Introduction
The clinical effectiveness of aligners is impacted by wear time (WT). This study aimed to objectively assess WT using microsensors and to evaluate whether patient awareness of monitoring affects WT.
Methods
This study was designed as a single-center, 2-arm, parallel-group randomized controlled trial in 43 adult patients. Orthodontic aligner treatment with an integrated microsensor was used to assess WT across 6 consecutive appointments, scheduled biweekly. Patients were randomly assigned to group 1 or 2 in a 1:1 ratio using an online randomization tool. Allocation was concealed using sequentially numbered opaque, sealed envelopes. Patients were either informed about the use of microsensors to register intraoral temperature (group 1, “unaware”) or to assess patient compliance based on the temperature data (group 2, “aware”). Operators and patients could not be blinded to the treatment allocation. However, outcome assessment was blinded. A linear mixed model was used to estimate the effect of study arm on WT per wave, with age and sex included as explanatory variables.
Results
At follow-up, 3 patients discontinued intervention. Forty patients were included in the final analysis (male, 12; female, 28; mean age, 36.2 ± 14.2 years). The linear mixed model showed that awareness of monitoring (group 2) resulted in significantly higher WT, with an increase of 4.41 h/d (95% confidence interval, 1.88-6.93, P = 0.0011). WT declined significantly at later visits. No harms or adverse effects were reported or observed.
Conclusions
Awareness of being monitored improved patient compliance and can be used as an effective strategy to enhance WT during aligner therapy in adults.
Trial Registration
German Clinical Trials Register (DRKS00029027).
Highlights
-
•
First study to objectively analyze aligner wear-time (WT) behavior in orthodontic patients.
-
•
High individual WT variability was observed among patients undergoing aligner treatment.
-
•
Most participants (67.5%) wore aligners between 10-20 h/d.
-
•
A linear mixed model showed that awareness, sex, and visit time impacted WT.
-
•
Awareness of monitoring (group 2) resulted in significantly higher WT, with an increase of 4.41 h/d.
Aligners are removable thermoplastic appliances used to perform orthodontic tooth movements. Introduced approximately 70 years ago for the treatment of mild malocclusions, such as dental crowding, aligner systems have since significantly advanced, especially by incorporating computer-aided design and manufacturing technology. , Today, aligner therapy addresses a wide range of clinical indications, including the treatment of patients with Class II and III malocclusion, open bites, and extractions. ,,, Nevertheless, the clinical effectiveness of aligners remains limited compared with fixed orthodontic multibracket appliances. , Virtually planned tooth movements are not fully achieved with aligners, demonstrating an overall mean accuracy between 50%-60%. ,,,
Several factors contribute to this discrepancy. Aligners primarily depend on their material properties to exert orthodontic force; however, the biomechanical force system required for specific tooth movements is not necessarily achieved by clinical deformation of the thermoplastic appliance. ,, The use of attachments, elastic chains, elastics, and other auxiliaries can enhance force application, but does not entirely compensate for the biomechanical limitations. ,
Another factor affecting clinical effectiveness is wear time (WT). Given that aligners are removable appliances, consistent wear is required to achieve the planned tooth movements. , For aligners, a daily WT of approximately 20-22 h/d is commonly recommended. WT can be assessed subjectively using questionnaires or objectively using microsensors. , Studies have shown that patients typically overestimate their WT when subjective and objective assessments are compared. , Objectively assessed WT was found to be shorter than orthodontist-recommended WT for all types of removable orthodontic appliances, typically by 3-6 h/d. , Compliance, defined as patients’ adherence to a prescribed WT, varies depending on appliance type, patient sex, age, treatment stage, psychosocial and other factors, though specific influences and mechanisms remain incompletely understood. ,,,
Microsensors have previously been used to objectively assess the WT of different removable appliances, including functional orthodontic appliances, retention appliances, and appliances for the treatment of obstructive sleep apnea. ,, However, such objective monitoring techniques have not yet been applied to aligner therapy.
Objectives and hypotheses
This randomized controlled trial aimed to objectively assess the WT of aligners using microsensors and to evaluate whether patient awareness of monitoring affects WT. The null hypothesis was that patient awareness of monitoring does not significantly influence aligner WT.
Material and methods
Trial design
The study was conducted as a prospective, single-center, 2-arm, parallel randomized controlled trial with a 1:1 allocation ratio. The study was approved by the Ethics Committee of the Ludwig-Maximilian-Universität München (LMU) Medical Faculty on March 24, 2022 (No. 21-1190) and registered at the German Clinical Trials register (DRKS-ID 00029027) on May 23, 2022. The trial protocol, full dataset, including statistical analysis code are publicly available via Open Data LMU. All methods were performed in accordance with the Declaration of Helsinki and relevant guidelines and regulations.
