A comparative assessment of stability and satisfaction between two mandibular lingual retainers in orthodontic patients. A 2-year follow-up, single practice-based randomized trial

Abstract

Objective

This study aimed to evaluate and compare the stability of mandibular dental arch dimensions and patient satisfaction between two types of fixed retainers—3-strand round twisted (RT) and 8-strand rectangular braided (RB)—both of which are bonded to all six anterior mandibular teeth.

Trial design

2-arm parallel, two-center prospective randomized controlled trial.

Methods: Participants

133 orthodontic patients (median age 24.6 years, 25th percentile = 17.2 years, 75th percentile = 32.4 years) were recruited.

Interventions

These patients were randomly assigned to receive either an RT or RB wire retainer at a 1:1 ratio.

Randomization

It was achieved using random permuted blocks of sizes 4, 6, or 8, which were concealed in sequentially numbered, opaque, sealed envelopes.

Outcomes

The primary outcome was the change in the irregularity index, with secondary outcomes including arch length; intercanine, interpremolar, and intermolar widths; and patient satisfaction. Evaluations were performed at baseline and at 3, 6, 12, 18, and 24 months after retainer placement. Dental cast measurements were analyzed using random effects linear regression, and satisfaction was assessed at each time point.

Blinding

Blinding of patients was not feasible. Only blinding the assessor for patient satisfaction was achieved.

Results

Cast measurements remained relatively stable from T1 to T6, with no significant difference attributed to the retainer type (RT or RB). Time significantly affected all cast measurements except for the irregularity index. There was no significant correlation between retainer type or time and satisfaction questionnaire responses, although the responses varied by question. No harms were observed.

Conclusions

Both RT and RB wire retainers effectively maintain mandibular arch alignment and are equally well tolerated by patients in the medium term.

Introduction

Fixed retainers (FRs) are commonly used worldwide in orthodontic patients following the active treatment phase [1–3]. In some countries, almost all orthodontists recommend them to their patients regardless of the type of malocclusion or other factors [1, 2], while in other regions, it is one of the most frequently used retention modalities [4]. A recently developed clinical practice guideline (CPG) [5] recommends, for example, the routine use of FRs in the mandibular arch. However, FRs can fail and the failure rate can exceed 50% [6, 7]. The most apparent consequence of FR failure can be dental relapse, but other implications also exist. A failed retainer needs to be repaired. Some detachments can be detected during routine checkups because they do not cause any irritation [6]. However, failures noticed by patients, particularly those requiring extra dental visits for repairs, are more likely to result in dissatisfaction. Additionally, dissatisfaction can arise from the physical properties of the retainer itself. For example, square or rectangular wires, commonly used in FRs, have sharper edges than round wires, potentially causing irritation to the tongue. Thus, patient dissatisfaction may be linked either to the physical characteristics of the retainer (e.g. tongue irritation) or to issues related to its failure (e.g. relapse or repair inconvenience).

While several randomized clinical trials (RCTs) have evaluated different aspects of FRs, only a few have explored patient satisfaction [8–10]. For instance, Forde et al. [8] found no significant difference in the embarrassment experienced by patients using FRs versus vacuum-formed retainers (VFRs). However, VFR users reported greater discomfort and speech interference compared to FR users, while also finding VFRs easier to clean. Pullisaar et al. [9] compared patient satisfaction between 6-stranded round FRs and CAD/CAM rectangular nitinol FRs, finding no significant differences. Interestingly, patient satisfaction improved significantly over time for most survey questions. Similarly, Jasim et al. [10] observed no significant differences in general satisfaction between patients wearing polyether-ether-ketone (PEEK) FRs and those wearing dead-soft coaxial FRs over six months.

Prolonged use of FRs without appropriate monitoring is associated with potential side effects, such as active displacement of the root from the alveolus, a phenomenon referred to as an ‘active retainer’ [11]. One proposed strategy to mitigate this risk—possibly caused by the untwisting of multistrand round wires—is the use of rectangular wires, which are less prone to untwisting and subsequent activation. Although this approach is recommended in the aforementioned CPG [5], it is important to note that the recommendation is based on clinical experience rather than robust research evidence, as no studies to date have evaluated the long-term effects of rectangular FRs.

