Trial design and any changes after trial commencement

This was a 2-arm parallel, randomized controlled trial with a balanced allocation ratio (42:38). The study was approved by the ethics committee of West China Stomatology Hospital (project number WCHSIRB-ST-2013-094). No changes to the methods after trial commencement occurred.

Participants, eligibility criteria, and settings

Initially, 110 consecutive patients who had been examined by an orthodontist (W.L.) at the last appointment before debonding were assessed for eligibility at West China Stomatology Hospital of Sichuan University between July 2013 and June 2015; 30 were excluded because of ineligibility (n = 17), or they declined to participate (n = 13). Finally, 80 patients aged between 12 and 33 years were recruited. The inclusion criteria were (1) patients who were healthy without systemic or oral diseases; (2) central incisors, canines, and first molars of both arches were present; and (3) the trial and signed informed consent forms were fully understood, either by themselves or their parents. Patients were excluded if they had oral habits such as bruxism and clenching, did not accept this type of retainers, or were unable to comply with follow-up during the 12-month period. Once the trial started, study methodologies were implemented consistently with no changes.

Interventions

All eligible subjects were randomized to receive either a 0.75-mm thick or a 1.00-mm thick COR. Basic information including name, sex, age, telephone number, address, dates of starting and ending treatment, and brief treatment plan was collected by another certified researcher (J.L.). All retainers were fabricated by a qualified technician in the Department of Orthodontics at Sichuan University. Standardized procedures were carried out in the fabrication process: dental casts were first examined for undesirable undercuts and blocked out when necessary; the distance from the bottom of working casts to the occlusal plane was unified to about 15 mm, the occlusal plane was parallel to the platform on the pressure thermoforming machine (Drufomat Scan; Dreve Dentamid, Unna, Germany), and the gingival margin of the model was placed flush with the edge of the machine, and round metal granules provided by the machine company were used to embed unneeded parts of the model during thermoforming. The buccal edges of the CORs were parallel with and at the same level as the gingival margins, and the lingual edges were parallel with but extended 3 to 4 mm beyond the gingival margins. All occlusal surfaces were tightly covered, and the most distal teeth were included. Two thermoplastic materials with different thicknesses (0.75 and 1.00 mm) were used to fabricate 2 types of CORs (Biolon; Dreve Dentamid).

Participants were instructed to wear the retainers 24 hours a day during the 12-month observation period, except during meals and cleaning. They were told to avoid pigmented and fizzy drinks.

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

The primary outcomes were survival time and total failure rates of the retainers. The secondary outcomes were rates of different types of failure (fracture, loss, nonfitting, and abrasion), and the tertiary results were comfort and quality of life after the wearing CORs.

Survival time (measured in days) was the time from the date the patient began wearing the original retainer to the date when either the maxillary or the mandibular retainer failed. Failure was defined as when a subject could no longer wear the retainers because of 1 of the following reasons: fracture, retainer loss, nonfitting, or serious abrasion. “No longer fitting” indicated that the retainers did not fit the teeth because of deformation or relapse. Retainers with slight cracks that did not affect wearing were not regarded as a failures. The numbers of each type of failure and fracture locations were totaled at the end of the follow-up period. Furthermore, failure events and survival times were separately recorded for adult and adolescent participants.

Patients’ comfort levels were assessed with a visual analog scale (VAS) twice that was used immediately after wearing the retainers and at the 1-month-appointment. The SF-36 scale was used once to evaluate the participants’ quality of life. It contains 36 items, mainly involving 8 dimensions: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. The 8 dimensions were further divided into 2 major areas: physical component summary and mental component summary. The participants were asked to complete this SF-36 scale at the 1-month appointment supervised by the researchers within a certain time (20 minutes).

Each participant was reexamined at 1, 3, 6, and 12 months (T1, T2, T3, and T4) after wearing the retainers. When the fixed orthodontic appliances were removed, the retention appliance was inserted (T0). Patients were told to contact the researcher (Y.Z.) as soon as their retainers were broken or lost, and the researcher arranged an appointment, manufactured a new retainer, and recorded the specific date, the reasons for failure, and the location of the fractured retainer. There were no outcome changes after trial commencement.

