European Journal of Orthodontics, 2017, 1–12

doi:10.1093/ejo/cjx058

Randomized controlled trial

Bonded versus vacuum-formed retainers: a

randomized controlled trial. Part 1: stability,
retainer survival, and patient satisfaction
outcomes after 12 months
Katherine Forde1, Madeleine Storey2, Simon J. Littlewood3,
Paul Scott4, Friedy Luther5 and Jing Kang6
1
Holywell House Orthodontics, Hinckley, Leicestershire, UK, 2Department of Orthodontics, Leeds Dental Institute,
Clarendon Way, UK, 3Department of Orthodontics, St Luke’s Hospital, Bradford, UK, 4Department of Orthodontics,
Montagu Hospital, Mexborough, Doncaster and Bassetlaw NHS Trust, UK, 5Sheffield Teaching Hospitals NHS
Foundation Trust, Department of Orthodontics, The Charles Clifford Dental Hospital, Sheffield, UK and 6Department of
Biostatistics, School of Dentistry, University of Leeds, UK

Correspondence to: Simon J. Littlewood, Department of Orthodontics, St Luke’s Hospital, Bradford BD5 0NA, UK. E-mail:
simonjlittlewood@aol.com

Summary
Background:  There is a shortage of evidence on the best type of retainer.
Objectives:  Evaluate upper and lower bonded retainers (BRs) versus upper and lower vacuum-formed
retainers (VFRs) over 12 months, in terms of stability, retainer survival, and patient satisfaction.
Trial design:  Two-arm parallel group multi-centre randomized controlled clinical trial.
Methods:  Sixty consecutive patients completing fixed appliance therapy and requiring retainers
were recruited from 3 hospital departments. They were randomly allocated to either upper and
lower labial segment BRs (n = 30) or upper and lower full-arch VFRs (n = 30). Primary outcome was
stability. Secondary outcomes were retainer survival and patient satisfaction. A random sequence
of treatment allocation was computer-generated and implemented by sealing in sequentially
numbered opaque sealed envelopes independently prepared in advance. Patients, operators and
outcome could not be blinded due to the nature of the intervention.
Results:  Thirty patients received BRs (median [Mdn] age 16 years, inter-quartile range [IQR] = 2)
and 30 received VFRs (Mdn age 17 years, IQR = 4). Baseline characteristics were similar between
groups. At 12  months, there were no statistically significant inter-group differences in post-
treatment change of maxillary labial segment alignment (BR = 1.1 mm, IQR = 1.56, VFR = 0.76 mm,
IQR = 1.55, P = 0.61); however, there was greater post-treatment change in the mandibular VFR group
(BR = 0.77 mm, IQR = 1.46, VFR = 1.69mm, IQR = 2.00, P = 0.008). The difference in maxillary retainer
survival rates were statistically non-significant, P = 0.34 (BR = 63.6%, 239.3 days, 95% confidence
interval [CI] = 191.1–287.5, VFR = 73.3%, 311.1 days, 95% CI = 278.3–344.29). The mandibular BR had
a lower survival rate (P = 0.01) at 12 months (BR = 50%, 239.3 days 95% CI = 191.1–287.5, VFR = 80%,
324.9 days 95% CI = 295.4–354.4). More subjects with VFRs reported discomfort (P = 0.002) and
speech difficulties (P = 0.004) but found them easier to clean than those with BRs (P = 0.001).
Limitations:  Results are after 1 year and we do not know how much the removable retainers were worn.
Conclusions:  After 1 year, there is no evidence of a significant difference in stability or retainer survival
in the maxilla. In the mandible, BRs are more effective at maintaining mandibular labial segment
alignment, but have a higher failure rate. In comparison with patients wearing VFRs, patients wearing

© The Author 2017. Published by Oxford University Press on behalf of the European Orthodontic Society. All rights reserved.
1
For permissions, please email: journals.permissions@oup.com

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

by U. of Florida Health Science Center Library user
on 16 January 2018
 2 European Journal of Orthodontics, 2017

BRs reported that they caused less interference with speech, required less compliance to wear them,
and were more comfortable to wear than VFRs. Patients found the BRs harder to keep clean.
Trial registration:  The trail was not registered.

