Evaluation of reservoirs in bleaching trays for at-home bleaching: a split-mouth single-blind randomized controlled equivalence trial

Abstract Objectives This randomized, split-mouth, single-blinded trial assessed whether the use of reservoirs in at-home bleaching trays is equivalent to non-reservoir trays. Our choice of an equivalence trial was based on the expectation that a non-reservoir tray is sufficient to produce a color change. Secondary outcomes such as tooth sensitivity (TS) and gingival irritation (GI) were also assessed. Methodology Forty-six patients were selected with canines shade A2 or darker. In half of the patient’s arch, bleaching trays were made with reservoirs and the other half, without reservoirs. At-home bleaching was performed with carbamide peroxide (CP) 10% (3 h daily; 21 days). Color change was evaluated with a digital spectrophotometer (ΔE, ΔE00, and Whiteness Index) and shade guide units (ΔSGU) at baseline, during and one-month post-bleaching. TS and GI were assessed with a numeric scale (NRS) and a visual analog scale (VAS). Results After one month, the equivalence of reservoir and non-reservoir groups were observed in all color instruments (p>0.05). Fifteen and sixteen patients presented pain (absolute risk: 33% and 35%, 95%, confidence interval (CI) 21-46% and 23-49%) in the reservoir and non-reservoir side, respectively. The odds ratio for pain was 0.8 (95%CI 0.2-3.0) and the p-value was non-significant (p=1.0). TS intensity was similar between both groups in any of the pain scales (p>0.05). No difference in the GI was observed (p>0.05). Conclusions The protocol with reservoirs is equivalent in color change to the non-reservoir, although no superiority of the latter was observed in terms of reduced TS and GI with at-home 10% carbamide peroxide bleaching. Clinical Relevance The presence of reservoirs in a bleaching tray did not improve color change or affect tooth sensitivity and gingival irritation.


Introduction
Dental bleaching is widely used to make teeth whiter and brighter, a common desire among patients. 1-3 The dentist-supervised dental bleaching technique can be performed using high-concentrate materials (in-office protocol) or by dispensing low concentrate-material in a custom bleaching tray (at-home use).
Among these available protocols, clinicians consider at-home bleaching safer as it employs low concentrate products 2 and therefore reduces the risk and intensity of tooth sensitivity. 4,5 Additionally, it is an easy protocol, requires reduced chair-time, and it is cheaper than the in-office protocol.
Since the introduction of at-home bleaching, several modifications of the protocol and materials occurred in the past years. Carbamide peroxide or hydrogen peroxide with varied concentrations can now be employed. 6 The daily usage time of the bleaching tray was reduced 7,8 and modifications in the manufacture of the bleaching trays 9 were proposed with the presence of reservoirs. 2,10,11 Reservoirs are modifications in the tray molds to increase the amount of bleaching material carried by the bleaching tray, seeking greater bleaching efficacy.
Fisher first introduced the use of tray reservoirs in 1992. 12 For such purpose, light-cured block-out resin or light-curing composites are applied on the buccal surface of teeth from the cast models to create an additional space between the tray and the teeth. The first report on the efficacy of reservoirs in bleaching trays come from the end of the 1990s. 11 This as well as other clinical studies 11,13,14 contested the efficacy of tray reservoirs in bleaching trays.
In a recent systematic review, the authors concluded that the majority of the studies that compared non-reservoir and reservoir were at unclear risk of bias, indicating the need for welldesigned clinical trials. 15 Some important aspects of well-designed clinical studies such as randomization, allocation concealment, and blinding were missing in the eligible studies 10, 11,13,14,[16][17][18] and the studies lacked standardized methods for reporting important outcomes, such as color change, tooth sensitivity, and gingival irritation. This prevented the authors from this systematic review 15 to conclude on the study's findings but to ask for the conduction of additional clinical trials that answer the same research question, since this modification related to the bleaching protocol, despite having the first reports published a long time ago, still generates controversies about its real effectiveness.
Therefore, the objective of this study was to conduct a randomized controlled equivalence trial with a split-mouth design to test that non-reservoir trays are 'as effective as' reservoir trays in terms of color change. The secondary outcomes risk of tooth sensitivity, intensity of tooth sensitivity, and gingival irritation were compared in a traditional superiority hypothesis testing.

