Trial design and changes after trial commencement

This study was a multicenter, 2-arm, parallel, randomized, triple-blinded, active-controlled clinical trial with a 1:1 allocation ratio. Institutional Review Board approval was obtained from the University of Washington Human Subjects Division before the study on May 15, 2015 (unique protocol number 49073-D). This study was registered at ClinicalTrials.gov with the original release on May 4, 2015 ( NCT02438280 ).

This study was originally designed to be conducted at 1 center in Seattle, Wash. In December 2015, with the approval of the University of Washington Human Subjects Division, an additional subject recruitment site in Vancouver, British Columbia, Canada, was added. This change was immediately reflected on ClinicalTrials.gov . No other changes or modifications were made to the original study protocol and methodology of the research.

The study was designed to evaluate the effects of a supplemental vibrational device (30 Hz and 0.25 N) when used in conjunction with clear aligners in adult subjects.

Participants, eligibility criteria, and setting

Orthodontic patients who were beginning their Invisalign (Align Technology, San Jose, Calif) treatment in 1 of the 2 approved private practices in Seattle or Vancouver were invited to participate in this study. Adults (18 years or older) with malocclusions whose treatment involved fewer than 25 sets of aligners were invited to participate. Also, no significant anteroposterior or transverse movements were planned. For example, patients whose treatment plans included correction from Class II to Class I molar relationships were excluded, as were patients whose treatment plans included correction of posterior crossbites. Subjects with systemic diseases or syndromes, a history or current use of bisphosphonates, current use of prostaglandin inhibitors, generalized moderate to severe periodontitis, or active oral hard or soft tissue lesions were excluded from the study.

Intervention

After the initial evaluation of the ClinCheck (Align Technology, Inc., San Jose, CA) treatment simulation by the treating orthodontist, the subjects were invited to participate in the study. They were told that the study was investigating the efficiency of a new vibrational device designed to accelerate tooth movement, and that they would be randomly assigned to 1 of 2 experimental devices, A or B, which differed in the magnitude or the frequency of vibration. As compensation for enrolling in the study, they received an experimental AcceleDent device during the study; after the study ended, they would receive a commercially available device at no charge to be used with any refinement sets of aligners. Subjects in both arms were instructed to use the device daily as recommended by the company (20 minutes per day) during their orthodontic treatment with aligners. Device A was the same as the commercially available AcceleDent Aura device. Device B was also the same as the commercially available AcceleDent Aura device, but the coupler that transmitted the vibration to the mouthpiece was missing. Thus, the motors on the B devices could be turned on and heard, but no vibrational force was delivered to the patient’s teeth.

The aligners were fabricated using the usual approach and prescription of the treating orthodontist, with respect to the sequencing of treatment, use of attachments or other treatment features, use of interproximal reduction, and so on. Tooth movement was limited to no more than 0.25 mm per aligner. If the treatment in 1 arch was projected to be completed before that of the opposing arch, Align Technology was instructed to “spread out” the tooth movement in the arch with fewer aligners so that it matched the number of aligners in the other arch. Thus, the rate of programmed tooth movement in the “spread-out” aligners was decreased. All subjects were placed on a 1-week regimen for changing aligners. They were given 3 aligners initially and then seen at 3-week intervals to check the fit of the aligners and to receive 3 more aligners.

At each visit, if the last aligner worn did not display an adequate fit as described below, the subject was removed from the study and was asked to change to a conventional 2-week regimen for the remainder of treatment. The criteria for lack of adequate fit were (1) spacing greater than 1.5 mm between the aligner and the occlusal or incisal surface of 2 or more teeth, (2) bubbles capturing less than 50% of 2 or more attachments, (3) aligner not fitting tightly over the gingival area of 2 or more teeth, and (4) any other gross aligner lack of fit that was not caused by distortion or damage to the aligner. An intraoral photograph was taken to document the lack of fit at the evaluation appointment ( Fig 1 ).

Bubbles capturing less than 50% of attachments and spacing greater than 1.5 mm between the aligners and the maxillary left premolar and molar.

After the first series of aligners, some subjects had to undergo additional sets of aligners as deemed necessary by the treating orthodontist. These stages of treatment were not included in the trial. The subjects were given the option to use a commercially available AcceleDent device during the additional sets of aligners. To ensure blinding throughout the entire study and to prevent the subjects from comparing devices, the commercial devices were not dispensed to these subjects until the experimental devices were returned to the study coordinator.

To document the clinical outcome, digital scans before the start of the treatment and immediately after treatment were taken using an iTero intraoral digital scanner (Align Technology).