Participants
Patients from the Department for Orthodontics and Dentofacial Orthopedics, LMU University Hospital, were asked to participate. Patients received oral and written information about the study. On agreement to participate, informed consent was obtained.
The following inclusion criteria were applied: (1) adult patients (aged >18 years), competent to provide informed consent; (2) indication for orthodontic treatment, and (3) dental malocclusion in the form of crowding or spacing.
The following exclusion criteria were applied: (1) skeletal malocclusions, (2) craniofacial anomalies or syndromes, (3) temporomandibular dysfunction, (4) mental health conditions, (5) diseases affecting motor or cognitive function, and (6) presence of dental implants.
Interventions
WT was objectively assessed using microsensors (Theramon; MC Technology GmbH, Hargelsberg, Austria), which were integrated into orthodontic aligners fabricated in-office as follows: Patients were scanned using an optical intraoral scanner (Trios 3, 3Shape, Denmark). After data export in standard tessellation language format, the scan data were processed using the orthodontic imaging software OnyxCeph (Image Instruments GmbH, Chemnitz, Germany). The aligner treatments were virtually planned with the Aligner3D module of the software. Treatment planning was conducted by a single experienced orthodontist (H.S.). The aligner setup models were exported in standard tessellation language format with a digitally integrated virtual analog of the microsensor ( Fig 1 , A ). The models were 3-dimensional printed using a liquid crystal display printer (Phrozen Sonic 4K; Phrozen Technology Co, Ltd, Hsinchu City, Taiwan) and postprocessed according to the manufacturer’s protocol. Aligners were produced by thermoforming using co-polyester sheets (CAPro+; Scheu GmbH, Iserlohn, Germany).
Placement of the sensor: A, Digital analog of microsensor; B, Intraoral photograph with clinically inserted aligners and microsensor.
All patients received the same intervention (aligner therapy) and standardized instructions regarding aligner handling and were instructed to wear the aligners for 20 h/d ( Fig 1 , B ). As part of the study, appointments were made approximately every 14 days for 6 consecutive appointments to read the microsensor and transfer it to the next aligner, corresponding to a clinical follow-up of approximately 12 weeks. The trial period began with the insertion of the first aligner and continued through the sixth aligner.
Outcomes
The primary outcome measure was WT in dependence on study arm (aware vs unaware) across 6 consecutive appointments. WT was assessed by temperature recordings of the microsensor every 15 minutes with an accuracy of ± 0.1°, recording 96 data points per day. No changes were made after trial commencement.
Sample size
No directly comparable studies objectively assessing WT in aligner therapy using microsensors were available. A related study investigated the effect of monitoring awareness on WT in patients wearing removable orthodontic appliances. However, it was not directly comparable, as patient awareness was introduced at the first follow-up rather than at the start of treatment, and the study assessed part-time wear (15 hours prescribed) in contrast to full-time wear, as in this study. Because of the lack of directly applicable data, assumptions were made based on published findings from that study. Sample size calculation revealed that 40 study subjects (20 per group) were required to detect an effect size (Cohen’s d ) of 0.9 in mean WT between appointments with a power of = 80% (version 3.1.9.6, G∗Power; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany). On the basis of published data on mean WT between appointments taken from this publication, an effect size of Cohen’s d of 0.9 corresponds to an average daily WT (eg, 8.1 ± 3.05 h/d in the test group and 5.7 ± 2.06 h/d in the control group). A dropout rate of 10% was anticipated and accounted for in the sample size calculation. To account for dropouts beyond the expected rate, eligibility issues after randomization, or administrative errors, 50 sealed randomization envelopes were prepared. However, per protocol, recruitment was terminated once 40 participants had completed the study.
Randomization and blinding
Patients were randomly allocated to either group 1 or 2 using sequentially numbered, opaque, sealed envelopes prepared and administered by a blinded, nonclinical scientist (U.B.). The 2 groups received different information about the study:
-
1.
Group 1 (“unaware”): Patients were informed about the use of microsensors to register intraoral temperature.
-
2.
Group 2 (“aware”): Patients were specifically informed about the use of microsensors to assess patient compliance based on the temperature data.
The sequence of the numbers 1-50 was randomized using the online program Research Radomizer (version 4.0; https://www.randomizer.org/ ; program setting: 1 set, 50 numbers per set with number range set to 1-50, each number unique, no sorting). Odd numbers were assigned to group 1, whereas even numbers were assigned to group 2. Fifty envelopes were prepared, numbered from 1 to 50. Two copies of the information sheet corresponding to the specific group designation were inserted, and the envelop sealed. The sealed envelope was handed over to the clinical investigators immediately before providing informed consent. Patients received both oral and written information about the study. On agreement to participate, informed consent was obtained. Two clinicians (V.V. and S.T.) were responsible for reading the microsensors at the appointments. The sensor IDs were registered to randomization numbers, allowing anonymous data collection across appointments. Data analysis was performed on the anonymized dataset by a blinded, nonclinician researcher (U.B.).