Introduction views. As a result, focus groups, in-depth interviews, and ques-

tionnaires are increasingly used in research and audit to assess
Retention is arguably the most challenging and unpredictable stage
patient satisfaction (23).
of orthodontic treatment and the most clinically effective form of
Several studies have investigated the attitudes and preferences
retention remains unclear (1).
of orthodontists towards various retention protocols but few have
In the UK, there has been a significant increase in the use of vac-
reported on the perceptions of patients (7, 24). The few studies which
uum-formed retainers (VFRs) and bonded retainers (BRs). Ideally, the
have explored patient satisfaction with orthodontic treatment have
choice of retainer appropriate for each individual patient, whether
highlighted a nearly universal dislike of orthodontic retainers. Some
fixed or removable, should be made considering not only the clinical
find them to be more inconvenient than their fixed appliances and
effectiveness, but also the patient’s ability to, and willingness to com-
even headgear (25–27). The reasons cited include difficulty speaking
ply with, the retention plan (2, 3).VFRs have been shown to be more
and eating, extra salivation, smell, embarrassment, and the ease with
clinically effective at maintaining the lower labial segment alignment
which they could be lost (25, 26). A remarkably high degree of non-
than Hawley retainers in the initial 6 months of retention (4). The
compliance was also noted, with most patients reporting little or no
retentive capacity of the VFR appears to be equally effective when
attention to dietary restrictions or flossing regimens (27).
worn full-time or at night only, during the initial 6–12 months (5, 6).
Hichens et al. compared levels of patient satisfaction of subjects with
Furthermore, VFRs have been shown to be preferred by both cli-
Hawley and VFRs during the first 6  months of the retention period.
nicians and patients owing to their comfort, ease of fabrication and
The results indicated that the majority of subjects showed a preference
fit, reduced failure rate, and cost-effectiveness (7).
for VFRs compared with Hawley retainers. Wearing VFRs caused less
BRs are commonly used as a means of permanent retention usu-
embarrassment to subjects, especially in terms of speech and appear-
ally in the form of a wire attached to the lingual or palatal surfaces
ance, and broke less frequently than Hawley retainers. Interestingly,
of the teeth within the labial segment (3, 8). They are particularly
although more subjects in the VFR group wore their retainers as
useful in cases of perceived instability, or high patient expectation.
instructed compared with the Hawley group, there was no difference in
There are many different types of BR. The most common is fabri-
complaints of discomfort between subjects in either group (28).
cated from multi-strand twistflex wire, which is bonded to each of
Jaderberg et al. evaluated patient experiences of subjects wearing
the labial segment teeth and is associated with a superior retentive
VFRs either full-time for 3 months followed by nights-only wear com-
capacity compared to the single-stranded wire bonded to the canines
pared with 1 week full-time wear followed by nights-only wear, for the
only (9–11). One of the clear advantages of BRs is that they are less
first 6 months of the retention period. Overall, the results of the ques-
reliant on patient compliance. However, notable disadvantages exist,
tionnaire showed that the VFRs were well tolerated, easy to get used to,
leading to reports of reduced popularity amongst some orthodon-
and most patients reported that they had no difficulties in remembering
tists (12). Placement is time-consuming and technique sensitive (13),
to wear the retainers at night. Soreness was reported in 13%, 22% had
failure has been reported to be as high as 53% (10)), and bonding
problems with speech, and 10% thought that they tasted bad (29).
failures may progress unnoticed by the patient, leading to localized
A recently updated Cochrane review looking at orthodon-
relapse, and the risk of caries or demineralisation (14).
tic retention emphasized the need for future research in this field
Wire detachment at the enamel–composite interface appears to be
to incorporate outcomes of stability, periodontal health, survival
the most common reason for failure (14–16), particularly within the
of retainers, and quality of life (1). Further well-designed prospec-
early stages of retention (17), with the canine region being the most com-
tive randomized controlled trials would therefore be desirable to
mon site for detachment failure (13, 18). Fracture of the wire is report-
strengthen the evidence comparing the most commonly used retainer
edly the least common failure type, and predominantly occurs mesial to
types—VFRs and BRs.
the maxillary canines, probably due to occlusal trauma (13, 17).
This is the first of two papers comparing upper and lower VFRs
Post-treatment changes may even occur with BRs in situ (19).
with upper and lower BRs. This paper aims to test the null hypoth-
Spacing and rotation of teeth still firmly attached to BRs has also
esis that there are no significant differences in clinical effectiveness
been reported when BRs have remained in place for several years
between the two retainer types. The primary outcome is stability and
and despite not breaking (20, 21), as has fraying of the multi-strand
the secondary outcomes are retainer survival, patient satisfaction,
wire which can cause irritation to the tissues (21).
and affect on periodontal health. The second paper will report on
At the time of starting this trial, the authors were aware of only
periodontal outcomes.
one high-quality published trial directly compared the VFR to the
BR. This prospective randomized controlled trial was conducted
over a 12-month period and the outcomes were published in the
Methods
form of conference abstracts (12, 22). More than three times as many
problems, such as loss or fracture of the retainers, were noted in the Trial design and ethics
VFR group. Labial segment alignment was statistically worse in the This was a 2-arm parallel group randomized controlled clinical
VFR group at 12  months and the authors related this to patient’s trial with a 1:1 allocation ratio. The study was approved by the
suboptimal compliance with removable retainers. National Research Ethics Service in December 2010 (REC number
There has been a growing emphasis on the importance of 10/H1306/79), and also the local ethical and research development
using qualitative research methods to gather data of patients’ departments of the three hospital trusts involved.

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

by U. of Florida Health Science Center Library user
on 16 January 2018
 K. Forde et al.
Participants, eligibility criteria, and settings
Consecutive patients nearing the completion of fixed appliance ther-
apy who required retainers were invited to take part in the trial.
Participants were recruited from the orthodontic departments of
two district general hospitals (St Luke’s Hospital, Bradford and York
Hospital) and one teaching hospital (Leeds Dental Institute) in the
UK. The retainers were placed and reviewed by two clinicians (KF
and MS), both of whom were Speciality Registrars in Orthodontics
in the early stages of their training. Treatment was provided for free
to patients under the UK’s National Health Service.
The following inclusion criteria were applied:
• Completion of a course of fixed appliance therapy (involving
both dental arches) with a satisfactory correction of the present-
ing malocclusion and dental alignment
• Good general health
• Demonstration of a good standard of oral hygiene (determined
through questioning and clinical examination)
• Subjects had a full and normal complement of teeth in the upper
and lower labial segments
• Teeth within the labial segments to be a regular size and shape
• Subjects willing to consent to the trial and comply with the trial
regime
Subjects may have presented with any malocclusion prior to ortho-
dontic treatment and may have been managed on an extraction (pre-
molar or molar) or a non-extraction basis. Subjects may have been
treated with removable or functional appliances in conjunction with
their orthodontic treatment, or may have had adjunctive surgery.
The following exclusion criteria were applied:
• Any general medical health problems which may influence gin-
gival health, such as those necessitating antibiotic cover as bac-
teraemia prophylaxis, diabetes mellitus, epilepsy, or physical or
mental disability
• Poor periodontal health, including the presence of supra-gingival
calculus or periodontal pocketing greater than 3 mm
• Periodontal problems with either a pending referral to the peri-
odontal department or a history of periodontal management dur-
ing the orthodontic treatment
• Gross or uncontrolled caries
• Absent or diminutive lateral incisors
• A starting malocclusion requiring extreme transverse correction
(involving rapid maxillary expansion or surgically assisted rapid
maxillary expansion)
• Cleft palate or severe facial deformities
Interventions
At the end of active treatment, two calibrated operators placed either
upper and lower BRs or upper and lower VFRs using standardized
Figure 1.  Bonded retainers with silicone positioning jig.
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
3
procedures and materials. All patients had been seen for supra-gingi-
val debridement prior to removal of the fixed appliances.
Upper and lower BRs (Figure 1) were prepared using 0.0195 in
(0.45  mm) 3-stranded twistflex stainless steel wire (Wildcat; GAC
International, Bohemia, New York, USA). The wire was shaped by
dental technicians against the dental casts to lie passively against the
lingual surfaces of the upper and lower incisors and canines. The
wire was not contoured interproximally. A  silicone positioning jig
was prepared to aid positioning of the retainer. This was moulded to
engage the wire and rest on the central incisors. The wire was bonded
using a low viscosity light-cured composite (Transbond™ LV; 3M
Unitek, Monrovia, California, USA), following the separate applica-
tion of etchant (37% phosphoric acid), and primer (Transbond™
XT adhesive primer; 3M Unitek). Care was taken not to leave any
bonding substance in contact with the gingival tissues.
Upper and lower VFRs (Essix™ C+; Figure 2) were constructed
using the ‘Essix™’ machine and cooled rapidly using Arctic spray
(Ortho-Care). Retainers were trimmed to cover all fully erupted
teeth and extend half way across the occlusal surface of the most
distal molar. Participants were instructed to wear their retainers only
at night, every night. At the time of retainer insertion, both written
and verbal oral hygiene instructions were given, including methods
for interdental cleaning around BRs. Participants were advised to
continue 6 monthly review appointments with their general dentist.
Patients were advised to contact the department as soon as pos-
sible if they had any queries, concerns, or problems. Failed retainers
were repaired on the same day adhering to the standardized protocol
for materials, construction, and technique of placement.
Outcome measures
The primary outcome was stability.
Secondary outcomes were:
• Retainer survival
• Patient satisfaction
• Periodontal health (reported in Paper 2)
Stability
The primary outcome was to evaluate the clinical effectiveness of
both retainer types in preventing post-treatment changes over the
initial 12-months of retention. Post-treatment changes were defined
as change in labial segment alignment. This was measured from digi-
tized study models (Figure 3) using Little’s Irregularity Index (LII).
In addition, changes in arch dimensions, occlusal relationships, and
re-opening of extraction spaces were also recorded. Table 1 provides
details of these measurements.
Alginate impressions were obtained at 4 time-points during the
trial (T0–T3) and were measured by the same examiner (KF) to an
accuracy of 0.01 mm.
 4 European Journal of Orthodontics, 2017

Figure 2.  Vacuum-formed retainers.

