Introduction

Self-ligating brackets (SLBs) have been widely adopted at clinics in recent years under claims of improved orthodontic effectiveness and efficiency. However, their clinical superiority has not been strongly proven yet. First introduced by Stolzenberg in 1935 [27], they were described as ‘Russell lock’ edgewise attachment. Since then, many other similar designs have appeared, including the Ormco Edgelok in 1972, Forestadent Mobil-Lock and Orec Speed in 1980, ‘A’ Company Activa in 1986, TimeLock bracket in 1998, and the Damon 2 and In-Ovation brackets in 2000 [24]. According to the mechanisms of lid closure, all these designs can be classified into two main categories: active and passive brackets. Active self-ligating brackets (ASLBs) have a spring clip that presses against the archwire for rotation and torque control. By contrast, passive self-ligating brackets (PSLBs) have a slide that closes without encroaching on the slot lumen, thus, exerting no active force on the archwire. However, the orthodontists are speculating if SLBs truly deliver the purported advantages and whether the conventional brackets (CBs) should be substituted by SLBs in clinics [18].

To date, many studies have investigated the clinical efficacy and efficiency of SLBs compared with CBs through various methods. Several systematic reviews of SLBs have been conducted in an attempt to form a conclusion, although these studies have varied greatly in methods and results [3, 4, 6, 10, 32]. Of these, two studies performed in 2010 obtained different results [4, 10]. One included 6 randomized controlled clinical trials (RCTs) and 11 controlled clinical trials (CCTs) but firm conclusions were not made due to lack of evidence [4]. The remaining studies included 2 RCTs, 10 CCTs, and 4 cross-sectional studies, which found SLBs to be advantageous with less chair time and incisor proclination [10]. Two systematic reviews concentrated on the specific aspects of these two techniques [3, 32]. One study focused on the initial pain, number of visits, and treatment time, which included only 4 studies; another focused on canine retraction and anchorage loss, which included 2 RCTs and 4 CCTs [3]. One meta-analysis of in vitro studies found that SLBs produce lower friction than CBs [6]. Until now, the published systematic reviews and meta-analyses have been based on in vitro or in vivo CCTs including prospective and retrospective trials due to the scarcity of RCTs. Since few RCTs of SLBs have been published in recent years, collecting and analyzing them will provide useful information for orthodontists.

Therefore, the goal of this study is to identify RCTs that compare SLBs with CBs and to investigate the transversal dimensional changes, space closure, and orthodontic efficiency of each design, also considering initial alignment and treatment time in patients undergoing fixed orthodontic treatment.

Methods

A customized procedure was preconducted in this study. The study selection, risk of bias assessment, and data extraction were performed by two expert reviewers. Furthermore, any discrepancies between them were resolved through a conference with a third reviewer to attain consensus. The following inclusion criteria were used: randomized controlled clinical trials, randomized split-mouth studies, participants who received fixed orthodontic treatment with self-ligating and conventional brackets, an intervention group of active or passive self-ligating brackets, a control group of conventional brackets, and outcomes of effectiveness demonstrating dimensional change, orthodontic space closures and efficiency in terms of alignment and treatment time. Exclusion criteria were as follows: published studies with repetitive data, patients with other related or unrelated diseases that may affect the results, and patients without a full arch under orthodontic treatment. There was no limitation for selection regarding the publishing year or language.

Search strategy

Details of related published or ongoing studies were obtained through the following electronic databases: MEDLINE (1948 to December 2016), Cochrane Central Register of Controlled Trials (CENTRAL, issue 12, 2016), EMBASE (via OVID, 1974 to December 2016), World Health Organization International Clinical Trials Registry Platform (December 2016), Chinese Biomedical Literature Database (1978 to December 2016), China National Knowledge Infrastructure (1994 to December 2016) and Chinese Medical Journal Database (1989 to December 2016). A total of 19 relevant Chinese journals were manually searched based on their titles and abstracts. The reference lists of all included studies were also screened for applicable information.

