How effective is a powered toothbrush as compared to a manual toothbrush? A systematic review and meta‐analysis of single brushing exercises

Abstract Objectives In adult participants, what is, following a single brushing exercise, the efficacy of a powered toothbrush (PTB) as compared to a manual toothbrush (MTB) on plaque removal? Methods MEDLINE‐PubMed and Cochrane‐CENTRAL were searched from inception to February 2019. The inclusion criteria were (randomized) controlled clinical trials conducted in human subjects ≥18 years of age, in good general health and without periodontitis, orthodontic treatment, implants and/or removable prosthesis. Papers evaluating a PTB compared with a MTB in a single brushing exercise were included. When plaque scores were assessed according to the Quigley‐Hein plaque index (Q&HPI) or the Rustogi modified Navy plaque index (RMNPI). From the eligible studies, data were extracted. A meta‐analysis and subanalysis for brands and mode of action being oscillating‐rotating (OR) and side‐to‐side (SS) were performed when feasible. Results Independent screening of 3450 unique papers resulted in 17 eligible publications presenting 36 comparisons. In total, 28 comparisons assessed toothbrushing efficacy according to the Q&HPI and eight comparisons used the RMNPI. Results showed a significant effect in favour of the PTB. The difference of Means (DiffM) was −0.14 (P < 0.001; 95%CI [−0.19; −0.09]) for the Q&HPI and −0.10 (P < 0.001; 95%CI [−0.14; −0.06]) for the RMNPI, respectively. The subanalysis on the OR mode of action showed a DiffM −0.16 (P < 0.001; 95%CI [−0.22, −0.10]) for the Q&HPI. For the SS mode of action using RMNPI, the DiffM showed −0.10 (P < 0.001; 95%CI [−0.15; −0.05]). The subanalysis for brands showed for the P&G OR PTB using the Q&HPI a DiffM of −0.15 (P < 0.001; 95%CI [−0.22; −0.08]) and the Colgate SS for RMNPI showed a DiffM of −0.15 (P < 0.001; 95%CI [−0.18; −0.12]). Conclusion There is moderate certainty that the PTB was more effective than the MTB with respect to plaque removal following a single brushing exercise independent of the plaque index scale that was used.


| INTRODUC TI ON
It is well established that natural oral self-cleaning mechanisms have no significant effect on dental plaque formation. Therefore, active removal of plaque at regular intervals is necessary. 1 Dental plaque leads to gingivitis and can eventually turn into chronic periodontitis. 2 Therefore, adequate oral hygiene is an essential habit for maintaining oral health. 3 Currently, there are numerous toothbrushes available on the market. The manual toothbrush (MTB) is a simple device which is widely accepted and affordable to most people. 1  ing to the efficacy of these new designs also continues to expand. 6 Whether powered brushing is superior to manual brushing has for long been a subject to controversy, as studies have demonstrated conflicting results. 7 However, the PTB has become an established alternative to the MTB. 8 The Cochrane Collaboration showed that the PTB is more effective in the reduction of plaque and gingivitis. This is based on studies with an evaluation time of 3 months or longer. 8 Single brushing exercise studies are considered to provide limited information since they do not take into account the benefits of gingival health. 9 Nevertheless, they are appropriate for assessing plaque removal, as they facilitate the control of confounding variables such as patient compliance. 10 Two previous published systematic reviews (SR) have determined the efficacy on plaque removal, following a single brushing exercise, on plaque removal of MTB and PTB separately. A head-to-head comparison with a SR approach of studies evaluating a PTB vs a MTB with a single brushing model is lacking. Collective evidence would help to guide the dental care professionals in making a well-considered recommendation for optimal plaque removal. Therefore, the purpose of this study was to review the effect on plaque removal of a PTB compared to a MTB following a single brushing exercise.

| MATERIAL AND ME THODS
This SR was prepared and described in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and the guidelines of Transparent Reporting of Systematic Reviews and Meta-analysis. [11][12][13] The protocol that details the review method was developed a priori following an initial discussion among the members of the research team.

| Focused question
In adult participants, what is, following a single brushing exercise, the efficacy of a PTB as compared to a MTB on plaque removal?