Therefore, the primary aim of this study was to compare two FRs—0.0215” 3-strand round twisted with a 0.0265” x 0.0106” 8-strand rectangular braided—in terms of patient satisfaction and dental arch stability over a 24-month period. Our research hypothesis (H_R) was that there would be no difference between the two FRs. The secondary aim was to evaluate the effect of failures (published in [6]) on satisfaction.

Material and methods

The study protocol (# KB/956/14, October 23, 2014) was approved by the Ethical Committee of Warsaw Regional Medical Chamber in Warsaw, Poland. Each participant provided written informed consent (with legal guardian consent if the participant was underage). However, the trial was not registered. No changes were made to the methods after the trial commenced.

Design, participants, eligibility criteria, and setting

This was a single-center 2-arm parallel-group randomized controlled trial employing a 1:1 allocation ratio. Participants were enlisted from December 2014 to April 2018, sourced exclusively from a solitary orthodontic private practice belonging to the senior author (PF). Inclusion criteria comprised individuals aged 15–50 at the time of debonding, possessing all mandibular permanent incisors and canines, with no active caries, restorations, or fractures on the mandibular incisors and canines, without periodontal disease, and treated with fixed labial appliances in both arches without extractions of permanent teeth (except of wisdom teeth). Their retention plan had to exclusively involve a retainer bonded from 3 to 3. Exclusion criteria encompassed inadequate hygiene, the necessity for restorative or surgical intervention, active periodontal disease, or the use of a removable retainer in addition to a bonded one. Each eligible patient received information about the trial, its objectives, and methodology, 2 months prior to planned debonding. Subsequently, patients were asked for their willingness to participate, with additional time for consideration provided upon request. Those who consented underwent the procedures outlined in the ‘Interventions’ section.

Sample size

Description of sample size calculation is presented in the 1^st part of this study [6]. In short, a minimum sample size of 58 subjects per group was determined based on the ability to detect a 25% difference (50% versus 25%) in the risk of first-time failure between the 2 trial arms.

Randomization

An online random number generator from www.sealedenvelope.com was used to generate random permuted blocks of 4, 6, or 8 patients to secure equal allocation to the two arms. Allocation concealment was done using sequentially numbered, opaque, sealed envelopes prepared before the trial commenced. The study coordinator was tasked with opening the subsequent envelope in sequence and executing the randomization process.

Interventions

One month prior to debonding, an alginate impression of the lower dental arch was taken and promptly dispatched to the laboratory within 24 hours, accompanied by specifications for the requested retainer type. The options included either a 0.0215” stainless steel 3-strand round twisted wire retainer (Ortho Organizers, Lindenberg, Germany) or a 0.0265” x 0.0106” 8-strand Bond-a-Braid wire retainer (Reliance Orthodontic Products, Itasca, IL, U.S.A). Approximately two weeks before debonding, each study participant underwent a scheduled session for scaling and tooth cleaning administered by an experienced hygienist.

The bonding process entailed the following steps: insertion of a cheek retractor, cleansing of the lingual surfaces of the six anterior teeth, application of 37% phosphoric acid etchant, thorough rinsing and drying, placement of dental floss segments to secure the retainer during bonding, coating of the etched tooth surfaces with primer (TransbondTM XT adhesive primer; 3M Unitek, Monrovia, California, USA), application of light-cured composite (TransbondTM Supreme LV; 3M Unitek), and curing the composite with light. Every precaution was taken to prevent moisture from interfering with the bonding process on the lingual surfaces of the teeth. Additionally, all retainers were bonded by the same experienced orthodontic assistant, boasting over 15 years of expertise in this procedure.