Table 1.  Single-arch and occlusal measurements recorded at each milestone.

Measurement Unit Description

Little’s irregularity index of the upper and lower
labial segments (LII)
Upper and lower inter-canine width (ICW)
Upper and lower inter-molar width (IMW)
Upper and lower arch length (AL)
Overjet (OJ)
Overbite (OB)
mm The sum of the five labial segment anatomical contact point displacements in a
labiolingual direction
mm Distance between the cusp tips of right and left canines
mm Distance between the mesiobuccal cusp tip of the right first molar to the me-
siobuccal cusp tip of the left first permanent molar.
Where the first permanent molar had been extracted, the second permanent molar
was used instead.
mm The sum of the right and left distances from the mesiobuccal cusp tip of the first
permanent molars to the interproximal contact point of the central incisors
mm The maximum distance between the upper incisor edge and the lower incisal
labial surface, horizontal to the occlusal plane
mm The maximum vertical overlap between the upper and lower incisors with the
models in maximal intercuspation.

• BRs could fail at: the wire–composite interface; the enamel–com-

posite interface; by wire fracture; or by complete detachment
from all teeth.
• VFRs were categorized as being lost, ill-fitting, or fractured.
The occlusal wear was also noted, particularly if there was
a breach in the full thickness of the retainer necessitating
replacement.

Patient satisfaction
Patient satisfaction (secondary outcome measure) was determined
by a questionnaire based on the most relevant questions in the
patient satisfaction questionnaire from the Hichens et  al. RCT
assessing patient satisfaction with Hawley and VFRs (28). The main
issues identified from their study were used to form the basis of
this questionnaire (i.e. it was a modified version of the previously
used questionnaire). A  preliminary version of this questionnaire
Figure 3.  Demonstration of the mandibular arch dimensions (ICW, IMW, AL). was piloted on a group of 10 patients attending the orthodontic
department for retainer reviews. These patients had agreed to be
involved in the pilot study and were not otherwise participating in
• T0Debond
the study. Any necessary modifications to ensure good readability
• T13 months post-debond
and reproducibility were made before using the questionnaires in
• T26 months post-debond
the main study.
• T312 months post-debond
The decision was made to administer the questionnaire to the
participants at their scheduled review appointments, rather than
Survival sending the questionnaire by post or e-mail to complete at home.
Retainer survival was recorded as the time to the first episode of fail- It was hoped that this would ensure that the questionnaires were
ure. The date of failure was recorded as the day the patient became completed fully, increasing the response rate, and to allow the inves-
aware of the problem, or alternatively, the date the clinician noted tigator to clarify any ambiguous questions.
the failure when participants were unaware of a failed retainer. The The data from the questionnaires was coded and assessed by one
pattern of failure was also recorded: operator (MS).