For subjects who left the study and were asked to switch to 2-week intervals, an additional scan was obtained before switching to the 2-week interval regimen. A scan of both arches was obtained when both arches completed the full series of aligners.

Compliance was tracked in 3 ways. First, subjects were asked to report on 1 week of device usage and aligner wear using questionnaires at 3 time points: baseline, midpoint (determined by the orthodontists based on total number of aligners), and completion of treatment. Second, Invisalign aligners were embedded with compliance indicators used in Invisalign Teen to track usage. Although this monitor does not record the usage precisely, it can provide an estimate of wear time. Third, the AcceleDent device was used to track the amount of use each day. This information was downloaded and evaluated from each device at each visit and at the end of treatment.

At various time points in the aligner regimen, we distributed questionnaires to the subjects to inquire about pain and QoL during treatment. The survey contained a numeric rating scale for pain assessment, because this method has been shown to be superior to other methods of pain assessment.

Pain was assessed twice for all subjects during the study. They were asked to complete a 1-week pain survey series during the initial and middle aligner stages. The questionnaires were given to the subjects to take home and fill out daily during those 2 weeks. The 1-week duration was chosen because previous studies have shown that most pain and discomfort returns to baseline 7 days after aligner delivery. Collecting longitudinal data allowed us to detect how pain changed during the immediate postdelivery period of a new series of aligners. In addition, collecting data at various time intervals helped us to account for subject adaptation or response shifts after becoming more familiar with the orthodontic experience.

Subjects were asked to refrain from taking over-the-counter pain medication (eg, acetaminophen, ibuprofen, naproxen) if possible, to prevent clouding of the pain experience. If subjects did take analgesic medications, they were asked to report this in their questionnaires.

The QoL questionnaire was modified from a well-validated oral health assessment index (short form of the Oral Health Impact Profile) to reflect impacts relevant to orthodontic treatment. QoL was assessed at 3 time points during the study. The subjects completed a baseline QoL assessment survey immediately after distribution of their first aligner. This survey was completed again at the midpoint of treatment and at the study completion visit.

Outcomes (primary and secondary) and any changes after trial commencement

As the main outcome, we compared the percentages of subjects who successfully completed their initial series of aligners as planned, using a 1-week change interval. They were given 3 aligners at a time and were seen at 3-week intervals. At each visit, an assessment was made to determine whether the current aligner had an adequate fit; if so, the patient would continue with the 1-week regimen.

The secondary outcome was the alignment of anterior teeth at the end of treatment for subjects who were able to complete their series of aligners with the 1-week change regimen. This was done by measuring the maxillary and mandibular incisor irregularity scores at the beginning and end of treatment. The end of treatment irregularity index and the reduction in the irregularity index were compared in groups A vs B. To measure the incisor irregularity index, digital models were viewed in OrthoCAD (Cadent, Carlstadt, NJ), oriented in the occlusal view with a 1-mm grid overlaid, and magnified (about 10 times) with the zoom function of OrthoCAD. The exact magnification was measured for each model. The absolute horizontal distances between each anatomic contact point were measured, and then the distance was scaled appropriately by using the overlaid 1-mm grid. There was 1 patient in group A and 3 patients in group B with missing anterior teeth. In those patients, the average irregularity index was calculated by adjusting for the number of contacts.

Compliance in groups A and B (active and control arms) was monitored using 3 methods: (1) self-reported data from questionnaires, (2) objective wear time of the aligners by the blue-dot indicators, and (3) AcceleDent use time obtained from the device.

Additionally, for all participants of the study, we collected and analyzed data from the numeric rating scale for pain and QoL questionnaires.

Sample size calculation

To justify the use and cost of the vibration device, we thought that a considerably larger percentage of patients should be able to complete treatment with a 1-week regimen when an active vibration device was used. Thus, the values we selected for our power calculations were a 99% completion rate in patients with an active device, compared with a 50% completion rate with the control device. These estimates indicated that 12 subjects were required in each arm. Additionally, we estimated that a 1.5-mm difference in the incisor irregularity index would be clinically meaningful between the patients in the active and control arms. With an estimated sample of 12 subjects in each arm and a standard deviation of 1, our power was calculated to be 0.91 for this outcome.

Randomization (random number generation, allocation concealment, implementation)

A block stratified randomization scheme was used in this study. Stratification was achieved by generating a separate block for each combination of age (older than 45 vs younger than 45 years) and sex covariates. Subjects were assigned to the appropriate block of covariates using a random number sequence in separate blocks of 2 to ensure that equal numbers of subjects would receive the active and control devices.

