Efficacy of power‐driven interdental cleaning tools: A systematic review and meta‐analysis

Abstract Objectives To summarize the available evidence on the efficacy of power‐driven interdental cleaning tools (PDICTs) as an adjunct to tooth brushing compared to tooth brushing alone or tooth brushing combined with any other non‐PDICT in terms of interproximal plaque and gingival bleeding reduction in gingivitis patients. Material and Methods A systematic literature search was performed in three databases until March 20, 2022 with the following main eligibility criteria: (1) randomized controlled clinical trials (RCTs) with (2) at least 28 days of follow‐up in (3) gingivitis patients. Interproximal plaque and bleeding values were defined as the primary outcome variables and used for pair‐wise meta‐analyses. Results Sixteen RCTs were identified including data from 1258 participants at the final evaluation. Eight studies each investigated the effect of either a liquid‐based or mechanical PDICT; one of these studies tested additionally a combined liquid‐based and mechanical PDICT. Tooth brushing combined with a liquid‐based PDICT compared to tooth brushing alone did not result in better interproximal plaque values but in significantly lower interproximal bleeding values. Tooth brushing combined with either a liquid‐based PDICT or with a mechanical PDICT compared to tooth brushing and flossing achieved comparable interproximal plaque and bleeding values. The majority of studies reporting on patient compliance/preference favored the use of a PDICT, and except for a single study, which was reporting soft tissue trauma in two subjects from improper use of a mechanical PDICT, none of the studies reported adverse events. Conclusions Daily use of PDICT as an adjunct to tooth brushing significantly reduces interproximal bleeding. This effect appears comparable to that of flossing, while PDICT may achieve higher patient acceptance/compliance.

However, it has only a marginal effect on the interdental spaces, which are often narrow and difficult to clean and therefore prone to develop caries, gingivitis, and periodontitis (Galgut, 1991). There are numerous interdental cleaning tools (ICTs) in the market to be used as an adjunct to tooth brushing, with the aim to improve the patient's oral health care and reduce the risk for disease. In theory, the choice of the device should be site-specific, depending on interdental space accessibility, which in turn is defined by local anatomical (e.g., tooth position/malposition) and iatrogenic (e.g., prosthetic restorations, orthodontic appliances, etc.) factors, that is, not all ICT suit all patients and situations. Nevertheless, flossing remains a frequently recommended standard by dental professionals, especially for children and adolescents (Särner et al., 2010), although studies have indeed shown better plaque removal with interdental brushes, followed by toothpicks (Sälzer et al., 2015;Schmid et al., 1976;Yost et al., 2006).
In this context, another important aspect is patient compliance in terms of using an ICT systematically and consistently, which is often difficult to achieve (Smith et al., 2019). For example, it is reported that only about 1/3rd of patients floss once a day, irrespective of age (W. P. Lang et al., 1995;Srinivasan et al., 2019;Winterfeld et al., 2015), while in a survey among >2000 US adults, it was disclosed that about 1/4th of the patients tend to lie to their dentists about their flossing habits. Additionally, 36% would prefer doing other unpleasant activities instead of daily flossing (e.g., cleaning the toilet, working on the taxes, or washing a sink full of dirty dishes) (American Academy of Periodontology, 2019). Recent studies have shown that power-driven ICTs (PDICTs), either liquid-based, mechanical, or a combination thereof, potentially achieve a higher patient preference compared to other ICT (e.g., flossing) Lyle, 2012;Sharma et al., 2008;Shibly et al., 2001). Liquid-based PDICT have been introduced in the 1960-1970s, with the first oral irrigator aiming to clean the interdental space with a water jet (Lobene, 1969). Modern powerdriven oral irrigators (e.g., WaterPik, Water Pik, Inc.) and air/liquid flossers (e.g., Sonicare AirFloss) use either a stream of liquid or a stream of air with liquid microdroplets, with either low or high pressure, to remove plaque. Both types can be used with water only or with antibacterial and/or anti-inflammatory mouth rinses, such as chlorhexidine (Flemmig et al., 1990; N. P. Lang & Raber, 1981).
Mechanical PDICT, on the other hand, are basically electric vibrating or rotating interdental brushes, toothpicks, and flossers.
In general, PDICTs have not achieved wide acceptance within the dental community, which may partly be due to the lack of any comprehensive systematic appraisal of the literature about their clinical efficacy (i.e., plaque removal and prevention of gingivitis).
Thus, the aim of the present systematic review was to answer the following focused question according to the Population, Intervention, Comparison, Outcomes, Study Design criteria (Miller & Forrest, 2001): "In gingivitis patients (P), what is the efficacy of any type of PDICT (e.g., liquid-based, mechanical, and/or combinations thereof) (I) as an adjunct to tooth brushing compared to tooth brushing alone (C1) or compared to tooth brushing combined with any other non-PDICT (NPDICT) (C2) in terms of plaque (O1) and gingival bleeding levels (O2)?" 2 | MATERIALS AND METHODS