Blinding
This study was a randomized, single-blind controlled trial, in which the evaluator was blinded to the group assignment. A researcher not involved in the implementation and evaluation process was responsible for the delivery and guidance on the administration of the bleaching trays.

Intervention
Two dentists, with more than 5 years of clinical experience (E.M. and F.M.C.) performed the bleaching procedure. They made alginate impressions of each participant's jaw, and after disinfection, filled them with dental stone. The upper arch models were used in the study. In one of the sides, a photopolymerized blocking material (TopDam, FGM, Joinville, SC, Brazil) was applied in the buccal surfaces of the central, lateral, canine and premolar teeth in one side of the patient's mouth to create reservoirs on these teeth.
The blocking resin was applied so that the labial surface was covered except for 1 mm in the mesial, We instructed all participants to use the bleaching tray with the bleaching agent (10% carbamide peroxide with potassium nitrate and fluoride, Opalescence PF, Ultradent Products) for 3 hours once a day for 21 days.
They were instructed to place an amount of gel to cover the buccal surface of all teeth (this amount being slightly higher in the reservoir-side of the bleaching tray). Participants were instructed to remove the tray after each bleaching period, rinse teeth with water and brush their teeth as usual.
As a measure of adherence to the experimental protocol study, participants received a diary in which they were asked to take note of the number of hours a day they used the tray during treatment. They were reminded of this procedure daily using the social media network group set up at the beginning of the study protocol.

Color evaluation
For the evaluation of this primary outcome, two experienced and calibrated dentists (kappa statistics higher than 80% after previous calibration), who were not involved in the randomization procedures, We performed the calibration procedure using 20 volunteers. The operators color-checked the canines independently, using shade guides, and when differences were noted, they had to reach an agreement. This procedure was repeated until they get a kappa equal to or higher than 80% in two consecutive measurements.
We performed the color evaluation using the shade guide VITA Classical and the VITA Bleachedguide In case of disagreement between operators, they were required to reach consensus.

Tooth sensitivity evaluation
We instructed patients to fill in a form to record daily dental sensitivity after bleaching. Patients were instructed in detail on how to perform this procedure.
These forms returned to the investigator at every clinical appointment.
For the 4-point numeric rating scale (NRS), we asked the patient to indicate the numeric value of the degree of sensitivity (from 0 to 4) for each of the periods, in which zero means no sensitivity, 1 means mild, 2 means moderate, 3 means considerable, and 4 means severe tooth sensitivity. In addition, the participants were also instructed to record pain intensity using the visual analog scale (VAS). This scale is a 10-centimeter horizontal line with scores of zero and ten at each end, in which 0 means no sensitivity, and 10 means severe tooth sensitivity.
The patient was required to mark with a vertical line across the horizontal line of the scale the intensity of the tooth sensitivity. Then, the distance in millimeters from the zero end was measured with the aid of a millimeter ruler.
We merged the daily data from the three weeks of bleaching for statistical purposes. For this purpose, the worst score (NRS scale) and the highest numerical value (VAS scale) from the total period were taken to represent the patient's sensitivity level throughout the study. If the participant scored zero (no sensitivity) in all time-assessments, this participant was considered to be insensitive to the bleaching protocol. In all other circumstances, the participants were believed to have bleaching-induced tooth sensitivity.