The study statistician (C.S.) created the randomization list by using R software (version 3.1.1; R Foundation for Statistical Computing, Vienna, Austria).

A randomization chart was kept at the University of Washington, remote from the clinical sites, at all times. At the time of enrollment, the orthodontist contacted the study coordinator by phone or text to receive the assignment based on age and sex.

Blinding

The devices were designated as A and B when they were shipped from the factory. Thus, this was a triple-blinded study, in which the subjects, investigators, and assessors were all blinded to the treatment arms.

The subjects were told that we were investigating the efficiency of a new vibrational device designed to accelerate tooth movement. They were told that they would be randomized to receive devices that differed in frequency or amplitude. In reality, the difference was that the control devices had an amplitude of zero. We used this deception to guard against poor compliance if a patient suspected that he or she had received an inactive device. The institutional review board was consulted regarding this issue before the study and the wording that was used.

A blinded examiner (M.K.) assessed all digital casts and abstracted data from the questionnaires. Likewise, our statistician was blinded to the study arms throughout the statistical analyses.

Statistical analysis (primary and secondary outcomes, subgroup analyses)

Descriptive statistics, including means and standard deviations, were used to compare the demographic characteristics of the subjects in each arm. Investigation of the percentage of subjects who successfully completed the initial 1-week treatment regimen between the 2 arms was carried out using the Fisher exact test, and associated 95% confidence intervals were calculated using exact binomial confidence intervals.

For the differences between the groups’ initial irregularity index scores, numbers of aligners, final irregularity index scores, and changes in irregularity index scores, P values and 95% confidence intervals were calculated using independent-samples t tests.

To assess the reliability of measurements, the final irregularity index values of the completers were measured again after 2 weeks. The Dahlberg measurement error calculations showed a 0.1-mm difference between the original and repeated measurements.

Compliance data downloaded from the devices were analyzed. The difference between the 2 arms of the study and the difference between completers and noncompleters of the 1-week aligner regimen were measured using the Wilcoxon rank sum test.

Self-reported aligner compliance data, pain level on the numeric rating scale and QoL data were gathered from the questionnaires and analyzed. We used the Wilcoxon rank sum test to measure the differences between the arms of the study.

Trial design and changes after trial commencement

This study was a multicenter, 2-arm, parallel, randomized, triple-blinded, active-controlled clinical trial with a 1:1 allocation ratio. Institutional Review Board approval was obtained from the University of Washington Human Subjects Division before the study on May 15, 2015 (unique protocol number 49073-D). This study was registered at ClinicalTrials.gov with the original release on May 4, 2015 ( NCT02438280 ).

This study was originally designed to be conducted at 1 center in Seattle, Wash. In December 2015, with the approval of the University of Washington Human Subjects Division, an additional subject recruitment site in Vancouver, British Columbia, Canada, was added. This change was immediately reflected on ClinicalTrials.gov . No other changes or modifications were made to the original study protocol and methodology of the research.

The study was designed to evaluate the effects of a supplemental vibrational device (30 Hz and 0.25 N) when used in conjunction with clear aligners in adult subjects.

Participants, eligibility criteria, and setting

Orthodontic patients who were beginning their Invisalign (Align Technology, San Jose, Calif) treatment in 1 of the 2 approved private practices in Seattle or Vancouver were invited to participate in this study. Adults (18 years or older) with malocclusions whose treatment involved fewer than 25 sets of aligners were invited to participate. Also, no significant anteroposterior or transverse movements were planned. For example, patients whose treatment plans included correction from Class II to Class I molar relationships were excluded, as were patients whose treatment plans included correction of posterior crossbites. Subjects with systemic diseases or syndromes, a history or current use of bisphosphonates, current use of prostaglandin inhibitors, generalized moderate to severe periodontitis, or active oral hard or soft tissue lesions were excluded from the study.

Intervention

After the initial evaluation of the ClinCheck (Align Technology, Inc., San Jose, CA) treatment simulation by the treating orthodontist, the subjects were invited to participate in the study. They were told that the study was investigating the efficiency of a new vibrational device designed to accelerate tooth movement, and that they would be randomly assigned to 1 of 2 experimental devices, A or B, which differed in the magnitude or the frequency of vibration. As compensation for enrolling in the study, they received an experimental AcceleDent device during the study; after the study ended, they would receive a commercially available device at no charge to be used with any refinement sets of aligners. Subjects in both arms were instructed to use the device daily as recommended by the company (20 minutes per day) during their orthodontic treatment with aligners. Device A was the same as the commercially available AcceleDent Aura device. Device B was also the same as the commercially available AcceleDent Aura device, but the coupler that transmitted the vibration to the mouthpiece was missing. Thus, the motors on the B devices could be turned on and heard, but no vibrational force was delivered to the patient’s teeth.