Sample size calculation

Because no similar study has been reported for reference, the sample size of our study could not be calculated exactly at the design phase. This sample size (n = 80) was determined by our empirical experience in conducting randomized controlled trials. Our results may serve as a reference for future studies in calculating sample sizes.

Interim analyses and stopping guidelines

No interim analyses were performed, and the study would be stopped if serious harm was observed during the study.

Randomization (random number generation, allocation concealment, implementation)

Randomization was accomplished by tossing a coin. Then, the allocation information (randomization results) was concealed in opaque and sealed envelopes. Baseline information was written on the outside of the envelopes. The practice manager (H.L.) was responsible for implementing the randomization process and allocation concealment, but was not involved in other parts of the trial.

Blinding

Because the researcher throwing the coins did not participate in the enrollment procedure and the 2 CORs were apparent, the participants and the researcher were both blinded after assignment to the interventions. The statistical analyst (J.F.) did not directly participate in our trial to ensure blinding. Therefore, blinding of patient, operator, and analyst were all achieved.

Statistical analysis (primary and secondary outcomes, subgroup analyses)

Statistical analysis was performed using SPSS software (version 19.0; IBM, Armonk, NY). The failure rates between the 2 groups were analyzed using the chi-square test. Moreover, the Fisher exact test was used to evaluate the constituent ratios of failure reasons and fracture locations. The survival times of the retainers were estimated using the Kaplan-Meier analysis, and the log-rank test was used to identify the significant differences among the groups. Hazard ratios (HR) and associated 95% confidence intervals (CIs) were calculated using Cox regression modeling. Analysis of variance (ANOVA) and the Student t test were used to compare the SF-36 and VAS scores between the 2 groups, respectively. A P value of 0.05 or less was considered to be statistically significant.

Trial design and any changes after trial commencement

This was a 2-arm parallel, randomized controlled trial with a balanced allocation ratio (42:38). The study was approved by the ethics committee of West China Stomatology Hospital (project number WCHSIRB-ST-2013-094). No changes to the methods after trial commencement occurred.

Participants, eligibility criteria, and settings

Initially, 110 consecutive patients who had been examined by an orthodontist (W.L.) at the last appointment before debonding were assessed for eligibility at West China Stomatology Hospital of Sichuan University between July 2013 and June 2015; 30 were excluded because of ineligibility (n = 17), or they declined to participate (n = 13). Finally, 80 patients aged between 12 and 33 years were recruited. The inclusion criteria were (1) patients who were healthy without systemic or oral diseases; (2) central incisors, canines, and first molars of both arches were present; and (3) the trial and signed informed consent forms were fully understood, either by themselves or their parents. Patients were excluded if they had oral habits such as bruxism and clenching, did not accept this type of retainers, or were unable to comply with follow-up during the 12-month period. Once the trial started, study methodologies were implemented consistently with no changes.

Interventions

All eligible subjects were randomized to receive either a 0.75-mm thick or a 1.00-mm thick COR. Basic information including name, sex, age, telephone number, address, dates of starting and ending treatment, and brief treatment plan was collected by another certified researcher (J.L.). All retainers were fabricated by a qualified technician in the Department of Orthodontics at Sichuan University. Standardized procedures were carried out in the fabrication process: dental casts were first examined for undesirable undercuts and blocked out when necessary; the distance from the bottom of working casts to the occlusal plane was unified to about 15 mm, the occlusal plane was parallel to the platform on the pressure thermoforming machine (Drufomat Scan; Dreve Dentamid, Unna, Germany), and the gingival margin of the model was placed flush with the edge of the machine, and round metal granules provided by the machine company were used to embed unneeded parts of the model during thermoforming. The buccal edges of the CORs were parallel with and at the same level as the gingival margins, and the lingual edges were parallel with but extended 3 to 4 mm beyond the gingival margins. All occlusal surfaces were tightly covered, and the most distal teeth were included. Two thermoplastic materials with different thicknesses (0.75 and 1.00 mm) were used to fabricate 2 types of CORs (Biolon; Dreve Dentamid).