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

by U. of Florida Health Science Center Library user
on 16 January 2018
 K. Forde et al.
Sample size calculation
The sample size was determined based on the primary objective
of comparing the efficacy of each retainer group in minimising the
1-year post-orthodontic treatment change in the anterior arch align-
ment as measured by LII (30).
Assuming a clinically relevant difference of 0.5 mm between the
two randomized groups, a common standard deviation of 0.5 mm, a
power of 90%, and a significance level of 5%, the study required 22
subjects in each group.
To account for a potential dropout rate of 20% and to increase
the sample size for the secondary outcome measures, the sample size
was increased to 30 per group resulting in a total of 60 subjects.
Randomisation
Consecutive participants eligible for inclusion were approached by
KF or MS and written informed consent was obtained from the par-
ticipant (and parent if the participants were adolescents). Having
consented to taking part in the study, the participant was allocated
randomly to receive either upper and lower BRs or upper and lower
VFRs using sequentially numbered, opaque, sealed envelopes, pre-
pared in advance of the trial by a colleague independent of the study,
using a computerized randomisation programme.
Blinding
It was not possible to blind the patient or clinician/operator in this
study following opening of the envelope due to the nature of the
intervention. Blinding of the outcome assessor looking at stability
on the digital models was also not possible, but blinding of the
patient satisfaction assessor was undertaken by removing patient-
identifying information from the completed questionnaires.
Patients leaving the study or refusing treatment
To account for potential attrition bias, an ‘intention-to-treat’ analysis
was employed for those patients who dropped-out of the study. Similarly,
if a subject failed to co-operate with the retention regime, the data were
still collected and an ‘intention-to-treat’ analysis was carried out.
Average values for that particular retainer group were substi-
tuted for the missing data. Subsequent analyses were based on the
new complete data set. The retainer survival time for subjects who
had dropped-out at 12  months was based on the 6-month review
data where the retainer had not failed previously.
Statistical methods
The SPSS software package was used for data analysis (version 20; SPSS,
Chicago, Illinois, USA) with statistical significance set at the 5% level.
The intra-class  correlation coefficient (ICC) was used to assess
the reliability of measurements made by the same operator (KF).
Twenty randomly selected digital models were benchmarked against
the ‘gold standard’ of digital callipers and plaster models. Results
ranged from 0.882 (95% confidence interval [CI] = 0.728–0.951) for
the arch length to 0.960 (95% CI = 0.805–0.987) for the inter-molar
width. Excellent agreement was associated with LII (ICC  =  0.951,
95% CI = 0.882–0.980). A further 20 randomly selected digital mod-
els were measured on two separate occasions. Excellent repeatability
was demonstrated for the LII (ICC = 0.921, 95% CI = 0.841–0.961).
Occlusal measurements, and the relapse data were found to be
non-normally distributed, therefore non-parametric analyses were
employed, using the median (Mdn) and inter-quartile range (IQR).
Mann–Whitney U-tests were used to compare relapse values between
retainer groups.
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
5
Retainer survival was analysed through the Kaplan–Meier sur-
vival plot and the log–rank test.
The patient satisfaction questionnaires were coded by the
research team and all responses entered into a database. A descrip-
tive analysis was used to compare the levels of satisfaction between
VFRs and BRs, whilst Chi-squared tests were used for trend and
significance.
Results
Participant flow and recruitment details
The CONSORT flow diagram is shown in Figure 4.
Of the 104 patients who were initially informed about the study,
44 were excluded as they either declined to be involved or did not
satisfy the inclusion criteria.
Of the 60 patients enrolled, 30 were provided with BRs (15 male,
15 female, Mdn age 16 years, IQR = 2) and 30 were provided with
VFRs (12 male, 18 female, Mdn age 17 years, IQR = 4). All patients
were recruited between March 2012 and September 2013. It was
not possible to conduct a 3-month review for eight participants
(BR  =  6, VFR  =  2) due to staff shortages in one of the hospitals.
One participant (BR = 1) did not attend their 6-month review and
3 (VFR  =  3) failed to attend their 12-month review. Furthermore,
four sets of study models taken at the 3-month review were irretriev-
able (VFR = 2, BR = 2), and one participant attending the 12-month
review declined impressions.
Baseline data
Baseline data were collected prior to the start of orthodontic treat-
ment and were found to be similar in both groups (Table  2). Any
changes in arch dimensions that occurred during active orthodontic
treatment were not found to be statistically different between the
retainer groups.
Primary outcome: stability
Occlusal measurements for each retainer were measured at debond
(T0), 3 months (T1), 6 months (T2), and 1 year (T3; Table 3). Post-
treatment changes are shown in Table 4.
In the mandible, post-treatment changes tended to be greater
in the VFR group, reaching statistical significance at both 3 and
12 months; mandibular LII values were BR = 0.77 mm IQR = 1.46,
VFR = 1.69 mm IQR = 2.00, P = 0.008 at 12 months. In the maxilla,
there was no significant difference in post-treatment changes between
the retainer groups after 12  months (BR  =  1.10  mm IQR  =  1.56,
VFR = 0.76 mm IQR = 1.55, P = 0.61).
The only statistically significant relapse measurement in arch
dimensions or occlusal relationships at 12  months was mandibu-
lar arch length (BR  =  +0.10  mm IQR  =  1.19, VFR  =  −0.70  mm
IQR = 1.25, P = 0.040).
There was no difference in the re-opening of the extraction
spaces during the initial 12  months of retention between BR and
VFR groups. Mdn space opening within the maxilla was 0.00 mm
(IQR  =  0.28) with BRs and 0.00  mm (IQR  =  0.00) with VFRs,
0.00  mm (IQR  =  0.42) with BRs and 0.00mm (IQR  =  0.25) with
VFRs within the mandible.
Secondary outcome: retainer survival
There were no statistical differences in the survival rates of BRs and
VFRs in the maxilla. The Kaplan–Meier plot for maxillary retain-
ers shows 63.3% of BR and 73.3% of VFR survived the initial
 6 European Journal of Orthodontics, 2017
Figure 4.  CONSORT flow diagram.
12  months of retention, and this difference was statistically non-
significant (P = 0.34). The mean survival time was 272.5 days (95%
CI = 226.1–319.0) for the BR, and 311.3 days (95% CI = 278.3–
344.29) for the VFR (Figure 5).
This contrasted with the mandibular retainers, where there was a
statistically significant difference in the 12-month survival of BR and
VFR (P = 0.01) with respective rates of 50% and 80%. The Kaplan–
Meier plot illustrates a mean survival time of 239.3  days (95%
CI = 191.1–287.5) for the BR, and 324.9 days (95% CI = 295.4–
354.4) for the VFR (Figure 5).
Looking at VFR survival rates after 12  months, 8 patients
(26.7%) with VFRs had experienced at least one episode of retainer
failure; either retainer loss or poorly fitting retainers or not fitting.
All eight patients exhibited failure of an upper retainer, whilst six
of these patients also had a lower retainer failure. Of these eight
patients three had repeat failures: two subjects experienced two
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
episodes of failure, whilst one patient exhibited three episodes. No
correlation was found between failure and patient gender, operator,
or study centre (P = 0.21, 0.66, 0.17, respectively).
After 12 months, 18 patients with BRs (60%) had experienced
at least one episode of retainer failure; 11 maxillary BRs and 15
mandibular BRs. Adhesive and cohesive failures were prevalent in
both arches. There was one episode involving a maxillary BR frac-
ture mesial to the right canine, and there were three episodes of the
mandibular BR completely detaching from all tooth surfaces simul-
taneously. Two episodes of maxillary BR failure occurred in three
subjects, and involved detachment at the wire-composite interface.
Three patients encountered repeated mandibular BR failures (two
patient with two episodes and three patients with three episodes).
Maxillary failure tended to occur between the wire and the compos-
ite, whereas, mandibular failure was more common at the enamel–
composite interface, particularly within the initial 6 months.
 K. Forde et al. 7