| Protocol and eligibility criteria
The present systematic review followed the criteria of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (Supporting Information: Appendix 1) (Page et al., 2021). The following inclusion criteria were applied: (1) articles written in English and published in peer-reviewed journals; (2) randomized controlled clinical trials (RCTs) with (3) at least 28 days of follow-up and (4) with the purpose to investigate the efficacy of PDICT as an adjunct to any type of tooth brushing (5) compared to tooth brushing alone or compared to tooth brushing combined with any NPDICT (6) in gingivitis patients (7) in terms of plaque and gingival bleeding reduction. Studies including patients with periodontitis, undergoing orthodontic treatment, or with psychological or physical disability, as well as studies focusing on implants, or in vitro, laboratory, and preclinical studies were excluded.

| Information sources and literature search
The search process included originally two sources (MEDLINE/ PubMed and the Cochrane central register of controlled trials [CENTRAL]), which were screened for articles from 1980 up to and including April 2020. The search was updated before submission for the period April 1, 2020 to March 20, 2022, and during the revision process for a third source covering the same time period (Embase).
Details on the search term are presented in Supporting Information: Appendix 2. A manual search from relevant literature and previous review articles was also performed. Finally, a forward search via Science Citation Index with the included papers was added.

| Data collection and extraction
Based on the above-listed eligibility criteria, the titles, abstracts, and finally full-texts were screened for relevance by two authors (P. E. J. and K. B.); if no abstract was available, the article was read in full-text.
In case of ambiguity, consensus through discussion was achieved together with a third author (A. S.). The following clinical parameters were extracted by two authors (P. E. J. and K. B.) from the included studies at baseline and at final evaluation for statistical analysis: plaque, bleeding, and gingival indices at interproximal sites only and as mean of all measured sites. Additionally, the following study details were extracted and summarized in tables, if reported: study design, number of participants, type of ICT investigated, type of intervention and control treatment, duration of the trial, loss to follow-up, patient preferences, adverse events related to the ICT, randomization process, blinding process, and participants' demographics, and overall and oral health status.

| Risk of bias assessment
All included articles were assessed for risk of bias using the RoB2 tool from Cochrane (Sterne et al., 2019). The following domains were evaluated at "low," "high," or "unclear" risk of bias: risk of bias arising from (1) the randomization process, (2) deviations from the intended interventions (effect of adhering to intervention), (3) missing outcome data, (4) measurement of the outcome, and (5) selection of the reported results. As the specific research question did not allow a proper blinding process, the effect of adhering instead of assignment to intervention was judged in the second domain. It was considered as unclear if no follow-up on adherence to the study protocol was reported and low if it was reported and not rising any concerns. The overall risk of bias for an individual study was judged as follows: low, if all criteria were evaluated to be of low risk; high, if at least one criterion was evaluated to be of high risk; and unclear, if at least one criterion was evaluated to be of unclear risk but no criterion of high risk.

| Synthesis of results
The following six comparisons were considered for grouping the studies: (1) Comparison 1: Brushing versus brushing + liquid-based PDICT.