The MeSH terms were combined with free text words as a search strategy, such as ‘active’, ‘passive’, ‘self-ligating’, ‘conventional’, ‘orthodontic,’ and ‘brackets.’ The Cochrane Highly-Sensitive Search Strategy was used to identify randomized trials. In the PubMed database, the terms used for searching were ‘Orthodontic Appliances’ [MeSH], ‘bracket’, ‘braces’ and ‘self-ligat’. Titles and abstracts of publications were scanned to find potentially eligible studies. Once relevant studies were identified, full texts were obtained for further consideration of these studies to obtain details.

Risk of bias assessment

The Cochrane Collaboration tool for risk of bias assessment on RCTs was adopted with the following domains: sequence generation, allocation concealment, blinding of outcome assessors, incomplete outcomes, selective data reporting, and other bias [13]. Factors such as blinding of clinicians and patients were omitted due to an inherent feasibility problem since orthodontists can expertly distinguish bracket types. A given study was considered to be ‘low risk of bias’ if all the pertinent domains were judged as ‘low risk,’ ‘unclear risk of bias’ if any domain was judged as ‘unclear risk,’ and ‘high risk of bias’ if any domain was judged as ‘high risk’.

Data collection and synthesis

Customized data extraction forms were independently developed by two reviewers and included the following items: article name, first author, published year of the study, demographic data of participants, settings, brands of SLB and CB, main inclusion criteria, treatment methods, time points, and results. Review Manager 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) was used for data analysis. Statistical heterogeneity was explored using the χ 2 test with a 10% significance level as the cutoff value due to its low power. The impact of statistical heterogeneity was quantified using the I 2 statistic. If I 2 > 50% and P ≤ 0.10, the causes of the heterogeneity were analyzed followed by a subgroup analysis. If heterogeneity was found to be high (I 2 > 50%), the random effects model was chosen for meta-analysis. Otherwise, the fixed effects model was adopted. For continuous data, the treatment effects were expressed as the mean difference (MD) at 95% confidence interval (CI). The statistical significance of the hypothesis test was set as P < 0.05 (by two-tailed z tests). A meta-analysis was performed on studies with similar methods and results, such as those that assessed the same measurement without high heterogeneity. Data that could not be pooled have been described herein. Subgroup analyses were addressed, depending on bracket type of included studies, to classify the self-ligating brackets into active and passive designs by the random effects model. Sensitivity analyses of study exclusion were adopted where heterogeneity may have been caused by one or several studies in the meta-analysis.

The GRADE profiler was employed to evaluate the quality of study methodology, precision of effect estimates, directness of evidence, risk of publication, and heterogeneity. The GRADE quality levels were set as high, moderate, low, and very low, and clinical recommendations were classified as strong or weak [12].

Results

Study selection and characteristics

A total of 114 published studies were identified by the electronic and manual search, 39 of which were considered eligible, and their full texts were assessed after screening their titles and abstracts. Finally, 17 studies with 536 participants in the SLB group and 440 participants in the CB group were included in the qualitative analysis, 11 of which qualified for the quantitative synthesis. The study inclusion process is depicted in Fig. 1.

Fig. 1
figure 1

Flowchart for study inclusion

Abb. 1 Flowchart zur Studienrekrutierung

Full size image

All included studies were randomized controlled clinical trials, 3 of which were three-arm parallel trials, 3 were split-mouth studies, and 11 were two-arm parallel trials. Three studies compared ASLB with CB, while 13 studies compared PSLB with CB, and 1 study compared both ASLB and PSLB with CB. The details of included studies are shown in Table 1.

Tab. 1 Details of included studies Tab. 1 Detaillierte Aufstellung der eingeschlossenen Studien
Full size table

Risk of bias within studies

A total of 11 studies were included in the meta-analysis and their risks of bias were evaluated. Two included studies had low risks of bias, while the other 9 studies had unclear risks of bias (Table 2).