Table 2.  Baseline characteristics at T0 and treatment summary of Discussion

the sample.
Key findings and comparison to available literature
Bonded Vacuum-formed Stability and retainer survival
retainer retainer The study found no evidence of a difference in post-treatment
(N = 30) (N = 30)
changes, or retainer survival rate, between BRs and VFRs in the
Sex upper arch. In the lower arch, BRs reduced post-treatment mandibu-
 Male 15 12 lar labial segment alignment better than VFRs; however, there was
 Female 15 18 greater failure rate with BRs.
Age (years) This randomized controlled trial was conducted by two opera-
  Median (IQR) 16 (2) 17 (4) tors and aimed to compare the clinical effectiveness of BRs and VFRs
Study centre during the course of the initial 12 months of retention. Its prospec-
 Leeds 8 9 tive nature is in contrast to much of the early retrospective literature,
 Bradford 16 14
which incorporates long-term follow-up along with the associated
 York 6 7
bias (13, 31–34). Within these studies, patients were treated by a
Initial incisor classification
 class I 5 4
number of operators with varying orthodontic appliances and treat-
  class II Division 1 13 14 ment strategies, and using a variety of retainer designs (including
  class II Division 2 3 8 the type of wire used, the number of teeth bonded, and the bonding
 class III 9 4 agent).
Initial skeletal classification At the commencement of this trial the authors were aware of
  Class 1 7 6 only one published trial within the literature, directly comparable
  Class 2 14 20 to this trial (12, 22). This investigated the clinical effectiveness of
  Class 3 9 4 mandibular BRs and VFRs and was published as a series of con-
Extractions
ference abstracts. Whilst the trend in relapse was similar, with the
 Non—extraction 11 8
BR being more effective at retaining the labial segment alignment
  Maxillary extractions only 10 9
  Mandibular extractions only 2 3 than the VFR, there was, however, a difference in the failure rates.
  Bimaxillary extractions 7 10 McDermott reported greater failure of the VFR than the BR (62.5%
Treatment modality and 15.6%, respectively), whilst this current trial found failure of the
  Fixed appliances only 16 21 BR to be greater (VFR = 20%, BR = 50%).
  URA and fixed appliances 3 2 Since completion of the study, a further randomized controlled
  Functional and fixed appliances 10 4 trial comparing BRs and VFRs has been published (35). In that study,
 Orthognathic surgery and fixed 1 3 BRs and VFRs were compared in the lower arch only. They found
appliances less relapse with the BR group after 6 months, but at 12 months and
18 months there was no difference between the two types of lower
retainers. Interestingly, the amount of relapse was exceptionally low
No statistical predictors were found for the survival of either
for both groups, which may be due to a low failure rate of the BRs
upper or lower BRs. The confounders tested included the operator,
and good adherence to wearing the VFRs.
hospital as well as patient factors such as gender incisor relationship
In this study, the BR failure rates for the maxilla and mandi-
and skeletal relationship.
ble were 36.7% and 50%, respectively, for the initial 12 months of
retention. These figures may be regarded as high with failure rates
Secondary outcome: patient satisfaction of BRs constructed from multi-strand twistflex wire ranging from
A total of 52 questionnaires were completed at 3 months (BR = 24, 7.8% (13) to 53% (10). The process of placing a BR is technique
VFR = 28), 60 at 6 months (BR = 30, VFR = 30), and 57 at 12 months sensitive. The operators in this study were at the beginning of their
(BR = 30, VFR = 27). Where subjects failed to attend review appoint- orthodontic career and a general reduction in the survival rate has
ments, questionnaires were posted to patients to complete and return been associated with less experienced operators (36, 37). These fac-
in an attempt to minimise missing data. Results are reported in tors may partially explain the higher failure rate in this study.
Supplementary Table 1. An interesting finding of this research relates to the higher prev-
There was no significant difference between patients’ opin- alence of BR failures occurring in the mandibular arch compared
ions of wearing their retainers compared to their fixed appliances with the maxilla. This differs from many of the published trials,
(P = 0.096) with both groups reporting that their retainers were bet- which demonstrate greater failure rates in the maxilla probably
ter, or at least no worse, than their fixed appliances. due to occlusal stress (10, 17, 20, 38). In fact, Dahl et al. found the
There was no difference in the embarrassment felt by partici- maxillary failure rates to be almost twice as high as those found in
pants wearing either retainer type (P  =  0.065) with only two sub- the mandible (13, 38). The higher mandibular failure rate reported
jects with VFRs reporting occasional embarrassment. More patients in this study was potentially influenced by the inexperience of the
reported discomfort from VFRs than BRs (P = 0.002). More patients operators. This correlates with Scheibe’s retrospective analysis of
with VFRs reported that their speech was affected by their retainers 1062 retainers placed by operators of varying experience where the
than patients with BRs (P = 0.004). Subjects with VFRs found them majority of mandibular failures were associated with less experi-
easier to clean than those with BRs (P = 0.001). enced operators (36).
Published reports of BR survival tend to show most failures
Harms occurring within the initial 6  months of placement (9, 10, 16, 17,
No serious harm associated with the retainers was observed. Effects 39) and link these early failures to adhesive failure. This is similar to
on periodontal health will be reported in Paper 2. the findings in this study, which show that BR failures tend to occur

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

by U. of Florida Health Science Center Library user
on 16 January 2018
 8

on 16 January 2018
Table 3.  Occlusal measurements (mm) at each time point using intention-to-treat analysis. Data are presented in the form of median (inter-quartile range) and Mann–Whitney P value due to
its non-normal distribution. AL, arch length; BR, bonded retainer; ICW, inter-canine width; IMW, inter-molar width; LII, Little’s irregularity index; OB, overbite; OJ, overjet; VFR, vacuum-formed
retainer. Bold indicates statistical significance P < 0.05.

Occlusal measurements at each review

Maxilla Mandible Inter-arch

LII ICW IMW AL LII ICW IMW AL OJ OB

T0 BR 0.00 (0.93) 35.20 (2.83) 50.11 (3.96) 73.94 (12.74) 0.29 (1.02) 27.53 (1.68) 44.05 (4.64) 66.74 (6.00) 2.37 (0.70) 1.29 (1.22)
Debond VFR 0.23 (0.66) 34.09 (2.22) 48.46 (4.24) 71.23 (9.67) 0.06 (1.23) 26.17 (1.13) 41.34 (5.72) 65.53 (12.94) 2.38 (2.40) 2.00 (1.21)
P value 0.31 0.003 0.01 0.101 0.382 0.003 0.071 0.188 0.203 0.083

by U. of Florida Health Science Center Library user

T1 BR 0.79 (0.91) 35.08 (1.92) 49.63 (2.92) 72.75 (10.57) 0.44 (0.63) 27.45 (2.11) 43.65 (3.67) 66.81 (8.68) 2.31 (1.04) 1.45 (0.71)
3 months VFR 0.30 (0.98) 33.69 (2.52) 47.90 (3.88) 69.81 (11.46) 0.44 (1.71) 25.69 (1.45) 41.70 (5.35) 63.43 (9.94) 2.62 (0.76) 1.87 (0.87)
P value 0.05 0.001 0.018 0.147 0.432 0.000 0.055 0.104 0.064 0.001
T2 BR 0.88 (1.07) 35.36 (2.41) 49.74 (3.46) 72.88 (10.35) 0.67 (1.47) 28.06 (1.65) 44.23 (2.34) 67.06 (8.83) 2.53 (1.09) 1.52 (0.92)
6 months VFR 0.64 (1.77) 34.21 (2.06) 48.84 (3.85) 70.19 (9.62) 1.30 (1.98) 26.26 (1.65) 41.84 (6.22) 65.17 (11.42) 2.74 (0.50) 1.92 (0.97)
P value 0.22 0.000 0.015 0.051 0.335 0.002 0.174 0.055 0.181 0.036
T3 BR 1.35 (1.98) 35.08 (2.31) 49.47 (3.88) 76.70 (10.81) 1.01 (1.28) 27.31 (2.21) 43.90 (4.32) 66.97 (8.21) 2.26 (1.07) 1.59 (0.78)
12 months VFR 0.97 (1.68) 33.21 (2.36) 47.70 (3.80) 68.86 (10.26) 1.73 (2.77) 25.56 (1.39) 41.32 (4.61) 62.57 (9.50) 2.59 (0.94) 2.01 (1.00)
P value 0.24 0.001 0.064 0.117 0.053 0.001 0.068 0.071 0.128 0.022

Table 4.  Relapse values (measured in mm) at 3 months, 6 months, and 12 months. Data are presented in the form of median (inter-quartile range) and Mann–Whitney P value due to its non-
normal distribution. AL, arch length; BR, bonded retainer; ICW, inter-canine width; IMW, inter-molar width; LII, Little’s irregularity index; OB, overbite; OJ, overjet; VFR= vacuum-formed retainer;
PAR, peer assessment rating. Bold indicates statistical significance P < 0.05.