The aligners were fabricated using the usual approach and prescription of the treating orthodontist, with respect to the sequencing of treatment, use of attachments or other treatment features, use of interproximal reduction, and so on. Tooth movement was limited to no more than 0.25 mm per aligner. If the treatment in 1 arch was projected to be completed before that of the opposing arch, Align Technology was instructed to “spread out” the tooth movement in the arch with fewer aligners so that it matched the number of aligners in the other arch. Thus, the rate of programmed tooth movement in the “spread-out” aligners was decreased. All subjects were placed on a 1-week regimen for changing aligners. They were given 3 aligners initially and then seen at 3-week intervals to check the fit of the aligners and to receive 3 more aligners.

At each visit, if the last aligner worn did not display an adequate fit as described below, the subject was removed from the study and was asked to change to a conventional 2-week regimen for the remainder of treatment. The criteria for lack of adequate fit were (1) spacing greater than 1.5 mm between the aligner and the occlusal or incisal surface of 2 or more teeth, (2) bubbles capturing less than 50% of 2 or more attachments, (3) aligner not fitting tightly over the gingival area of 2 or more teeth, and (4) any other gross aligner lack of fit that was not caused by distortion or damage to the aligner. An intraoral photograph was taken to document the lack of fit at the evaluation appointment ( Fig 1 ).

Bubbles capturing less than 50% of attachments and spacing greater than 1.5 mm between the aligners and the maxillary left premolar and molar.

After the first series of aligners, some subjects had to undergo additional sets of aligners as deemed necessary by the treating orthodontist. These stages of treatment were not included in the trial. The subjects were given the option to use a commercially available AcceleDent device during the additional sets of aligners. To ensure blinding throughout the entire study and to prevent the subjects from comparing devices, the commercial devices were not dispensed to these subjects until the experimental devices were returned to the study coordinator.

To document the clinical outcome, digital scans before the start of the treatment and immediately after treatment were taken using an iTero intraoral digital scanner (Align Technology).

For subjects who left the study and were asked to switch to 2-week intervals, an additional scan was obtained before switching to the 2-week interval regimen. A scan of both arches was obtained when both arches completed the full series of aligners.

Compliance was tracked in 3 ways. First, subjects were asked to report on 1 week of device usage and aligner wear using questionnaires at 3 time points: baseline, midpoint (determined by the orthodontists based on total number of aligners), and completion of treatment. Second, Invisalign aligners were embedded with compliance indicators used in Invisalign Teen to track usage. Although this monitor does not record the usage precisely, it can provide an estimate of wear time. Third, the AcceleDent device was used to track the amount of use each day. This information was downloaded and evaluated from each device at each visit and at the end of treatment.

At various time points in the aligner regimen, we distributed questionnaires to the subjects to inquire about pain and QoL during treatment. The survey contained a numeric rating scale for pain assessment, because this method has been shown to be superior to other methods of pain assessment.

Pain was assessed twice for all subjects during the study. They were asked to complete a 1-week pain survey series during the initial and middle aligner stages. The questionnaires were given to the subjects to take home and fill out daily during those 2 weeks. The 1-week duration was chosen because previous studies have shown that most pain and discomfort returns to baseline 7 days after aligner delivery. Collecting longitudinal data allowed us to detect how pain changed during the immediate postdelivery period of a new series of aligners. In addition, collecting data at various time intervals helped us to account for subject adaptation or response shifts after becoming more familiar with the orthodontic experience.

Subjects were asked to refrain from taking over-the-counter pain medication (eg, acetaminophen, ibuprofen, naproxen) if possible, to prevent clouding of the pain experience. If subjects did take analgesic medications, they were asked to report this in their questionnaires.

The QoL questionnaire was modified from a well-validated oral health assessment index (short form of the Oral Health Impact Profile) to reflect impacts relevant to orthodontic treatment. QoL was assessed at 3 time points during the study. The subjects completed a baseline QoL assessment survey immediately after distribution of their first aligner. This survey was completed again at the midpoint of treatment and at the study completion visit.