| Definition of a powered toothbrush
In the dental literature, "electric" and "powered" are used interchangeably for identical toothbrushes. It may be described in general as a powered device that consists of a handle having an electromotor which converts electricity into a mechanical action that is transferred to a shaft that propels the brush-head. 14 A large variety of PTBs are available to the consumer. For the purpose of this review, only toothbrushes with rechargeable batteries were included. Brushes containing a normal battery to provide an electric current, those that do not have a moving brush-head or those using a "switched off" mode, were not considered. 14

| Search strategy
A structured search strategy was designed to retrieve all relevant studies that evaluated the efficacy of a single brushing exercise in adults using either a PTB or a MTB. The National Library of Medicine, Washington, DC (MEDLINE-PubMed) and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched from inception to February 2019 for appropriate papers that evaluated the effect on dental plaque in a single brushing exercise in healthy adults.
The reference lists of the included studies were hand-searched to identify additional potentially relevant studies. For details regarding the search terms used, see Table 1.

| Screening and selection
Titles and abstracts from the studies obtained by the searches were independently screened by two reviewers (TAE, NAMR) to select studies that potentially met the inclusion criteria. Only papers in the English language were accepted. Based on the title and abstract, the full-text versions of potentially relevant papers were obtained. These were categorized (TAE, DES) as definitely eligible, definitely not eligible or questionable. Disagreements concerning eligibility were resolved by consensus or if disagreement persisted, by arbitration through a fourth reviewer (GAW). The papers that fulfilled all of the inclusion criteria were processed for data extraction.
The inclusion criteria were as follows: •

| Assessment of heterogeneity
Factors used to evaluate the heterogeneity of outcomes of different studies were categorized as follows: study design, subject characteristics, regimen details, mode of action, brands and plaque indices.

| Quality assessment
Two reviewers (TAE and DES) independently scored the individual methodological qualities of the included studies using the checklist as presented in Appendix S1 according to the method described in detail by Keukenmeester et al 20 In short, a study was classified as having a "low risk of bias" when random allocation, defined inclusion/ exclusion criteria, blinding to the examiner, balanced experimental groups, identical treatment between groups (except for the intervention) and reporting of loss to follow-up were present. Blinding to the participant was not taken into account as the participants could always see whether they used a PTB or a MTB. Studies that had five of these six criteria were considered to have a potential moderate risk of bias. If two or more of these six criteria were absent, the study was considered to have a high risk of bias. 21

| Data extraction
From the papers that met the selection criteria, the data were processed for analysis. If possible, the mean plaque scores for pre-brushing, post-brushing, change and standard deviations were independently extracted. This data extraction was performed by the three independent reviewers (TAE, NAMR and DES) using a specially designed data extraction form. Disagreement between the reviewers was resolved through discussion and consensus. If a disagreement persisted, the judgement of a fourth reviewer (GAW) was decisive. Some of the studies provided standard errors (SE) of the mean. If needed and where possible, the authors calculated standard deviation (SD) based on the sample size (SE = SD/√N). If the 95% CI, mean and sample size were provided, using Omni calculator (https :// www.omnic alcul ator.com/stati stics/ confi dence-interval), 22 the SD was calculated. For those papers that provided insufficient data to be included in the analysis, the first and/or corresponding authors were contacted to request additional data.

| Data analysis
Pre In studies consisting of multiple treatment arms and data from one particular group compared with more than one other group, the number of subjects (n) in the group was divided by the number of comparisons. A meta-analysis was only performed if there could be two or more comparisons included. 23 The difference of means (DiffM) between PTB and MTB was calculated using a "random or fixed effects" model where appropriate. A fixed-effect analysis was implemented if there were fewer than four studies because the estimate of between-study variance is poor for analysis with low numbers of studies. 11 The formal testing for publication bias was used as proposed by Egger et al 24 with a minimum of 10 comparisons.

| Grading the "body of evidence"
The Grading of Recommendations Assessment, Development and Evaluation (GRADE) 25 was used to appraise the evidence. 26 Three reviewers (TAE, DES, GAW) rated the quality of the evidence and the strength and direction of the recommendations 27 according to the following aspects: risk of bias, consistency of results, directness of evidence, precision and publication bias, and magnitude of the effect. Any disagreement among the three reviewers was resolved after additional discussion.