Statistical methods
All measurement data were transferred to a Microsoft Excel spreadsheet (Microsoft Corp, Redmond, Wash). Statistical analysis was performed using SPSS (version 29.0.2.0; IBM, Armonk, NY) and R (version 4.4.3; R Core Team, Vienna, Austria). Quantitative variables (eg, WT, age, and days between appointments) were reported as mean with standard deviation and median with interquartile range (IQR). Qualitative variables (eg, sex, visit, and study arm) were reported as frequency counts and the proportion of observations for each category. A linear mixed model was established to compare WT over the observation period. WT (continuous) was considered as the response variable, and study arm (group 1 = 0; group 2 = 1), wear period (visit T0-T1 = 1, …, visit T5-T6 = 6), sex (male = 0; female = 1), and age (continuous) as explanatory variables. To determine the individual random variability, subject ID (nominal; 1…40) was included as a random intercept. The explanatory variables and interactions between them (study arm × sex and study arm × age) were added one after the other to the model (estimated using machine learning and nloptwrap optimizer), and their contribution to the model was estimated using likelihood ratio tests and Akaike information criterion. The finally accepted model was estimated using REML and nloptwrap optimizer, and contained study arm, wear period, sex, and study arm × sex interaction as explanatory variables. Standardized parameters were obtained by fitting the model on a standardized version of the dataset. Confidence intervals (CIs) values and P values were computed using a Wald t distribution with Kenward-Roger approximation. Model quality parameters were visually inspected, including checks related to linearity, heteroscedasticity (homogeneity of variance), influential observations (Cook’s distance), collinearity, normality of residuals, and of random effects. Potential outlier observations and influential subjects were addressed by selectively removing these from the dataset and refit of the model. To deal with the nonnormality of model residuals, several transformations of the response variable were assessed, including cubic and Box-Cox transformations, and their effect on the normality of residuals was visually determined. The level of significance was set to an α of <0.05.
Model fitting was done using the R packages lme4 (version 1.1.37) and lmerTest (version 3.1.3). , Model quality was assessed using the R package performance (version 0.15.1). Box-Cox transformation of WT was determined and applied to the “reduced” model using car (version 3.1.3) and effects (version 4.2.4). Multiple comparisons of model factors and their effect on WT were calculated with emmeans (version 1.11.2.8). Reports of the models were prepared with the aid of the R packages from easystats (version 0.7.5) including report (version 0.6.1) and the performance’ package, sjPlot (version 2.9.0), and the packages from the tidyverse (version 2.0.0). All analysis steps were documented in a Quarto script and the corresponding HTML file using RStudio (version 2025.05.1.513; Posit Software, PBC, Boston, Mass). , These 2 files, together with the raw data, are publicly available (see statement on data availability).
Results
Participant flow and recruitment
Patient enrolment started on January 4, 2023, and ended on October 8, 2024. A total of 55 patients were initially screened for eligibility. Twelve patients were excluded because of the exclusion criteria or because they declined to participate ( Fig 2 ). A total of 43 patients were randomized, and 3 discontinued the intervention, resulting in 40 patients included in the final analysis.
Participant flow diagram illustrating patient recruitment, allocation, follow-up, and analysis stages.
Baseline data
Forty patients (median age, 31.1 years; IQR, 25.8-48.2 years; range 17-72 years) were included for analysis, 20 in group 1 (5 males and 15 females) and 20 in group 2 (7 males and 13 females). The baseline data for the total study and both study arms concerning patients’ age and sex are shown in Table I . Although extraction patients were not excluded by the eligibility criteria, none were included in the final sample. All patients received aligner treatment in both arches.
Table I
Patient demographics and distribution by study arm, including age, gender, and total number of participants
| Variables | Total (N = 40) |
Group 1 (unaware)
(n = 20) |
Group 2 (aware)
(n = 20) |
|---|---|---|---|
| Age (y) | |||
| Mean ± SD | 36.2 ± 14.2 | 33.4 ± 14.7 | 39.0 ± 13.4 |
| Median (IQR) | 31.1 (25.8-48.2) | 30.0 (23.4-33.1) | 41.4 (26.4-51.8) |
| Sex, n (%) | |||
| Male | 12 (30.0) | 5 (12.5) | 7 (17.5) |
| Female | 28 (70.0) | 15 (37.5) | 13 (32.5) |
SD , standard deviation.
Outcomes
Patients were scheduled for 6 clinical appointments following a biweekly protocol. However, because of factors such as professional obligations, illness, and other circumstances, strict adherence to the prescribed schedule was not always possible. The median interval between 2 consecutive appointments was 14.3 days (IQR, 13.9-15.0 days) in group 1 and 14.5 days (IQR 14.0-15.5 days) in group 2 ( Table II ). Supplementary Data presents the distribution of visits among the patients, illustrating both the clustering of appointment waves and the variability in individual time intervals between consecutive appointments.
