Abstract
Purpose Dental floss has been promoted reduce the effects from interdental microbial biofilm, however its efficacy has been questioned in the literature. The purpose of this study was to compare daily flossing instructions using an adapted horizontal vertical flossing technique (AHVFT) and routine oral hygiene on gingival inflammation as indicated by bleeding on probing (BoP).
Methods This randomized single-blinded controlled clinical trial was conducted with non-smoking adults presenting with gingivitis and no other systemic diseases. Eligible participants were recruited from a dental school patient population and were randomly assigned to one of two groups. Group A (experimental group) was instructed in how to use the AHVFT once daily and Group B (control group) was asked to continue with their regular oral hygiene practices. Clinical evaluations (interproximal BoP measurements) were performed by blinded, calibrated examiners at two, four, and eight-week intervals; the percentage of sites with interproximal BoP was recorded. Descriptive and inferential statistics were used to analyze the data.
Results A total of 36 eligible participants were enrolled in the study: Group A (n=18), Group B (n=18). The mean (±SD) percentage of interproximal sites with BoP was 26.3 (±4.7), 8.6 (±7.3), 7.2 (±5.2), and 7.9 (±5.8) for Group A at baseline, two weeks, four weeks, and eight weeks, respectively. The corresponding values for Group B were 25.8 (±9.9), 15.6 (±12.4), 12.3 (±12.2), and 18.0 (±13.1), respectively. The percentage of sites with BoP was significantly lower for Group A than for Group B (p=.015 at two weeks, p=.009 at four weeks, and p<.001 at eight weeks). Within each group, the percentage of sites with BoP was significantly lower when compared with baseline (p<.008). At the final visit, the percent reduction in BoP for Group A was 70% and 30% for Group B compared with baseline. Most (88.2%) of Group A participants demonstrated mastery of the AHVFT at eight weeks.
Conclusion Participants who received Instruction with the daily use of the AHVFT were shown to have reductions in interproximal BoP as compared to participants who had not received instructions in the AHVFT. Positive gingival health outcomes with dental flossing may be technique sensitive.
- dental flossing
- dental plaque
- bleeding on probing
- gingival health
- interdental cleaning
- oral hygiene instructions
INTRODUCTION
Epidemiologic studies have demonstrated that gingivitis is found in all age populations and is the most common form of periodontal disease.1-3 The American Academy of Periodontology and the European Federation of Periodontology classify gingivitis in part by the presence of gingival inflammation as evidenced by >10% sites with bleeding on probing (BoP).4.
The most popular dental devices that are marketed to help remove and thereby reduce the effects from interdental microbial biofilm include: dental floss or dental tape, interdental brushes, toothpicks, rubber tips, and oral irrigators. Dental floss/tape disrupts and removes interdental plaque with thin strands of nylon or other materials, which pass through the interproximal contacts of teeth and must be adapted to the proximal surfaces of the teeth.5
The effects of flossing to reduce gingival inflammation has been reviewed in the literature. The United States (US) Department of Health and Human Services (HHS) Food and Drug Administration has recommended flossing since 1979.6 Subsequently, the Associated Press reported that flossing was removed from the guidelines due to weak scientific evidence based on a 2015 systematic review of the literature and meta-analysis.7 Studies included the 2015 review, used by the Food and Drug Administration, had not investigated the effects of instructed daily flossing utilizing a specific technique to demonstrate a reduction in interproximal gingival bleeding and inflammation. Only two clinical trials in the review utilized the American Dental Association (ADA) flossing instructions as published on their webiste.8 This lack of consistency led to increased variability amongst the studies. A survey conducted in hospital and private practice settings found that the prevalence of flossing is only 20%, and of those, only 40% presented with acceptable flossing techniques. Effective flossing techniques, as described by the ADA were described as relative to the number of times the floss is moved in a corono-apical direction with a horizontal adaptation to the proximal surfaces of a tooth.9 Another study using an in-home dental examination of over 300 participants demonstrated that only a third of the study population was flossing, and only 22% of participants of those participants satisfied the criteria for an acceptable flossing technique.10 A third study showed the same number to be at 20% of given study sample present with acceptable flossing techniques.11 Thus, the daily use of dental floss has not been shown to be a common practice, nor the recommended technique for its effective use.
Recognizing the importance of interdental plaque removal, oral health care providers continue to recommend daily flossing even though there is limited evidence on the details of the flossing technique in the literature. Most previous studies have not investigated the effect of instructed daily flossing with a specific technique to demonstrate a reduction in interproximal gingival bleeding and inflammation.
The adapted horizontal vertical flossing technique (AHVFT) is similar to the flossing technique promoted by the ADA except that it adds a horizontal buffing component (bucco-lingually as well as corono-apically). The purpose of this study was to compare instructed daily flossing using the AHVFT with non-instructed, routine oral hygiene practices on gingival inflammation, as evidenced by interproximal bleeding on probing (BoP). The secondary and tertiary aims were to compare BoP scores at baseline with scores at each of the follow-up visits and evaluate the role of repeated flossing instructions on efficacy as shown by interproximal BoP.