Outcomes

There were 2 types of outcomes assessed in this part of the study:

• 1) Irregularity index [12], intercanine width, interpremolar width (between buccal and between lingual cusps), intermolar width (between mesial buccal and between mesial lingual cusps), and arch length (on right and left side)—all were measured on plaster casts

• 2) Satisfaction evaluated with a questionnaire.

Data collection

Data were collected at 6 time points: T1 (baseline, at debonding and retainer placement) and T2, T3, T4, T5, and T6 (3, 6, 12, 18, and 24 months respectively after retainer placement) by the same orthodontist (EW). In case of retainer failure, study participants were requested to contact the practice as soon as possible. At the end of the visit the subsequent appointment was scheduled in advance to ensure regular follow-up.

At each visit for data collection, in addition to the steps described our previous article [6], an alginate impression of the lower dental arch was made, and the patient was asked to complete a questionnaire containing 11 questions related to comfort with the retainer. The questionnaire (Supplementary material) used a 4-point scale: 1—no/cannot say, 2—barely, 3 -moderately, and 4—certainly.

The alginate impression was sent to the laboratory within 24 hours. It was stored in a refrigerator before being sent. The previously mentioned measurements—irregularity index, intercanine width, interpremolar width, intermolar width, and arch depth—were done by DT using a digital caliper with an accuracy of 0.5 mm. The results were recorded on a spreadsheet in real time.

Blinding

Blinding of patients was not feasible due to the nature of the intervention. Blinding the assessor for cast measurements was not implemented—the retainer was visible during the measurements. However, blinding the assessor for patient satisfaction was achieved by removing patient identifying information from the completed forms.

Method error assessment

Plaster casts made at T1 to T6 of 20 patients (i.e. 120 dental casts in total) were remeasured after more than 1 month by DT to establish method error.

Statistical analysis

For the cast measurements analysis, descriptive statistics were calculated per time point. The direct likelihood approach was used to account for the missing data by taking advantage of the multiple time points and thus for each dental cast outcome, random effects linear regression analyses for the effect of treatment (retainer type) adjusted for time and using the patient as the random effect were carried out.

For the satisfaction assessment percentages of responses per time, question and retainer type were tabulated and plotted. A mixed ordinal logistic regression model with patient as the random intercept was fit using the 4-level categorical response as the outcome and retainer type, time and question as predictors in order to examine their main effects on the outcome. Again, the direct likelihood approach accounted for the missing data by taking advantage of the multiple time points. On an exploratory basis, and to assess the population average effect of failure on the satisfaction score (i.e. the sum of the 1–11 question scores per patient and timepoint) adjusted for time a generalized estimating equations (GEE) [13] model was fit using non-parametric bootstrap (500 iterations) to calculate the 95% confidence intervals.

Intraclass correlation coefficients (ICCs) along with 95% confidence interval were calculated for duplicate dental casts measurements for error assessment.

All analyses were conducted in Stata 18 (StataCorp, College Station, TX, USA), and R Software version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Participants

One hundred thirty-three participants (42 males, 91 females) were enrolled in the study, with a median age of 24.6 years (25th percentile = 17.2 years, 75th percentile = 32.4 years, minimum = 15.1 years and maximum = 49.8 years). They were randomly assigned to one of two groups: the 3-strand round twisted wire retainer (RT group, 65 participants, 22 males, 43 females) or the 8-strand rectangular braided wire retainer (RB group, 66 participants, 19 males, 47 females), both of which were bonded to all mandibular incisors and canines. The CONSORT flowchart (Fig. 1) illustrates the flow of participants through the study. Out of the 133 participants, 132 received their allocated treatment. One participant discontinued intervention as they requested retreatment for a displaced premolar, thus being excluded from the analyses. Additionally, four participants who missed 1 to 3 data collection appointments were still included in the analysis.

Missing data

Missing cast measurements ranged from 9.6%–15.03%, while missing questionnaire responses ranged from 0.89–13.53% and increased over time. On average, 8.74% of the responses were missing and were considered to be missing at random (MAR).