| Statistical analysis
For the above-listed outcome parameters and comparisons, randomeffects pair-wise meta-analyses using the inverse variance method with a comparable follow-up period (i.e., either approximately 30 days or ≥90 days) were implemented. Restricted maximum likelihood [REML] to calculate heterogeneity (τ 2 ) was used and the Knapp-Hartung standard error adjustment to account for the small number of studies. Standardized mean differences (SMD: Hedges's g) were used to compensate for the different indices recorded in the original studies. In case of at least three studies, the 95% prediction interval was estimated and displayed in the forest plots. Statistical

| Study selection
A flowchart of the literature search is presented in Figure 1. In total, 1259 titles were identified, and after screening, 29 publications were read in full text and assessed for eligibility; 13 publications were excluded for various reasons (Supporting Information: Appendix 3).

| Study characteristics
An overview of the study design, study population, type of intervention and interdental tools used, and outcome measures applied is summarized below and provided in Table 1.

| Study design
All studies were single-blinded and arranged in parallel groups, with a single study (Hague & Carr, 2007) applying a cross-over design.
Further, all studies assessed full-mouth values, except for two studies, that either used the first premolars as study teeth (Stauff et al., 2018) or the Ramfjord teeth, that is, tooth numbers 16, 21, 24, 36, 41, and 44 (Walsh et al., 1989). The follow-up period ranged from 28 to 180 days, with 12 studies having a follow-up of 28-30 days, one study reporting data after 28 and 56 days, and three studies with a follow-up of 42, 90, or 180 days, respectively.

| Study population
The sample size of the individual studies ranged from 48 to 147 patients. Herein, the data of 1258 participants are summarized (1342 participants in total at baseline, with 84 lost during follow-up). No study included patients <18 years old and the mean age-if reported-varied from 22 to 48 years.

| Interventions
Tooth brushing was performed in two studies (Goyal et al., 2012;Lyle et al., 2020) with a powered toothbrush, while in 13 studies with a manual toothbrush; one study (Walsh et al., 1989) had four groups, that is, two groups used a manual and two a powered toothbrush. In the control groups of five studies (Frascella et al., 2000;Goyal et al., 2012Goyal et al., , 2018Lyle et al., 2020;Walsh et al., 1989) brushing alone was performed (i.e., no interdental cleaning), while in 10 studies the participants of the control group used brushing and flossing for interdental cleaning; one study (Hague & Carr, 2007)  One of these studies tested additionally a prototype, which was classified as a combined liquid-based and mechanical PDICT. Further, eight studies investigated the efficacy of a mechanical PDICT as an adjunct to tooth brushing; four of these mechanical PDICT devices F I G U R E 1 Flowchart of the inclusion process of studies for the systematic review were classified as powered flossers, one as a powered toothpick, and two different types of powered interdental cleaners.

| Outcome measures
Fifteen studies reported on various plaque indices, 14 studies on various gingival indices, and 13 studies on various bleeding indices.
Primarily mean interproximal and/or full mouth scores were reported, while some studies also reported separately buccal and palatal/ lingual scores.

NR
No adverse events, but two subjects of the test group expressed some concern about filament breakage, but this was considered as of no clinical significance. Gordon et al. (1996)  to use it. Two studies (Gordon et al., 1996;Shibly et al., 2001) indicated a clearly higher preference for the mechanical PDICT compared to flossing, and in one study (Stauff et al., 2018), the participants rated the liquid-based PDICT more comfortable, but less effective compared to flossing. All studies reported on the occurrence of adverse events related to the ICT, and in all but one, no adverse events occurred; a single study (Hague & Carr, 2007) reported soft tissue trauma from improper use of the mechanical PDICT, in two subjects.

| Risk of bias assessment
Except for a single study (Goyal et al., 2018), which had a low risk of bias, all studies presented overall with an unclear risk of bias. Mostly Domain 1 (randomization process) and Domain 5 (selection of the reported results) presented some concerns, either due to lack of information on the randomization process or due to lack of an a priori published study protocol allowing to compare the intended and finally performed the statistical analysis. For an overview see Supporting Information: Appendix 4.