Tab. 2 Risk of bias assessment Tab. 2 Einschätzung des Bias-Risikos
Full size table

Transversal dimensional expansion

Six studies that were included in this analysis investigated the dimensional changes in the arch [1, 7, 11, 21, 22, 25], 4 of which reported changes in the mandibular intercanine and intermolar arch width and compared PSLBs with CBs [7, 21, 22, 25]. Figures 2 and 3 show that the meta-analyses show a higher efficacy of CBs in expanding mandibular intercanine width (MD − 0.51, 95% CI − 0.85 to − 0.17), while SLBs are apparently more efficient in expanding mandibular intermolar width (MD 0.50, 95% CI 0.11–0.90). However, the sample size used in this study was small and the risks of bias of two of the included studies were found to be unclear. Clinical heterogeneity such as variations in the arch structures was also seen [21, 25]. Thus, a sensitive analysis was performed; the results of which did not indicate any significant statistical difference between PSLB and CB in regard to mandibular intermolar width change (MD 0.37, 95% CI − 0.09 to 0.83, Fig. 4). Two other studies have also described noticeable maxillary arch width changes [1, 11]. One study compared both ASLBs and PSLBs with CBs and found no apparent difference between them [11]. Another study comparing PSLBs with CBs in Class I maxillary constriction patients demonstrated the similarity in their functions during transverse dimension changes. However, in the study, a quad-helix appliance was used in the CB group but not in PSLB group indicating the presence of heterogeneity [1].

Fig. 2
figure 2

Meta-analysis of mandibular intercanine width change of passive self-ligating brackets (PSLB) and conventional brackets (CB). 95% CI 95% confidence interval, SD standard deviation

Abb. 2 Metaanalyse zur Änderung der intercaninen Distanz im Unterkiefer mit passiven selbstligierenden Brackets (PSLB) und konventionellen Brackets (CB). 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image
Fig. 3
figure 3

Meta-analysis of mandibular intermolar change of passive self-ligating brackets (PSLB) and conventional brackets (CB). 95% CI 95% confidence interval, SD standard deviation

Abb. 3 Metaanalyse zu transversalen Veränderungen im Molarenbereich im Unterkiefer mit passiven selbstligierenden Brackets (PSLB) und konventionellen Brackets (CB). 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image
Fig. 4
figure 4

Sensitive analysis of expanding mandibular molar width change of passive self-ligating brackets (PSLB) and conventional brackets (CB). 95% CI 95% confidence interval, SD standard deviation

Abb. 4 Sensitivitätsanalyse zur Expansion im Molarenbereich mit passiven selbstligierenden Brackets (PSLB) und konventionellen Brackets (CB). 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image

Space closure

Five studies selected for this analysis focused on space closure in orthodontic treatment [2, 16, 17, 26, 28]. Three of these were split-mouth studies detailing canine retraction velocity with maxillary first premolar extraction [2, 16, 17], while the meta-analysis showed no difference between SLB and CB (MD 0.13, 95% CI − 0.09 to 0.35, Fig. 5). Two other studies that investigated the space closure without analyzing the canine retraction rate were excluded from the meta-analysis [26, 28]. Of these two analyses, one was discontinued after 3 months due to the absence of differences between PSLB and CB [28]. This might have been due to variation in the individual patient responses to the applied force. The second study compared ASLB, PSLB, and CB with different observation intervals. No variation was seen in the bracket types but differences were observed in the treatment of maxilla and mandible [26].