Measurement error

The mean ICC for dental cast measurements was 0.967. The ICC ranged from 0.824 (95% confidence interval: 0.640 to 0.925) to 0.999 (95% confidence interval: 0.996 to 0.999). Detailed ICC values are provided in the Supplementary material.

Stability of the dental arch

Table 1 demonstrates the equivalence of the RT and RB groups at baseline (T1) regarding dental arch measurements. Overall, all measurements remained relatively stable from T1 to T6 (Fig. 2) with changes being small and clinically insignificant. Random effects linear regression analyses assessing the effect of retainer type on plaster cast measurements, adjusted for time, revealed no significant impact of retainer type—RT or RB—on the cast measurements. However, time had a significant effect on the change of all cast measurements except for the Irregularity Index, for which no significant effect of time was observed (Table 2). The interaction effect between the type of retainer and the progression of time in our model (retainer*time interaction) was non-significant for all the measurements indicating similar evolution between retainer groups over time. The Wald test yielded a P-value of < .001 for the 3–3, 4–4 buccal and lingual, 6–6 buccal and lingual, and 1–6 right and left distances.

Satisfaction

The supplementary table 1 (see Supplementary material) illustrates the percentages of responses per question, retainer type, and timepoint. Overall, participants expressed satisfaction with both types of retainers throughout the 2-year observation period (Fig. 3). Initially, at baseline (T1), immediately following retainer placement, a relatively high proportion of participants from both groups indicated feeling ‘moderately’ or ‘certainly’ to question 1 (‘Do you feel that your tongue has less space?’), suggesting potential tongue irritation due to the retainer. However, after 3 months (T2), fewer responses suggested reduced impression of limited tongue space. Conversely, there was an increased proportion of ‘moderately’ or ‘certainly’ responses to question 11 (‘Do you have problems with flossing?’) from time point 2 to 6, indicating that participants quickly realized the difficulty of dental flossing in the retainer area.

Analysis revealed no evidence of an association between response and retainer type or time, but a significant association was observed between the response and question (Table 3). This means that the question was a significant predictor for the response or that the responses are not homogeneous across the questions. Overall, after adjusting for time and question, there was no discernible difference in responses between the two types of retainers (Table 4).

Effect of retainer failure on satisfaction

A fitted GEE model to get the average effect of failure on the satisfaction score (i.e. the sum of the 1–11 question scores per patient and timepoint) adjusted for time using non-parametric bootstrap to calculate the 95% confidence intervals showed no difference in satisfaction between situation of retainer failure vs. non-failure. Only time was a significant predictor (table 5).

Harms

No harms were observed.

Discussion

In this study we assessed stability of the dental arch and patients’ satisfaction after 2 types of retainers were bonded to all 6 anterior teeth (canine to canine) in the mandible. Our research hypothesis was that there was no difference between groups for any of the outcomes.

Patients in both groups demonstrated considerable stability in dental arch alignment, with the irregularity index and intercanine width remaining virtually unchanged over the two years of observation. Additionally, despite the wire retainers being bonded only from canine to canine, the interpremolar and intermolar widths and arch lengths showed remarkable stability. Our findings agree with other studies, which commonly report good 2-4-year posttreatment stability of dental arch dimensions in patients with bonded canine-to-canine retainers [3, 14–16]. Surprisingly, the lower dental arch retained with vacuum-formed retainers demonstrated similar stability to that retained with canine-to-canine retainers [3, 17, 18]. A Cochrane review [3] indicated that regardless of the type of bonded retainer or wear regimen for vacuum-formed thermoplastic retainers, the intermolar width and arch length were almost always comparable between the bonded canine-to-canine and vacuum-formed retainer groups. Given the limitations of the studies included in the Cochrane review, it suggested that premolar and molar widths in the mandible remain stable, whether retained or unretained.

In a previous part of this investigation, we found that approximately 50% of patients—regardless of retainer type—experienced retainer failure during the two years of observation [6]. Failures were usually described as detachment of the retainer wire from a single tooth due to inadequate adhesion of the composite to the tooth surface or composite damage and subsequent ‘release’ of the wire. Complete retainer detachments were very rare. Regardless of the origin, retainer failure can lead to relapse of anterior alignment.