| Search and selection results
Searching the MEDLINE-PubMed and Cochrane-CENTRAL databases resulted in 3450 unique papers (for details, see Figure 1). Screening of the titles and abstracts resulted in 83 papers, which were obtained in full text. Based on a detailed reading of these papers, 66 papers were excluded. The reasons were no full-mouth scores, no single use or conducted among children. The other 11 studies did not fit the eligibility criteria of which one study was due to the fact that the PTB that was used was a prototype 28 and another did a long-term study including a single brushing exercise but unfortunately did not report the data. 29 In total, 17 papers were selected. [30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46] In various trials, more than one brush type was used to obtain data on plaque removal efficacy, resulting in 36 comparisons for inclusion in this review.

| Assessment of heterogeneity
The selected papers showed considerable heterogeneity in study design, participant characteristics and entry criteria, period of plaque accumulation prior to the brushing experiment, used products, PTB mode of action, brands, instruction method, brushing duration and plaque indices used. Appendix S2 showed an overview of these items in the selected studies.

| Study design
Of the 17 selected studies, 16 were randomized controlled trials or "acute lesions" or "hard tissue lesions" were defined as exclusion criteria. These descriptions were summarized as "dental neglect."

F I G U R E 1 Search and selection results
The method of instruction in oral hygiene practices was classified as "none" as reporting normal regimen or no instruction. Instructions according to the manufacturer or written instructions or leaflet were considered as "written." Professional instructions by a dental care professional, video instructions or if feedback was provided are classified as "visual." Plaque accumulation varied from 12 hours to 4 days, and brushing duration was 30 seconds till unrestricted time for self-performed brushing (for details, see Appendix S2).

| Plaque indices
Of the 36 comparisons, 28 comparisons used the Q&HPI or a modification. 15,16 In eight comparisons, the plaque scores were assessed according to the criteria as described for the RMNPI. 18,19

| Methodological quality assessment
The potential risk of bias was estimated based on the methodological quality aspects of the included studies as presented in the Appendix S1. Based on a summary of the proposed criteria, the potential risk of bias was estimated to be high for Kulkarni et al 46

| Study outcomes results
The Appendix S3 presents the data as extracted per study when the Q&HPI was used and Appendix S4 when RMNPI was used.
Consequently, studies are categorized by index and ordered by year of publication. Data are presented with respect to prebrushing, postbrushing, changes in plaque scores and the absolute difference in terms of percentage of plaque scores.   (Table 4).   Presented as overall and a subanalysis of the mode of action  Tables 3 and 4 show a summary of the MA outcomes. Detailed information regarding the forest plots and funnel plots can be found in the Appendices S6-S14. Table 5 presents a summary of the various factors used to rate the quality of evidence and to appraise the strength and direction of recommendations according to GRADE 25 including the level of certainty. There is a small difference in plaque removal in favour of the PTB. The single brushing design is rather direct as it does not reflect long-term use. As the risk of bias varied from "low to high" and many studies were industry-financed reporting bias cannot be ruled out. The strength of the recommendation was estimated to be "strong" due to the precision and rather consistent results of the plaque scores. Given the strength of this recommendation, there is a moderate rate of certainty of the beneficial effect of a PTB removing more dental plaque than a MTB. or the PTB. 14 These reviews showed that on average the plaque removing efficacy for the MTB was 42% and 46% for the PTB. Rosema et al 14 showed that brushes with rechargeable batteries yield higher reductions in plaque scores then replaceable battery-operated designs. 14 It was therefore decided a priori to include only rechargeable PTBs in the present review. Terézhalmy et al 28 was included in the SR of Rosema in 2016 as a replaceable PTB, but after critically re-reading the paper, it was excluded in this review because in the description of the brush, it was mentioned that this was a prototype and a special rechargeable battery was used. In addition, as a result of a search update, new studies (Re et al, 43 Gallob et al, 44