Gingival irritation evaluation
Participants were instructed to fill out a form to record the daily GI after bleaching. These forms returned to the researcher at the next appointment (with one-week intervals), during the three weeks of treatment. For the GI questionnaire, the participant was asked to indicate if they felt any discomfort in the gingiva and if there was discomfort, the side should be indicated.
As with tooth sensitivity assessments, when the participants reported no gingival irritation at all threeweek bleaching time evaluations, they were considered insensitive to the bleaching protocol. In all other circumstances, participants were considered to have GI induced by bleaching.
Evaluation of reservoirs in bleaching trays for at-home bleaching: a split-mouth single-blind randomized controlled equivalence trial J Appl Oral Sci. 2020;28:e20200332 5/11

Statistical analysis
All participants received the intended protocol and had their outcomes measured, meaning that the intention-to-treat protocol and per-protocol analysis resulted in the same findings ( Figure 1). The statistician was blinded to the groups.
Two one-sided t-tests for paired samples (TOST-P) were used to test the equivalence of the study groups at the different assessment points (baseline vs.

Results
We examined a total of 59 participants according to the inclusion and exclusion criteria, but only 46 participants remained for the clinical trial ( Figure   1). The baseline color of the participants was 9.8 from Vita Classical scale and ∆SGU from Vita Bleached guide. The two-sided 90% CI of the difference of the means are within the predetermined equivalence margins of -3 and +3 for ∆E, ∆E00 and ∆Wi, and -1 and +1 for ∆SGU.
The traditional one-way repeated ANOVA measures detected significant whitening over time (Table 1; p>0.05). After one month, a whitening effect of near eight units in the Vita Classical scale and 11 units in the Vita Bleached guide were observed. Regarding to spectrophotometers measurements, we observed a color change of 13 units for ∆E, 9 units for ΔE00 and 22 units for ΔWi (Table 1).

Risk of tooth sensitivity
A total of fifteen patients reported pain in the experimental arch side (absolute risk: 33%, 95% CI 21 to 46%), and from these, 5 patients did not report pain in the control side. Sixteen patients (absolute risk: 35%, 95% CI 23 to 49%) reported pain in the control group, and from these, 6 did not experience pain in the experimental side. In relative terms, the odds ratio for pain was 0.8 (0.2 to 3.0; Table 2), and

Risk of gingival irritation
A total of 16 patients reported pain in the experimental arch side (absolute risk: 35%, 95% CI 23 to 49%), and from these, 4 patients did not report pain in the control side. Seventeen patients reported pain in the control group (absolute risk: 37%, 95% CI 25 to 51%), and from these, 5 did not experience pain in the experimental side. In relative terms, the odds ratio for pain was 0. 8 (0.16 to 3.7;

Intensity of tooth sensitivity
The statistical analysis did not show any significant difference in the TS intensity between groups in any of the pain scales (p=0.64 for NRS scale, and p=0.23; for VAS scale; Table 6). The mean difference of pain intensity in VAS scale was on average -0.2 units, a difference far from clinically important. Pain was positively correlated in both groups (Table 4).
Correlation was moderate and significant for both pain scales. For NRS scale, the Spearman correlation was 0.52 (p<0.01) and for VAS scale, the Pearson correlation was 0.69 (p<0.01).