Fig. 5
figure 5

Meta-analysis of orthodontic space closure of passive self-ligating brackets (PSLB) and conventional brackets (CB). 95% CI 95% confidence interval, SD standard deviation

Abb. 5 Metaanalyse zum Lückenschluss mit passiven selbstligierenden Brackets (PSLB) und konventionellen Brackets (CB). 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image

Efficiency of initial alignment

Six studies compared SLB with CB in terms of alignment efficiency [8, 19, 21, 25, 26, 29], 4 of which calculated the required time for initial alignment as a part of the meta-analysis [8, 21, 25, 26]. No prominent difference was observed between the results obtained for PSLB and CB (MD 26.48, 95% CI − 12.75 to 65.71). However, results of one study comparing ASLB, PSLB, and CB deviated from the group (Fig. 6) [26]. The sensitive analysis indicated no difference in the alignment efficiency of PSLB and CB (MD − 4.69, 95% CI − 22.28 to 12.91). The two studies that were not included in the meta-analysis calculated and compared the irregularity scores to present alignment efficiency but did not find any difference between the two. One of these two studies compared ASLB with CB at intervals of about 10 weeks [19], while the other compared PSLB with CB at monthly intervals [29].

Fig. 6
figure 6

Meta-analysis of initial alignment efficiency of self-ligating brackets (SLB) and conventional brackets (CB). PSLB passive self-ligating brackets, ASLB active self-ligating brackets, 95% CI 95% confidence interval, SD standard deviation

Abb. 6 Metaanalyse zur Effizienz des initialen Alignments mit selbstligierenden Brackets (SLB) und konventionellen Brackets (CB). PSLB passive selbstligierende Brackets, ASLB aktive selbstligierende Brackets, 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image

Efficiency of treatment time

Three included studies analyzed the number of months required for a complete orthodontic treatment [5, 9, 14], whereby the meta-analysis indicated that CBs demand less treatment time than SLBs regardless of type—active or passive (MD 2.20, 95% CI 0.43–3.97). Moreover, when SLBs were further classified into ASLBs and PSLBs for detailed analysis, no significant statistical variation was observed (Fig. 7).

Fig. 7
figure 7

Meta-analysis of treatment time of self-ligating brackets (SLB) and conventional brackets (CB). PSLB passive self-ligating brackets, ASLB active self-ligating brackets, 95% CI 95% confidence interval, SD standard deviation

Abb. 7 Metaanalyse zur Behandlungszeit mit selbstligierenden Brackets (SLB) und konventionellen Brackets (CB). PSLB passive selbstligierende Brackets, ASLB aktive selbstligierende Brackets, 95% CI 95%-Konfidenzintervall, SD Standardabweichung

Full size image

GRADE assessment

According to the GRADE quality analysis, the quality of the evidence for mandibular intermolar width and space closure outcomes was found to be low. In contrast, the quality of the evidence for mandibular intercanine width, initial alignment, and treatment time are moderate, and further research is required for confirmation.

Discussion

There are many uncontrolled factors which possibly influence clinical orthodontic treatment with brackets such as slot design and mode of operation (active or passive). The described factors may also affect the resistance to sliding, thus, affecting the bracket friction system and other facets of clinical applications [20]. However, in contrast to the present meta-analysis, most recent meta-analyses have not classified SLB into ASLB and PSLB. Seventeen RCTs focusing on the efficacy and competence of SLBs were included. Their clinical heterogeneities were also taken into consideration such as differences in treatment procedures, clinical and statistical methods, and the uniformity of response and oral healthcare habits of the patients [15]. Thus, only 11 of the 17 RCTs were further utilized for the meta-analysis.

Data for transversal dimensional arch width change and orthodontic space closure were extracted and synthesized from the results of the included RCTs. The meta-analysis showed CB to be more beneficial in mandibular intercanine width expansion and SLB as more effective in intermolar use without stability. No major difference was seen between SLB and CB in incisor inclination control and canine retraction speed. However, the results should be assessed carefully while considering all kinds of factors. As for the dimensional change, a published meta-analysis that included one RCT and two CCTs found no differences in their results [4]. Although the present meta-analysis includes four RCTs, the included sample size was limited and heterogeneity such as differences in race may not have been effectively avoided [31]. Moreover, one included RCT was a multicenter study and two had unclear risks of bias [25]. Furthermore, all studies included for meta-analysis of orthodontic space closure were split-mouth trials with maxillary first premolar extraction but with different bracket types and closing methods, such as NiTi springs or elastic coils. One published systematic review also considered orthodontic space closure, but included only one CCT and no definite conclusion was made [10].