| Synthesis of results
All studies reporting data at approximately 30 days were pooled for meta-analyses. Meta-analyses were possible for at least one of the outcome variables regarding Comparisons 1 (i.e., brushing vs.

| Comparison 4: Brushing + flossing versus brushing + mechanical PDICT
Five studies (Cronin et al., 1997(Cronin et al., , 2005Gordon et al., 1996;Hague & Carr, 2007;Shibly et al., 2001) contributing with six comparisons assessed interproximal plaque. Only two studies favored slightly the use of a PDICT and there was significant heterogeneity among the studies (I 2 = 63.4%, p = .02). Meta-analysis did not show a significant difference between the two interventions (SMD: 0.08; 95% CI: −0.43, 0.60; p = .69) ( Figure 2). The quality of evidence was judged as low (Supporting Information: Appendix 6c). Five studies (Anderson et al., 1995;Cronin et al., 1997Cronin et al., , 2005Gordon et al., 1996;Shibly et al., 2001)  bleeding indices; however, this seems not to be due to better plaque control, as captured by plaque indices. Furthermore, it appears that the clinical efficacy in terms of plaque and bleeding control of either liquid-based or mechanical PDICT is well comparable to flossing as an adjunct to brushing.
These findings are in general in line with those of previous reviews (Husseini et al., 2008;Ng & Lim, 2019;Worthington et al., 2019) showing that compared to brushing alone, the use of a liquid-based PDICT significantly improves gingival health, while in regards with plaque control, inconsistent results are obtained. The finding that the use of an ICT as an adjunct to brushing results in improved gingival health comes without surprise; it is well established that interdental cleaning for plaque removal is pivotal to prevent oral diseases such as caries, gingivitis, and periodontitis in susceptible individuals (Chapple et al., 2015;Sälzer et al., 2015Sälzer et al., , 2020. The finding, however, that the positive effect of liquid-based PDICT on interproximal bleeding scores is not associated with reduced amounts of interproximal plaque was indeed unexpected. It has been previously discussed that this observation may be due to an impact of PDICT on biofilm composition, thickness and/or maturation, and/or a stimulated immune response (for an overview, see Husseini et al., 2008); however, a proper, purpose-designed study is missing.
In this context, although PDICT did not show any clinical benefit compared to flossing, they might have advantages in terms of patient compliance/preference. In all six studies included in this review and reporting on patient compliance and/or preferences, PDICT scored most often clearly better. A similar higher preference for and/or more often the wish to continue to use a liquid-based PDICT has been F I G U R E 2 Forest plot for interproximal plaque reported in a recent study from our group, not included in the present review . As already mentioned, patient compliance in terms of using an ICT systematically and consistently is a matter of concern (Smith et al., 2019) and only a small fraction of patients floss daily (W. P. Lang et al., 1995;Srinivasan et al., 2019;Winterfeld et al., 2015), while a considerable number tends to lie to their dentists about their flossing habits (Periodontology, 2019). It may thus seem reasonable to assume that the use of PDICT may be beneficial in the long term in controlling/ preventing gingivitis/periodontitis, by facilitating better patient compliance; however, this has yet to be assessed in properly designed RCT with long-term follow-up. Nevertheless, in the only two studies (Cronin & Dembling, 1996;Isaacs et al., 1999) (Salles et al., 2021) or in the treatment of peri-implant mucositis (Bunk et al., 2020;Magnuson et al., 2013;Tütüncüoğlu et al., 2022). Thus, studies including comparisons with other relevant NPDICT (e.g., interdental brushes) and reporting on interproximal values, long-term data, patient-reported outcome measures, and other patient groups, are needed.
In conclusion, considering the limited number of original studies and/or comparisons, the following conclusions can be drawn regarding the use of PDICT in gingivitis patients: − Liquid-based PDICT significantly reduces interproximal inflammation compared with brushing alone; however, this seems not to be due to better plaque control.
− Liquid-based and mechanical PDICT show a similar efficacy as flossing in terms of plaque and inflammation control.
− PDICT may achieve higher patient acceptance/compliance.