Table II
Number of days between the appointments (periods), and overall mean time between appointments in days
| Wear periods | Group 1 (unaware) (n = 20) | Group 2 (aware) (n = 20) |
|---|---|---|
| T0-T1 (d) | ||
| Mean ± SD | 13.9 ± 1.9 | 15.2 ± 3.3 |
| Median (IQR) | 14.0 (13.0-14.0) | 14.0 (14.0-15.0) |
| T1-T2 (d) | ||
| Mean ± SD | 14.0 ± 2.3 | 15.0 ± 2.0 |
| Median (IQR) | 14.0 (14.0-14.0) | 14.0 (14.0-15.0) |
| T2-T3 (d) | ||
| Mean ± SD | 15.1 ± 4.2 | 15.4 ± 3.9 |
| Median (IQR) | 14.0 (14.0-14.0) | 14.0 (13.5-15.5) |
| T3-T4 (d) | ||
| Mean ± SD | 13.9 ± 0.9 | 14.2 ± 1.5 |
| Median (IQR) | 14.0 (14.0-14.0) | 14.0 (14.0-14.0) |
| T4-T5 (d) | ||
| Mean ± SD | 14.7 ± 2.8 | 14.6 ± 1.6 |
| Median (IQR) | 14.0 (14.0-15.0) | 14.0 (14.0-14.0) |
| T5-T6 (d) | ||
| Mean ± SD | 17.1 ± 7.9 | 15.0 ± 3.9 |
| Median (IQR) | 14.0 (14.0-15.0) | 14.0 (14.0-15.0) |
| T0-T6 (d) | ||
| Mean ± SD | 14.8 ± 1.8 | 14.9 ± 1.1 |
| Median (IQR) | 14.3 (13.9-15.0) | 14.5 (14.0-15.5) |
SD , standard deviation.
Trajectories of mean daily WT for each study participant showed substantial individual variability, both in the initial mean daily WT in the first wear period (T0-T1), but also in its progression over time ( Fig 3 ). Only a few subjects who were unaware of WT monitoring (group 1) consistently wore the aligners for fewer than 10 h/d between appointments (IDs 2, 13, and 34). In contrast, several participants who were aware of monitoring (group 2) consistently achieved WTs around or exceeding 20 h/d (IDs 8, 14, 22, 23, 24, 27, and 40), as did some participants unaware of monitoring (IDs 5, 7, 18). Most participants (27 of 40 [67.5%]) wore the aligners for 10-20 h/d.
Mean daily WT (hours) and slope for each of the 40 individual patients across the 6 follow-up appointments. The individual slope was calculated using ordinary least squares fitting implemented in the base R function lm .
Descriptive statistics of daily WT at each of the 6 consecutive follow-up intervals, as well as the overall mean WT throughout the entire observation period, are summarized in Table III .
Table III
Descriptive statistics of mean WT (in hours) between the appointments (ie, wear periods), and overall mean time between appointments (wear period T0-T6) for the complete study population and for each study arm
| Wear periods | Group 1 (unaware) (n = 20) | Group 2 (aware) (n = 20) |
|---|---|---|
| T0-T1 | ||
| Mean ± SD | 14.3 ± 4.3 | 17.1 ± 3.1 |
| Median (IQR) | 14.7 (12.8-17.0) | 17.4 (14.9-19.6) |
| T1-T2 | ||
| Mean ± SD | 14.5 ± 4.6 | 17.3 ± 3.4 |
| Median (IQR) | 15.8 (12.0-17.1) | 17.7 (15.4-20.1) |
| T2-T3 | ||
| Mean ± SD | 14.2 ± 5.0 | 16.9 ± 3.9 |
| Median (IQR) | 14.7 (11.8-17.5) | 17.2 (15.3-19.9) |
| T3-T4 | ||
| Mean ± SD | 13.7 ± 4.9 | 16.7 ± 3.5 |
| Median (IQR) | 14.2 (10.8-17.0) | 16.9 (15.4-19.5) |
| T4-T5 | ||
| Mean ± SD | 13.1 ± 4.9 | 16.7 ± 3.2 |
| Median (IQR) | 12.6 (10.3-16.6) | 17.4 (14.4-19.4) |
| T5-T6 | ||
| Mean ± SD | 12.3 ± 4.6 | 15.7 ± 3.9 |
| Median (IQR) | 12.5 (9.1-14.9) | 16.0 (12.2-19.4) |
| T0-T6 (overall) | ||
| Mean ± SD | 13.7 ± 4.4 | 16.7 ± 3.2 |
| Median (IQR) | 14.8 (11.7-16.1) | 16.7 (14.6-19.7) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
