Self-reported bleeding on brushing as a predictor of bleeding on probing: Early observations from the deployment of an internet of things network of intelligent power-driven toothbrushes in a supportive periodontal care population

Abstract

Background: Bleeding on brushing (BoB) is an important sign of gingival inflammation. Recently, the use of intelligent toothbrushes and oral health Apps has shown potential to improve oral and periodontal health. In the context of the introduction of an Internet of things network of intelligent power-driven toothbrushes in a supportive periodontal care programme, the aim of this study was (a) to audit the adoption and retention of this new technology; and (b) to preliminarily assess the feasibility to gather data on BoB and associate them with clinical periodontal parameters.

Methods: 100 subjects with different periodontal case diagnoses participating in supportive periodontal care (SPC) were provided with and instructed on the use of an intelligent power-driven toothbrush connected with an App (I-Brush). Brushing sessions and occurrence of BoB were recorded through the App and stored in a data protection compliant backend. Subject retention was audited over time. BoB recorded in the two weeks before the SPC appointment was associated with clinical parameters observed by the clinician blind to App data.

Results: 75% of subjects provided data from using the power toothbrush and the App for a minimum of 10 brushing sessions over an average period of 362 days. Compared with baseline, subjects' oral hygiene, bleeding on probing and prevalence of residual pockets improved gradually while using the I-Brush. The number of BoB episodes in the two weeks leading to the SPC appointment and the number of residual pockets predicted BOP (p < .001) detected during the examination. App use in the previous two weeks was associated with lower plaque scores at SPC.

Conclusions: These preliminary observations indicate good adoption and retention of a mobile health system built around an intelligent power toothbrush in a SPC population. The App was able to gather clinically relevant information predicting the observed degree of gingival inflammation. Deployment of mHealth systems seems feasible in dental practice and may bring significant oral health benefits. More investigations are needed in this area.

Keywords: adherence; bleeding on brushing; bleeding on probing; gingival inflammation; mHealth; periodontitis; supportive periodontal care.

Conflict of interest statement

MST is a member of the European Advisory Board of the Procter & Gamble Company, the manufacturer of the intelligent power toothbrush and App that has been used in this study and has received personal fees not related to this work. Susanne Thurnay is an employee of the Procter & Gamble Company. All other authors report no conflict of interest with this study.

© 2020 The Authors. Journal of Clinical Periodontology published by John Wiley & Sons Ltd.

Figures

FIGURE 1

User retention. The figure illustrates subject retention up to 400 days since the time the I‐Brush and App were configured on their smartphones. Data collection was stopped in March 2019 while many users were still actively using the system. 14 subjects showed a retention of more than 600 days until that date [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Recorded brushing sessions for a representative subject. Each dot in the time of the day versus date diagram represents a single recorded brushing session. The subject has been active for more than a year. Green dots represent brushing sessions without self‐reporting bleeding on brushing, while the red dots illustrate those with bleeding on brushing. The open triangles show the dates of the SPC appointments in the clinic [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 3

Prediction profiler based on multiple linear mixed model analysis with bleeding on probing at a SPC session as dependent variable. Prediction plots of the significant factors in the model after stepwise backwards elimination of non‐significant factors (p > .15). In each plot, the grey‐shaded area represents the 95% confidence interval of the prediction. The red dotted lines represent the reference point for the prediction and the confidence intervals; they are set at the population average for these plots (an interactive prediction profiler is available in Appendix 1). FMBS is the percentage of sites with bleeding on probing at the SPC appointment, Pockets is the number of sites with probing depths 4 mm or deeper, Dx Perio is the periodontal case definition (health/gingivitis = 0, stage I to stage IV periodontitis = 1–4), Age is in years, count of App sessions in the 2 weeks before the appointment, brushing time is in seconds as recorded by the App and count of BoB episodes recorded by the App in the 2 weeks before the SPC appointment [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 4

Prediction profiler based on multiple linear mixed model analysis with oral hygiene level (FMPS) at a SPC session as dependent variable. Prediction plot of the significant factor in the model after stepwise backwards elimination of non‐significant factors (p > .15). In the plot, the grey‐shaded area represents the 95% confidence interval of the prediction. The red dotted lines represent the reference point for the prediction and are set at the population average for count of app sessions in the past 2 weeks (P2W) (an interactive prediction profiler is available in Appendix 1). FMPS decreased significantly with greater number of use of the App in the previous two weeks leading to the SPC appointment. This analysis includes subjects without the use of the App (count of 0) in the two weeks before the SPC appointment [Colour figure can be viewed at wileyonlinelibrary.com]