| Plaque indices
The RMNPI 18 and the Q&HPI 15 and their modifications are the two indices most commonly used for assessing plaque removal efficacy with toothbrushes. Although these indices score plaque in different ways, there appears to be a strong positive correlation between them. 49 The MA showed that the PTB is more effective than the MTB, independent of the overall plaque score used (Appendices S6a, S8a, S10a, S12a, S13a, S14a). Sicilia et al 50

| Publication bias and risk of bias
The analysis of funnel plots provides a useful test for the possible presence of bias in the MA. The capacity to detect bias will be limited when MA is based on a few number of small trials due to the fact that the methods for detecting publication bias relate effect size to sample size. 8,24 Publication bias in this SR might be subjectively Risk of bias (Appendix S1) Low to high Consistency (  (Tables 3 and 4) Precise

Reporting bias Likely
Magnitude of the effect (Tables 3 and 4

) Small
Strength of the recommendation based on the quality and body of evidence Strong Direction of recommendation:With respect to the removal of dental plaque, there is moderate certainty to advise a PTB over a MTB inferred since the funnel shape is asymmetrical (Appendices S7a-c, S9a-c, S11a-c). In the lower part of the funnel plots, studies are missing and the assumption is that these showed no beneficial effect and were therefore not published. 24,51,52 Publication bias can therefore not be ruled out.
In 14 studies, the instructions were given according to what the manufacturer did advice. Only three studies gave written instructions to the users of the PTB but no instructions to the MTB users. 34,37,45 This aspect can potentially introduce a bias as emphasis on the brushing method in the form of a written instruction can change the individual brushing skills. This may enhance the effect of the PTB over the MTB which will have an impact on the overall outcome. However, this was not apparent when a sensitivity analysis was performed. It does have an effect on the estimated potential risk of bias because the treatment was not identical for both the interventions.

| Familiarization phase and learning effect
Glavind et al 53

| Indication for clinical practitioners
Both the use of PTBs or MTBs has been reported to have positive effects on plaque and gingivitis reduction in many RCTs. Therefore, recommending the use of a toothbrush to patients is supported by evidence. 56,57 Many factors may be of influence for the effectiveness of toothbrushes including filament arrangement, filament orientation and angulation, filament size, filament shape and filament flexibility, brushhead size and brush-head shape. For PTBs, in particular, this may also be the brushing speed 58 as well as the presence or absence and characteristics of a timer. 4 The Cochrane Collaboration review concluded that the PTB reduces plaque and gingivitis more than a MTB both in the short and long term. 8 Based on the present review, it is justifiable to state that independent of the mode of action a PTB is more effective in reducing plaque as compared to a MTB.

| LI M ITATI O N S
The English language restriction could have introduced a language bias. However, over the years, the extent and effects of such a possible bias may have diminished because of the shift towards publication in English. 11 Blinding for the participant was not possible due to the fact that they see and experience whether they use a PTB or a MTB which cannot be excluded. For the examiners, blinding to the toothbrush is feasible. Blinding the examiner to the single brushing exercise deserves special attention mainly regarding the sound. Some of the studies have reported on this particular aspect. 35,40,45 Only full publications were taken into account. No abstracts from scientific meetings or data on file of manufacturers were sought.

| CON CLUS ION
There is moderate certainty that the PTB was more effective than the MTB with respect to plaque removal following a single brushing exercise independent of the plaque index score that was used.

| Scientific rationale for the study
Toothbrushing is generally accepted as the most efficient oral hygiene method.
Traditionally, MTBs are used, but the last decades' PTBs became more popular. Data from a comparison of MTB vs PTB in single brushing exercises have at present not been systematically evaluated.

| Principle findings
PTB and MTB are both effective oral hygiene devices for removing dental plaque. There is a small but significant difference observed in plaque score reduction in favour of a PTB.

| Practical implications
Consequently, for plaque removal in daily oral hygiene, with moderate certainty the PTB can be recommended over a MTB independent of the mode of action.

ACK N OWLED G EM ENTS
The authors wish to acknowledge their gratitude to Joost Bouwman, head librarian of the ACTA library and Laura Wiggelinkhuizen, for their help to retrieve papers used in this study. They are also grate-