Outcomes (primary and secondary) and any changes after trial commencement

As the main outcome, we compared the percentages of subjects who successfully completed their initial series of aligners as planned, using a 1-week change interval. They were given 3 aligners at a time and were seen at 3-week intervals. At each visit, an assessment was made to determine whether the current aligner had an adequate fit; if so, the patient would continue with the 1-week regimen.

The secondary outcome was the alignment of anterior teeth at the end of treatment for subjects who were able to complete their series of aligners with the 1-week change regimen. This was done by measuring the maxillary and mandibular incisor irregularity scores at the beginning and end of treatment. The end of treatment irregularity index and the reduction in the irregularity index were compared in groups A vs B. To measure the incisor irregularity index, digital models were viewed in OrthoCAD (Cadent, Carlstadt, NJ), oriented in the occlusal view with a 1-mm grid overlaid, and magnified (about 10 times) with the zoom function of OrthoCAD. The exact magnification was measured for each model. The absolute horizontal distances between each anatomic contact point were measured, and then the distance was scaled appropriately by using the overlaid 1-mm grid. There was 1 patient in group A and 3 patients in group B with missing anterior teeth. In those patients, the average irregularity index was calculated by adjusting for the number of contacts.

Compliance in groups A and B (active and control arms) was monitored using 3 methods: (1) self-reported data from questionnaires, (2) objective wear time of the aligners by the blue-dot indicators, and (3) AcceleDent use time obtained from the device.

Additionally, for all participants of the study, we collected and analyzed data from the numeric rating scale for pain and QoL questionnaires.

Sample size calculation

To justify the use and cost of the vibration device, we thought that a considerably larger percentage of patients should be able to complete treatment with a 1-week regimen when an active vibration device was used. Thus, the values we selected for our power calculations were a 99% completion rate in patients with an active device, compared with a 50% completion rate with the control device. These estimates indicated that 12 subjects were required in each arm. Additionally, we estimated that a 1.5-mm difference in the incisor irregularity index would be clinically meaningful between the patients in the active and control arms. With an estimated sample of 12 subjects in each arm and a standard deviation of 1, our power was calculated to be 0.91 for this outcome.

Randomization (random number generation, allocation concealment, implementation)

A block stratified randomization scheme was used in this study. Stratification was achieved by generating a separate block for each combination of age (older than 45 vs younger than 45 years) and sex covariates. Subjects were assigned to the appropriate block of covariates using a random number sequence in separate blocks of 2 to ensure that equal numbers of subjects would receive the active and control devices.

The study statistician (C.S.) created the randomization list by using R software (version 3.1.1; R Foundation for Statistical Computing, Vienna, Austria).

A randomization chart was kept at the University of Washington, remote from the clinical sites, at all times. At the time of enrollment, the orthodontist contacted the study coordinator by phone or text to receive the assignment based on age and sex.

Blinding

The devices were designated as A and B when they were shipped from the factory. Thus, this was a triple-blinded study, in which the subjects, investigators, and assessors were all blinded to the treatment arms.

The subjects were told that we were investigating the efficiency of a new vibrational device designed to accelerate tooth movement. They were told that they would be randomized to receive devices that differed in frequency or amplitude. In reality, the difference was that the control devices had an amplitude of zero. We used this deception to guard against poor compliance if a patient suspected that he or she had received an inactive device. The institutional review board was consulted regarding this issue before the study and the wording that was used.

A blinded examiner (M.K.) assessed all digital casts and abstracted data from the questionnaires. Likewise, our statistician was blinded to the study arms throughout the statistical analyses.

Statistical analysis (primary and secondary outcomes, subgroup analyses)

Descriptive statistics, including means and standard deviations, were used to compare the demographic characteristics of the subjects in each arm. Investigation of the percentage of subjects who successfully completed the initial 1-week treatment regimen between the 2 arms was carried out using the Fisher exact test, and associated 95% confidence intervals were calculated using exact binomial confidence intervals.

For the differences between the groups’ initial irregularity index scores, numbers of aligners, final irregularity index scores, and changes in irregularity index scores, P values and 95% confidence intervals were calculated using independent-samples t tests.

To assess the reliability of measurements, the final irregularity index values of the completers were measured again after 2 weeks. The Dahlberg measurement error calculations showed a 0.1-mm difference between the original and repeated measurements.

Compliance data downloaded from the devices were analyzed. The difference between the 2 arms of the study and the difference between completers and noncompleters of the 1-week aligner regimen were measured using the Wilcoxon rank sum test.

Self-reported aligner compliance data, pain level on the numeric rating scale and QoL data were gathered from the questionnaires and analyzed. We used the Wilcoxon rank sum test to measure the differences between the arms of the study.