Data regarding the efficiency of initial alignment and complete treatment time were synthesized from different RCT studies. The conducted meta-analyses favored CBs for shortened treatment time and these results were consistent with another published systematic review that included only two studies having a standard mean difference. However, when comparing the time taken for complete treatment using SLBs, clinical heterogeneities such as variations in the time interval and patient compliance must not be overlooked [3]. A published meta-analysis included three retrospective cohort studies and found no difference between SLB and CB in regard to the total treatment time [4]. Hence, further research is required to obtain substantial data for these aspects. As for the initial alignment, the sensitive analysis was conducted to confirm consistency in initial alignment efficiency between PLSB and CB. A novel approach to network meta-analysis was carried out to compare ASLB, PSLB, and CB for initial orthodontic alignment, including 6 RCTs and 4 CCTs, which provided direct, indirect, and mixed evidence by advanced statistical methods [23]. The efficiency of initial orthodontic alignment was also reported in a systematic review that included 6 RCTs and 11 CCTs, although results were described only superficially without actual quantitatively synthesized evidence [10]. In another meta-analysis, only RCTs were included but the study effectively focused on the oral hygiene of self-ligating brackets, neglecting the determination of effectiveness or efficiency [30].

We only included randomized controlled clinical trials that have been considered to have high-quality clinical evidence in the present systematic review and meta-analysis. Moreover, our review has been conducted strictly according to standards set forth in the Cochrane Handbook and have included analysis of reasonable search strategy, risk-of-bias assessment, and multianalyses. However, some potential limitations could not be avoided. For example, selection bias could not be completely avoided, which is a common problem for all systematic reviews. Electronic databases and ongoing studies were searched thoroughly for relevant information. Manual searching was also performed, which may have restricted the endemic bias to a certain degree. However, the manual searching was limited to Chinese language publications due to the unavailable access to other language journals. Clinical heterogeneity was discovered to be pervasive and not entirely avoidable due to differences in brands of SLBs and CBs, treatment strategies, appointment intervals, and archwire sequences. Moreover, the participant population included in each study was not large, resulting in deficient statistical power and lack of prominent significant results. In addition, the number of published studies included in this meta-analysis was limited. There is a strong need for more quality analyses in the future in order to obtain clear, definite, and reliable outcomes.

Our findings in this systematic review and meta-analysis of RCTs do not suggest the superiority of SLBs for the orthodontic clinical application. In fact, our results appear to support CBs as being more advantageous in regard to expanding the mandibular intercanine width and overall treatment, which requires immediate and detailed exploration. Although initial results seem to favor PSLB to expand the mandibular intermolar width, the sensitive analysis did not confirm this. In addition, the results show no difference at low or moderated GRADE quality. Taken together, the obtained results indicate that more high-quality orthodontic studies are necessary to obtain significant comparative data on the effectiveness and efficiency of SLB and CB. When choosing bracket design for clinical use, it is recommended that orthodontists consider factors like patient preference, main treatment purpose, and cost-effectiveness over the popularity of a technique.

Conclusions

On the basis of this systematic review and meta-analysis, which collected and synthesized current evidence from RCTs regarding the clinical use of orthodontic brackets, CBs appear to have an advantage over SLBs in expanding the mandibular intercanine arch width, while SLBs appear to be superior for utilization on intermolars without stability. Furthermore, no difference in the clinical effectiveness of orthodontic space closure between bracket types was found. Moreover, current evidence does not favor SLBs or CBs in terms of alignment efficiency, although investigated evidence supports the utilization of CBs in terms of overall treatment efficiency. However, because evidence is not compelling, more high-quality RCTs with low risk of bias are required for the confirmation of these results. For the best results, orthodontist’s choice of bracket types should be based on an understanding of clinical evidence, which has been largely covered in this study.