Changes in measurements with time

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

Maxilla Mandible Inter-arch

LII ICW IMW AL LII ICW IMW AL OJ OB

3-Month relapse (T1–T0) BR 0.29 (0.55) 35.08 (1.92) 49.63 (2.92) 72.75 (10.57) 0.15 (0.32) 27.45 (2.11) 43.65 (3.67) 66.81 (8.68) 2.32 (1.04) 1.44 (0.71)
VFR 0.04 (0.32) 33.69 (2.52) 47.90 (3.88) 69.81 (11.46) 0.33 (0.61) 25.69 (1.45) 41.70 (5.35) 63.43 (9.94) 2.62 (0.76) 1.87 (0.87)
P value 0.010 0.562 0.712 0.304 0.048 0.036 0.712 0.802 0.34 0.26
6-Month relapse (T2–T0) BR 0.39 (1.08) 0.18 (0.81) −0.63 (1.79) 0.00 (1.42) 0.53 (0.72) 0.84 (0.75) 0.19 (1.40) 0.00 (0.87) 0.08 (0.53) 0.20 (0.50)
VFR 0.30 (1.24) 0.01 (0.86) 0.15 (0.71) 0.35 (0.70) 0.70 (1.33) −0.18 (0.77) 0.49 (1.12) −0.49 (1.25) 0.14 (0.46) 0.02 (0.59)
P value 0.30 0.13 0.734 0.515 0.116 0.031 0.188 0.268 0.308 0.652
12-Month relapse (T3–T0) BR 1.10 (1.56) 0.25 (1.16) −0.16 (2.60) −0.36 (1.77) 0.77 (1.46) −0.26 (0.90) 0.11 (2.07) 0.10 (1.19) −0.09 (0.74) 0.30 (0.71)
VFR 0.76 (1.55) 0.19 (1.02) 0.02 (0.91) 0.10 (1.44) 1.69 (2.00) −0.45 (0.86) 0.54 (1.67) −0.70 (1.25) 0.12 (0.69) 0.15 (0.83)
P value 0.61 0.124 0.802 0.918 0.008 0.071 0.544 0.040 0.408 0.965
European Journal of Orthodontics, 2017
 K. Forde et al. 9

Figure  5.  Kaplan–Meier cumulative survival plots and log-rank P values for the time until the first episode of failure during 12  months of follow-up for (a)
maxillary retainers and (b) mandibular retainers.

early. The Kaplan–Meier plots within this study show failure of both
maxillary and mandibular BRs occurring within the initial 6 months
of retention. The majority of mandibular failures occurred at the
enamel–composite interface, whereas, problems more commonly
occurred at the composite–wire interface in the maxilla, perhaps in
response to occlusal interference.
The extent of relapse associated with VFRs within this current
study is comparable to reports within the literature. Within our
study, we noted similar failure rates between maxillary and man-
dibular VFRs and the majority of these were due to participants los-
ing their retainers. In contrast, others have found retainer breakages
to be the most common reason for failure, followed by lost retain-
ers. Sun et al. instructed all participants to wear their retainers on a
full-time basis, and perhaps this increased wear contributed to the
higher breakage rates (40).
Patient satisfaction
There was no significant difference between patients’ opinions of
wearing their retainer compared to their fixed appliances with both
groups reporting that their retainers were better or at least no worse
than their fixed appliances. This contrasts with the results of previ-
ous studies which have highlighted patients’ disappointment with
retainers, noting that they found retainers very difficult to live with,
more so than braces and headgear (25–27), mainly due to difficulties
with speech, eating, increased salivation, smell, losing their retainers,
and embarrassment over the appearance of the retainers (25, 26).

Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723

by U. of Florida Health Science Center Library user
on 16 January 2018
 10
Only 33% of subjects with VFRs reported that they always wore
their retainers every night as instructed, though a further 37% stated
they followed this advice most of the time. This finding is similar to
Sawhney et al. who stated that self-reported compliance with remov-
able retainers ranged between 75% and 85%, regardless of the regi-
men or type of retainer (41), and Wong and Freer who reported more
than 50% of patients admitted that they did not wear retainers as
instructed (24).
Eighty-three percent of participants with VFRs and 100% with
BRs stated that they were not embarrassed to wear their retainers,
though VFRs were associated with more discomfort (37%) and dif-
ficulty speaking (47%) than BRs, though the relevance of difficult
speaking with the retainer is questionable when VFRs are prescribed
for night-time wear only.
Jaderberg et al. also found similar problems with VFRs, albeit to
a lesser extent, reporting soreness in 13%, and problems with speech
in 22% of subjects (29). Similarly, Sawhney et al. found BRs affected
speech less than both VFRs and Hawley retainers (41).
Perhaps unsurprisingly, subjects in the VFR group found it easier
to clean their retainers than those with BRs, supporting previous
findings (41).
Limitations
In this study, it was not possible to blind the operators and patients
to the treatment allocations.
Although the minimum sample size was achieved, the power cal-
culation was based on the primary outcome measure: the efficacy of
each retainer group in minimising the 1-year post-orthodontic treat-
ment change in arch alignment. As such, it is possible that the study
is underpowered with respect to the secondary outcome measures,
increasing the chance of making a type II error.
The amount of VFR wear was not measured, so it is not clear
whether the extra relapse in the mandible was due to the retainer
itself, or a lack of patient compliance with the prescribed retention
regimen. It could be argued that this is the ‘real-life’ situation.
Post-treatment changes were measured in this study, rather than
attempting to measure relapse towards the starting malocclusion.
This is because it was felt that increases in incisor irregularity after
treatment was more clinically relevant for patients. Post-treatment
changes are unpredictable and do not always represent a move back
towards the original malocclusion. This study therefore measures the
ability of BRs and VFRs to reduce post-treatment changes in inci-
sor irregularity, which may or may not reflect a relapse towards the
presenting malocclusion.
LII was chosen as the principle measurement of relapse as, anec-
dotally, incisor irregularity is often the patient’s main concern relat-
ing to post-orthodontic movement. This method is quick, simple,
and widely used within research, therefore enabling direct compari-
son between studies. However, recognized limitations of this method
include a disregard to spacing, or mutual rotations, where the con-
tact point remained intact. Furthermore, the index is cumulative;
therefore, a LII of 1.5 mm may have been awarded to a patient with
evenly dispersed (but clinically insignificant) minor displacements,
or equally to another patient with a single contact point displace-
ment of the same value. It is likely that if the irregularity is limited
to one larger displacement, then a patient will notice it more than if
the same total amount of irregularity is evenly distributed across a
number of teeth.
A statistical difference in relapse of mandibular irregularity was
found between the two groups. We considered whether to use a
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
European Journal of Orthodontics, 2017
Bonferonni correction to allow for the possible affects of multiple
testing, but in line with other similar studies, this was not used in this
study for this outcome. This may be a statistical approach that could
be considered in future research.
Randomisation is undertaken to ensure equivalence between the
two groups. It should be noted that there was statistical difference
in the upper and lower inter-canine width at debond, although the
differences were so small that it is unlikely to be clinically significant.
The randomisation process did not enable exclusion of subjects
where placement of the BR was deemed difficult, such as those with
heavy, labial segment occlusal contacts. The retainer wire was con-
toured gingivally (where possible) to reduce, and hopefully avoid,
traumatic occlusal contact. The influences of the incisor classifica-
tion and overbite on failure could be investigated separately with
a larger sample. All retainers were placed by orthodontic trainees.
In an attempt to minimize the impact of operator inexperience,
standardized positioning jigs were employed to aid placement of the
retainers and reduce the potential for operator error and the risk of
moisture contamination.
In this study, the BRs were used in isolation to best assess the
effects caused by them without the influence of an additional remov-
able retainer. Some operators supplement a BR with a removable
retainer in an attempt to maintain buccal segment alignment, retain
space closed in extraction cases, and to act as a safeguard in case of
breakage of the BR. A  future study looking at this dual retention
may be useful.
Patient satisfaction is a complex area and is related to a number
of different factors, which are not evenly weighted. The question-
naire used in the study was based on the most relevant questions in
the patient satisfaction questionnaire from the Hichens et al. rand-
omized controlled trial assessing patient satisfaction with Hawley
and VFRs (28). To aid direct comparison of the data obtained with
that from a previous study. Further research may be needed to devise
a more appropriate and validated questionnaire for different types
of retainers. This may involve the use of focus groups and patient
interviews to identify specific concerns or problems.
One of the challenges of orthodontic research is following up
patients over a long period and this is particularly pertinent to
research into retention. Due to the problem of large drop-outs in
long-term orthodontic studies, this study only measured changes
over 1  year. It is important to remember that the findings of this
study only relate to 12 months of retention and it is possible that the
findings may be different over a longer period.
Generalisability
The generalisability of these findings is limited to the retainer mate-
rials, design, and technique of placement, in addition to the opera-
tor experience. The operators were orthodontic trainees, therefore
the clinical outcomes, in particular the survival rates, may be not
be transferable to those with greater experience in the placement
of BRs.
Implications for clinical practice
Within the initial 12 months of retention, BRs were more effective at
retaining mandibular labial segment alignment than VFRs. However,
there was a significantly higher failure rate relating to mandibu-
lar BRs, which had often progressed unnoticed by the patient.
Therefore, it may be worthwhile considering whether a reduction
of, on average, 0.92 mm of mandibular labial segment relapse war-
rants the additional clinical time spent during the supervision and
 K. Forde et al.
amending of BR failures, particularly if the relapse is negligible once
distributed across the full labial segment. If BRs are intended for the
labial segment, it may be prudent to arrange more frequent recalls,
particularly in the first 6  months after debond, to supervise BRs,
especially if a less experienced operator had placed the BR. This
would be beneficial not only in terms of minimising relapse, but also
for the development of the operator’s awareness and technique. Prior
to retainer placement, patients should also be counselled about the
possibility of failure, and the importance of monitoring the integrity
of the retainer.
Within the maxilla, there was no evidence of a difference in
relapse or retainer failure rates when comparing BRs and VFRs,
so other outcomes may affect the choice of retainer. This will be
explored further in Paper 2
Future research
Future research should investigate the outcomes and implications of
retainers over the longer term in terms of post-treatment changes,
failure rates, adverse effects, and patient compliance. In addition, it
would be useful to explore different types of BRs and VFRs.
Conclusions
After 12 months of retention, this randomized controlled trial com-
paring bonded with VFRs has shown that:
- There is no evidence of a difference in post-treatment changes
between patients who wear BRs or VFRs in the maxilla
- The mandibular BR is more effective at reducing post-treatment
changes in the labial segment alignment