Discussion
The use of reservoirs in the bleaching trays was initially seen as positive, since higher accumulation of material could provide the patient with greater treatment efficacy. 11 After the emergence of this new technique, some clinical trials 11,13,14 observed that the efficacy of this treatment was not dependent on reservoirs but rather in the exposure area and gel application time. 10 However, although these studies reported these findings, they were considered at unclear risk of bias in a recent systematic review. 15 Additionally, these earlier studies had low statistical power. Negative results of low-powered studies may not indicate that one group is different from another, but rather that these results may be due to chance alone.
These earlier studies 10, 11,13,14,[16][17][18] also lacked the use of standardized outcomes to report their findings of color change, tooth sensitivity and gingival irritation, which reduced the reliability of the study's findings.
The limitations above of the previous studies on this issue motivated us to conduct this randomized clinical trial.
In the present study, we measured color change by using subjective methods (matching with different shade guide units) along with objective methods (spectrophotometer). It is reported that measurement with a spectrophotometer provides more accurate results than visual shade matching with shade guides 25,26 as it is less prone to subjective judgments; however, results published in ∆E are less clinically tangible. It worth to mentioning that, in this study the CIEDE2000 system and Whiteness Index for Dentistry were also used. 23,24 According to Sharma, Wu and McNemar's test (p=1.0); Spearman correlation between paired data =0.47; p-value=0.0001.     Bleaching-induced TS is directly related to the flow of hydrogen peroxide to the pulp chamber. 36 As the presence of reservoirs offers a higher amount of bleaching gel, it is believed that the tooth sensitivity could be aggravated, as well as gingival irritation. 13 However, the present study showed that there is no significant difference between groups whitened with and without reservoirs for gingival irritation or tooth sensitivity. Approximately 30% of the patients reported tooth sensitivity with a very low intensity (about 1.6 unit in the VAS scale), and no type of additional desensitization was required.
The similar risk of TS between both groups reinforces the fact that the amount of material placed on the enamel surface does not affect the bleaching outcome. The penetration of the bleaching agent is not driven by the mass (amount) of product placed on the surface but by the diffusion coefficient of the bleaching product itself on the dental substrate. 10,33 This diffusion coefficient is dependent on the nature of the substance under diffusion and on the area of application. Factors such as viscosity and solution properties (concentration, pH and, temperature) which were not altered between groups can affect diffusion, but not the product mass. 33,34 Therefore, one can expect that the amount of hydrogen peroxide that achieved the pulp chamber was similar in both sides of the patient's arches, leading to a similar risk of bleaching-induced TS; this may explain the fact that our findings agree with previous studies, which also found no differences in TS between groups with and without reservoirs. 10,11,14,16 For calculation of risk and intensity of TS, we summarized the tooth sensitivity data based on the worst episode of pain in the bleaching period. In the author's opinion, the experience of considerable pain makes the experience negative for patients, even if it is a single episode or multiple episodes. This provides us with the worst scenario; other ways to report the adverse effect of tooth sensitivity, however, do exist.
For instance, in an exploratory analysis we calculated the mean number of days patients experienced tooth sensitivity (reservoir side: 3.9 ± 5.0; non-reservoir side; 4.6 ± 5.9), and also the intensity of TS (VAS scale) by taking the mean of the daily TS during the bleaching period (reservoir: 0.8 ± 1.0; non-reservoir: 0.9 ± 1.0). By using appropriate statistics, we reached up with the conclusion of no significant difference between groups which makes the results of the present investigation robust and not affected by these prior decisions.
Apart from not bringing benefits to the at-home bleaching, designing bleaching trays with reservoirs will require a more significant amount of bleaching material used to whiten teeth, as well as more time for tray fabrication, increasing the costs associated with this procedure. Some companies, such as Ultradent Products still recommend the manufacturing of J Appl Oral Sci. 2020;28:e20200332 10/11 bleaching trays with reservoirs. This recommendation is not based on the findings of controlled trials but probably on the fact that more material is required to fill in the bleaching tray with reservoirs. While this may be an advantage for the company, the same is not true for for clinicians and patients, who will spend more in bleaching material than one would if bleaching was performed with trays without reservoirs.
Unfortunately, clinicians have much more access to the manufacturer's instruction of bleaching products than access to findings of randomized clinical trials or even laboratory studies, which also found no advantages in the presence of reservoirs. 37 Finally, we should mention the limitations of the present study. We have just evaluated one brand of material in this clinical trial. Although this may be seen as a limitation, bleaching agents have very similar composition which contrasts with the majority of the dental materials used in the daily practice. Researchers should conduct further clinical trials using different brand of materials. As the majority of the participants were young in this clinical trial, results should not be generalizable to older populations without care.

Conclusion
The presence or absence of reservoirs in a bleaching tray did not affect color change, tooth sensitivity, or gingival irritation in a dentist-supervised, at-home bleaching performed with 10% carbamide peroxide gel.

Funding
The work was supported by the research funding Capes agency.

Compliance with ethical standards
The clinical investigation was approved (2.124.508) by the scientific review committee and by the committee for the protection of human participants of the State University of Ponta Grossa and was conducted in accordance with the protocol established by the Consolidated Standards of Reporting Trials statement with extension for within-person designs