Participants were instructed to wear the retainers 24 hours a day during the 12-month observation period, except during meals and cleaning. They were told to avoid pigmented and fizzy drinks.

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

The primary outcomes were survival time and total failure rates of the retainers. The secondary outcomes were rates of different types of failure (fracture, loss, nonfitting, and abrasion), and the tertiary results were comfort and quality of life after the wearing CORs.

Survival time (measured in days) was the time from the date the patient began wearing the original retainer to the date when either the maxillary or the mandibular retainer failed. Failure was defined as when a subject could no longer wear the retainers because of 1 of the following reasons: fracture, retainer loss, nonfitting, or serious abrasion. “No longer fitting” indicated that the retainers did not fit the teeth because of deformation or relapse. Retainers with slight cracks that did not affect wearing were not regarded as a failures. The numbers of each type of failure and fracture locations were totaled at the end of the follow-up period. Furthermore, failure events and survival times were separately recorded for adult and adolescent participants.

Patients’ comfort levels were assessed with a visual analog scale (VAS) twice that was used immediately after wearing the retainers and at the 1-month-appointment. The SF-36 scale was used once to evaluate the participants’ quality of life. It contains 36 items, mainly involving 8 dimensions: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. The 8 dimensions were further divided into 2 major areas: physical component summary and mental component summary. The participants were asked to complete this SF-36 scale at the 1-month appointment supervised by the researchers within a certain time (20 minutes).

Each participant was reexamined at 1, 3, 6, and 12 months (T1, T2, T3, and T4) after wearing the retainers. When the fixed orthodontic appliances were removed, the retention appliance was inserted (T0). Patients were told to contact the researcher (Y.Z.) as soon as their retainers were broken or lost, and the researcher arranged an appointment, manufactured a new retainer, and recorded the specific date, the reasons for failure, and the location of the fractured retainer. There were no outcome changes after trial commencement.

Sample size calculation

Because no similar study has been reported for reference, the sample size of our study could not be calculated exactly at the design phase. This sample size (n = 80) was determined by our empirical experience in conducting randomized controlled trials. Our results may serve as a reference for future studies in calculating sample sizes.

Interim analyses and stopping guidelines

No interim analyses were performed, and the study would be stopped if serious harm was observed during the study.

Randomization (random number generation, allocation concealment, implementation)

Randomization was accomplished by tossing a coin. Then, the allocation information (randomization results) was concealed in opaque and sealed envelopes. Baseline information was written on the outside of the envelopes. The practice manager (H.L.) was responsible for implementing the randomization process and allocation concealment, but was not involved in other parts of the trial.

Blinding

Because the researcher throwing the coins did not participate in the enrollment procedure and the 2 CORs were apparent, the participants and the researcher were both blinded after assignment to the interventions. The statistical analyst (J.F.) did not directly participate in our trial to ensure blinding. Therefore, blinding of patient, operator, and analyst were all achieved.

Statistical analysis (primary and secondary outcomes, subgroup analyses)

Statistical analysis was performed using SPSS software (version 19.0; IBM, Armonk, NY). The failure rates between the 2 groups were analyzed using the chi-square test. Moreover, the Fisher exact test was used to evaluate the constituent ratios of failure reasons and fracture locations. The survival times of the retainers were estimated using the Kaplan-Meier analysis, and the log-rank test was used to identify the significant differences among the groups. Hazard ratios (HR) and associated 95% confidence intervals (CIs) were calculated using Cox regression modeling. Analysis of variance (ANOVA) and the Student t test were used to compare the SF-36 and VAS scores between the 2 groups, respectively. A P value of 0.05 or less was considered to be statistically significant.