- There is no evidence of a difference in the survival rate of maxil-
lary BRs and VFRs
- BRs are more likely to fail than VFRs
- Patients find VFRs easier to clean than BRs. However, patients
report that BRs cause less discomfort and speech difficulties, and
require less compliance
Supplementary Material
Supplemental data are available at the European Journal of
Orthodontics online.
Acknowledgements
We would like to acknowledge and thank Nadia Ahmed for her assistance
during the early stages of the trial, and each of the laboratory technicians,
reception staff, referring clinicians, and each patient and parent at the three
hospitals.
Conflict of Interest
None to declare.
References
1. Littlewood, S.J., Millett, D.T., Doubleday, B., Bearn, D.R. and Worthington,
H.V. (2016) Retention procedures for stabilising tooth position after treat-
ment with orthodontic braces. Cochrane Database of Systematic Reviews
2016, Issue 1. Art. No.: CD002283. DOI: 10.1002/14651858.CD002283.
pub4.
2. Zachrisson, B.U. (2007) Long-term experience with direct-bonded retain-
ers: update and clinical advice. Journal of Clinical Orthodontics, 41, 728.
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
11
3. Renkema, A.M., Sips, E.T., Bronkhorst, E. and Kuijpers-Jagtman, A.M.
(2009) A survey on orthodontic retention procedures in The Netherlands.
The European Journal of Orthodontics, 31, 432–437.
4. Rowland, H., Hichens, L., Williams, A., Hills, D., Killingback, N., Ewings,
P., Clark, S., Ireland, A.J. and Sandy, J.R. (2007) The effectiveness of Haw-
ley and vacuum-formed retainers: a single-center randomized controlled
trial. American Journal of Orthodontics and Dentofacial Orthopedics,
132, 730–737.
5. Gill, D.S., Naini, F.B., Jones, A. and Tredwin, C.J. (2007) Part-time versus
full-time retainer wear following fixed appliance therapy: a randomized
prospective controlled trial. World Journal of Orthodontics, 32, 300–306.
6. Thickett, E. and Power, S. (2010) A randomized clinical trial of thermo-
plastic retainer wear. The European Journal of Orthodontics, 32, 1–5.
7. Bondemark, L. Holm, A.K., Hansen, K., Axelsson, S., Mohlin, B.,
Brattstrom, V., Paulin, G. and Pietila, T. (2007) Long-term stability of
orthodontic treatment and patient satisfaction: a systematic review. The
Angle Orthodontist, 77, 181–191.
8. Valiathan, M. and Hughes, E. (2010) Results of a survey-based study to
identify common retention practices in the United States. American Jour-
nal of Orthodontics and Dentofacial Orthopedics, 137, 170–177.
9. Al-Nimri, K., Al Habashneh, R. and Obeidat, M. (2009) Gingival health
and relapse tendency: a prospective study of two types of lower fixed
retainers. Australian Orthodontic Journal, 25, 142.
10. Störmann, I. and Ehmer, U. (2002) A prospective randomized study of
different retainer types. Journal of Orofacial Orthopedics/Fortschritte der
Kieferorthopädie, 63, 42–50.
11. Årtun, J., Spadafora A.T. and Shapiro, P.A. (1997) A 3-year follow-up
study of various types of orthodontic canine-to-canine retainers. The
European Journal of Orthodontics, 19, 501–509.
12. McDermott, P., Field, D., Erfida, I. and Millett, D.T. (2007) Operator and
Patient Experiences With Fixed or Vacuum Formed Retainers. Interna-
tional Association of Dental Research, Cork, Vol. 17.
13. Dahl, E.H. and Zachrisson, B.U. (1991) Long-term experience with direct-
bonded lingual retainers. Journal of Clinical Orthodontics: JCO, 25, 619–
630.
14. Bearn, D.R. (1995) Bonded orthodontic retainers: a review. American
Journal of Orthodontics and Dentofacial Orthopedics, 108, 207–213.
15. Bearn, D.R., McCabe, J.F., Gordon, P.H. and Aird, J.C. (1997) Bonded
orthodontic retainers: the wire-composite interface. American Journal of
Orthodontics and Dentofacial Orthopedics, 111, 67–74.
16. Foek, D.L.S., Ozcan, M., Verkerke, G.J., Sandham, A. and Dijkstra, P.U.
(2008) Survival of flexible, braided, bonded stainless steel lingual retain-
ers: a historic cohort study. The European Journal of Orthodontics, 30,
199–204.
17. Saidler, G., J. McColl, and Stat, C. (1999) Breakage incidence with direct
bonded lingual retainers. British Journal of Orthodontics, 26, 191–194.
18. Renkema, A.-M., Renkema, A., Bronkhorst, E. and Katsaros, C. (2011)
Long-term effectiveness of canine-to-canine bonded flexible spiral wire
lingual retainers. American Journal of Orthodontics and Dentofacial
Orthopedics, 139, 614–621.
19. Abudiak, H., Shelton, A., Spencer, R.J., Burns, L. and Littlewood, S.J.
(2011) A complication with orthodontic fixed retainers: A  case report.
Orthodontic Update, 4, 112–131.
20. Andrén, A., Asplund, J., Azarmidohkt, E., Svensson, R., Varde, P. and
Mohlin, B. (1997) A clinical evaluation of long term retention with
bonded retainers made from multi-strand wires. Swedish Dental Journal,
22, 123–131.
21. Luther, F. and Nelson-Moon, Z. (2012) Chapter 9 in Orthodontic Retain-
ers and Removable Appliances: Principles of Design and Use. John Wiley
& Sons, Chichester, UK.
22. McDermott, P., Millett, D.T., Field, D., Van den Heuvel, A. and Erfid, I.
(2008) Lower Incisor Retention With Fixed or Vacuum Formed Retainers.
International Association of Dental Research, Toronto, pp. 642.
23. Williams, A. (2003) How to… Write and analyse a questionnaire. Journal
of Orthodontics, 30, 245–252.
24. Wong, P. and Freer, T.J. (2005) Patients’ attitudes towards compliance with
retainer wear. Australian Orthodontic Journal, 21, 45–53.
 12
25. Bennett, M.E. and Tulloch, J. (1999) Understanding orthodontic treatment
satisfaction from the patients’ perspective: a qualitative approach. Clinical
Orthodontics and Research, 2, 53–61.
26. Travess, H.C., Newton, J.T., Sandy, J.R. and Williams, A.C. (2004) The
development of a patient-centered measure of the process and outcome of
combined orthodontic and orthognathic treatment. Journal of Orthodon-
tics, 31, 220–234.
27. Vig, K.W., Weyant, R., O’Brien, K. and Bennett, E. (1999) Developing out-
come measures in orthodontics that reflect patient and provider values.
Seminars in Orthodontics, 5, 85–95.
28. Hichens, L., Rowland, H., Williams, A., Hollinghurst, S., Ewings, P., Clark,
S., Ireland, A. and Sandy J. (2007) Cost-effectiveness and patient satisfac-
tion: Hawley and vacuum-formed retainers. European Journal of Ortho-
dontics, 29, 372–378.
29. Jäderberg, S., I.  Feldmann and Engström, C. (2012) Removable ther-
moplastic appliances as orthodontic retainers—a prospective study of
different wear regimens. The European Journal of Orthodontics, 34,
475–479.
30. Little, R.M. (1975) The irregularity index: a quantitative score of man-
dibular anterior alignment. American Journal of Orthodontics, 68, 554–
563.
31. Zachrisson, B.U. (1977) Clinical experience with direct-bonded orthodon-
tic retainers. American Journal of Orthodontics, 71, 440–448.
32. Little, R.M., Wallen, T.R. and Riedel, R.A. (1981) Stability and relapse of
mandibular anterior alignment—first premolar extraction cases treated by
traditional edgewise orthodontics. American Journal of Orthodontics, 80,
349–365.
Downloaded from https://academic.oup.com/ejo/advance-article-abstract/doi/10.1093/ejo/cjx058/4558723
by U. of Florida Health Science Center Library user
on 16 January 2018
European Journal of Orthodontics, 2017
33. Little, R.M., Riedel, R.A. and Artun, J. (1988) An evaluation of changes in
mandibular anterior alignment from 10 to 20 years postretention. Ameri-
can Journal of Orthodontics and Dentofacial Orthopedics, 93, 423–428.
34. Little, R. (2009) Clinical implications of the University of Washington
post-retention studies. Journal of Clinical Orthodontics: JCO, 43, 645.
35. O’Rourke, N., Albeedh, H., Sharma, P. and Johal, A. (2016) Effectiveness
of bonded and vacuum-formed retainers: a prospective randomized con-
trolled clinical trial. American Journal of Orthodontics and Dentofacial
Orthopedics, 150, 406–415.
36. Scheibe, K. and Ruf, S. (2010) Lower bonded retainers: survival and fail-
ure rates particularly considering operator experience. Journal of Orofa-
cial Orthopedics/Fortschritte der Kieferorthopädie, 71, 300–307.
37. Schneider, E. and Ruf, S. (2011) Upper bonded retainers: survival and fail-
ure rates. The Angle Orthodontist, 81, 1050–1056.
38. Segner, D. and Heinrici, B. (2000) Bonded retainers–clinical reliability.
Journal of Orofacial Orthopedics/Fortschritte der Kieferorthopädie, 61,
352–358.
39. Taner, T. and Aksu, M. (2012) A prospective clinical evaluation of man-
dibular lingual retainer survival. The European Journal of Orthodontics,
34, 470–474.
40. Sun, J., Yu, Y.C., Liu, M.Y., Chen, L., Li, H.W., Zhang, L., Zhou, Y.,
Ao, D., Tao, R. and Lai, W.L. (2011) Survival time comparison between
Hawley and clear overlay retainers: a randomized trial. Journal of Dental
Research, 90, 1197–1201.
41. Sawhney, B. (2013) Orthodontic retainers: a survey of patient compliance
and satisfaction. Electronic Thesis and Dissertation Repository, The Uni-
versity of Western Ontario, Citeseer.