The current findings demonstrated that retainer failure had no effect on the irregularity index. In other words, detachment of the retainer from the tooth did not result in significant displacement of the released tooth. It is possible that the short time from the occurrence of failure to its detection—patients were scheduled for regular checkups—prevented the development of a more serious problem. It is, however, equally likely that a single tooth detachment, with adjacent teeth being attached to the wire, allows for only limited displacement of the tooth away from the retainer wire. If the original tooth position to which it could relapse was ‘toward the wire’, displacement might not have occurred. Our results align with overall findings of other studies that canine-to-canine retainers in the mandible are effective in preserving good dental arch alignment [3, 15, 19, 20]. Even relatively frequent single-tooth failures as observed in our study, did not result in relapse.

Orthodontic treatment often lasts several years, making the end of therapy and removal of appliances a highly anticipated event. Many patients view the day of debonding as a form of ‘liberation’—a moment when the discomfort associated with various aspects of orthodontic treatment finally comes to an end. Consequently, the beginning of retention phase, especially when prolonged retention is advised, can be associated with stress, discomfort, and even dissatisfaction because the expected end has not come. Ideally, retention appliances that interfere less with normal function and have fewer failures seem more likely to be accepted by patients. Overall, patient satisfaction with both types of retainers—round twisted and rectangular braided—was high throughout the observation period. This indicates that the rectangular retainer wire with relatively sharp edges did not cause any discernible discomfort compared to the round wire. This is consistent with the study by Pullisaar [9] which demonstrated that patient satisfaction does not depend on the specific type of bonded retainer. On the other hand, the rigid retainer bonded only to canines, which was not tested in this study, was preferred by patients compared to the rectangular chain [3].

As expected, the retainer bonded from canine to canine in the mandible was quickly sensed by patients as limiting the space for the tongue. The initial discomfort related to decreased space for the tongue was most noticeable at the beginning but happily diminished over time. The difficulty in flossing in the retainer’s area, in turn, was perceived as a source of some discomfort during the observation period. Our results suggesting that this aspect of retention was the most difficult for patients to accept, align with findings of Pullisaar et al. [9].

As mentioned before, both types of retainers—round twisted and rectangular braided—had comparable and relatively high failure rates. Patients who noticed failure required extra appointments to repair a failed retainer. Additional time and effort devoted to the visit to the orthodontist’s office can be associated with increased dissatisfaction. In our study, we found no effect of retainer failure on overall patient satisfaction. It is possible that failures were mostly undetected by patients—they were identified during thorough checkups at scheduled control appointments. In other words, failures seemed ‘hidden’ and did not interfere with daily activities or oral function. Our findings are consistent with those of Jowett et al. [21] who reported high levels of satisfaction with two types of retainers (Memotain® vs. Flex-tech) despite over 50% of upper Memotain retainer failures at six months, with only 26% of patients with Flex-tech retainer reporting similar issues.

Limitations

Several limitations of this study should be noted. First, the study was conducted in a single practice, which may limit the generalizability of the results. Second, we used a questionnaire that was not formally validated, which could have affected validity of our findings relating to satisfaction. Third, differences in the subjective assessment of patient satisfaction may be influenced by various factors, such as personal preferences and tolerance for discomfort. Fourth, the study protocol was not registered prior commencement of the study. Finally, the relatively wide age range of participants may have influenced the outcomes.

Conclusions

Our research hypothesis was accepted—both the 3-strand round-twisted and the 8-strand rectangular-braided wire retainers are effective in maintaining mandibular arch alignment and are well-tolerated by patients in the medium-term perspective. Therefore, the choice of retainer type can be based on clinical preferences and material availability, without concern for differences in clinical effectiveness and patient satisfaction.

Conflict of interest

Nothing to declare.

Funding

This work received no external funding

Data availability

Data underlying this publication will be shared on reasonable request to the corresponding author.

References