancy arose; all studies were of high methodological quality and were carried out in similar settings, with analogous age and gender distribution.11,17,18 The failure to highlight a significant bracket-related effect is compatible with previous research, which has failed to demonstrate a link between archwire material29 or dimension30 and pain experience. Clearly, pain is influenced by a variety of factors, with individual susceptibility being critical. Consequently, to definitively address this question, a well-designed, prospective study of a large sample is required.

Prospective research considering surrogate measures of treatment efficiency, including the efficiency of orthodontic alignment and rate of space closure, has shown little difference between fixed appliance types, with remarkable consistency. These findings are incompatible with retrospective research findings5,6 and with manufacturers' claims of superior clinical performance. However, statistical comparison of these studies was not performed in view of differences in measuring alignment, methodological inadequacies related to some of the research, and incomplete reporting of results.

Arch dimensional changes arising with SLBs and conventional systems appear to be similar: identical levels of incisor proclination and intercanine expansion developed in both systems.16,23–25 This outcome is at odds with claims that low-friction systems respond differently under soft tissue pressures. Nevertheless, two studies have suggested that greater mandibular intermolar expansion develops during alignment with SLBs.24,25 

The finding of lower bacterial and streptococcal loads surrounding SLBs compared with conventional brackets during the initial stages of orthodontic treatment is of interest.20 Longer term follow-up has highlighted the capacity of periodontal tissues to recover from this initial insult following appliance removal.22 Nevertheless, it is unclear whether increased plaque accumulation has other detrimental effects, particularly decalcification. Further research is required to investigate this relationship further.

While evidence regarding the clinical application of SLBs is beginning to accumulate, the influence of bracket type on oral health-related quality of life is uninvestigated. There has also been no direct prospective comparison of overall treatment duration with conventional brackets and SLBs. Further research should be reported in accordance with the CONSORT guidelines28 and should have adequate sample size to avoid Type II error.

• There is insufficient evidence to support the use of self-ligating fixed orthodontic appliances over conventional appliance systems or vice versa.

• SLBs do not confer particular advantage with regard to subjective pain experience.

• There is insufficient evidence suggesting that orthodontic treatment is more or less efficient with SLBs.

Dr Valeria Marinho for her kind help with the database search. Drs Nikalaos Pandis, Peter Miles, and Angus Pringle for providing further information and data on their research studies and Dr Pandis for providing access to unpublished material.

1 RANDOMIZED CONTROLLED TRIAL.pt. (272711)

2 CONTROLLED CLINICAL TRIAL.pt. (79394)

3 RANDOMIZED CONTROLLED TRIALS.sh. (0)

4 RANDOM ALLOCATION.sh. (64632)

5 DOUBLE BLIND METHOD.sh. (101818)

6 SINGLE BLIND METHOD.sh. (12964)

7 or/1-6 (408191)

8 (ANIMALS not HUMANS).sh. (3290537)

9 7 not 8 (377322)

10 CLINICAL TRIAL.pt. (453141)

11 exp Clinical Trial/ (576329)

12 (clin$ adj25 trial$).ti,ab. (160255)

13 ((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).ti,ab. (101760)

14 PLACEBOS.sh. (27991)

15 placebo$.ti,ab. (116671)

16 random$.ti,ab. (445674)

17 RESEARCH DESIGN.sh. (56256)

18 or/10-17 (970709)

19 18 not 8 (898433)

20 19 not 9 (534857)

21 9 or 20 (912179)

22 exp ORTHODONTICS/ (34694)

23 orthod$.mp. (39934)

24 22 or 23 (43726)

25 (bracket$ or brace$ or appliance$).mp. [mp = title, original title, abstract, name of substance word, subject heading word] (26892)

26 (self ligat$ or ligat$ or low friction$).mp. [mp = title, original title, abstract, name of substance word, subject heading word] (60515)

27 25 and 24 and 26 (317)

28 27 and 21 (42)