METHODS
This was a parallel-group, randomized single blinded controlled clinical trial conducted over an 8-week interval conducted in the Tufts University School of Dental Medicine (TUSDM), Boston, MA, USA. The TUSDM Institutional Review Board (IRB) approved the study (IRB# 12560) and it was registered on clinicaltrials.gov (Identifier: NCT04218994). Eligible participants were randomly assigned to one of two groups: experimental Group A (instructed in AHVFT flossing) and control Group B (“do what you normally do for dental hygiene”) with a 1:1 allocation ratio. The randomization scheme was created using the “sample” function of a statistical software program (R Statistical Package v. 3.1.2).
Sample Population
Thirty-six participants were recruited from January 2018 to December 2018. The number of participants was chosen based on a power calculation using statistical software (nQuery Advisor v. 7.0). Previous study results were used to calculate the anticipated effect size. Specifically, Walsh and Heckman found a mean decrease in interproximal bleeding of 12% for the test group and a mean increase of 17% for the control group (for an overall difference of 29% between groups).12 As a between-group difference of 20% or more was deemed to be clinically significant,4 the between-group difference of 29% was in the range of values considered clinically significant. The standard deviation of the change score was estimated to be 18.4% for the test group and 15.6% for the control group based on the above reference. Assuming a mean between-group difference of 29% and the standard deviations above, a final sample size of n=14 per group yielded greater than 99% power at the α=5% significance level. Moreover, this sample size still produced power more than 80% at the α=5% significance level when assuming a clinically significant between-group difference of 20% (rather than a between-group difference of 29%). Thus, a target final sample size of n=14 per group was specified. To account for approximately 20% attrition, the sample size was set at n=18 per group or n=36 for the population sample.
Participants eligible for consideration needed to be 18 years or older, demonstrate >20% interproximal BoP, have had a professional dental hygiene therapy (PDHT) appointment within four months prior to entering the study, not have interproximal probing depths greater than 4mm, and present with a diagnosis of gingivitis or stage I, according to the AAP 2017 Classification of Periodontal and Peri-implant Diseases.3 A maximum of four months since the last PDHT visit was chosen to eliminate heavy calculus as a confounder, and was not to interfere with the regular PDHT interval for the subject during the eight-week study. Dental students were eligible to join the study provided they met all of the inclusion criteria. Candidates were excluded from the study if they used tobacco products, had orthodontic appliances, splinted fixed prostheses, defective interproximal restorations, interproximal caries that was clinically evident and negatively impacting gingival health, malposed teeth that precluded the use of floss, probing depth > 5 mm interproximally, a systemic disease that could affect the gingiva, or were taking medications that could affect the gingiva. Individuals who were physically incapable of flossing their teeth and pregnant women (self-reported) were also excluded. An initial survey, prior to a clinical examination, was taken to screen participants. The survey also included questions about brushing and flossing habits.
PROCEDURES
Calibration of the blinded and non-blinded examiners was performed prior to beginning the trial. A UNC 1-15 mm periodontal probe (Hu-Friedy Group; Chicago, IL, USA) was used to measure gingival probing depths, and all examination findings were recorded. Probing depths were documented on six sites per tooth. The blinded examiners used to qualify the subjects for the study were calibrated for probing depth measurement and BoP. Calibration entailed recording the probing depths in the same quadrant twice. Both sets of periodontal chartings needed to be within 1mm of each other; these numbers were used as the standard. Recorded results were compared between examiners and needed to be within 95% agreement. The non-blinded examiners were calibrated for the initial screening purposes before group assignments and for teaching the AHVFT by demonstrating the technique on dental models and on each other.
Experimental Group
Participants randomly assigned to Group A were instructed to use their toothbrush as usual and to floss once daily using the AHVFT. The instructions provided for the daily AHVFT were similar to the method recommended by the ADA with the added technique that the floss should be moved in short buccal-lingual motions (horizontal) as well corono-apically (vertical) motion and be well-adapted to the interproximal surfaces and the line angles of the teeth (Figure 1). The instructions were first explained by a non-blinded examiner on a dental model, and then the examiner demonstrated the flossing technique on one side of the participant’s mouth. The participant was then asked to perform the technique on the other side of the mouth; corrections were made as necessary. Group A participants were also instructed not to use any other interproximal devices or mouthwash products. Group A participants were asked to use the dental floss that was provided throughout the study (Sunstar GUM Butlerweave, 60-yard, mint waxed; Sunstar Americas, Inc., Schaumburg, IL, USA). Group A participants were also given a daily flossing diary. The non-blinded calibrated examiner evaluated the flossing diaries and observed and reinstructed AHVFT as needed at each visit.
Instructions for the AHVFT1
Flossing
Done once a day, with no preference on timing of the day.
Technique
Use an approximate 32 cm of floss/tape (~15–18 inches), which is adapted on the ring (4th fingers of each hand) approximately at the level of the 1st knuckle.
The floss should be adapted enough that when the hands are held with fingers extended and the hands are pulled apart, the floss will be approximately 5–6 inches apart (~10–12 cm) and will not come off of the fingers.
The floss is then held between the first finger (index finger) and the thumb on each hand.
The floss is controlled by the thumb or index finger to prevent it from cutting into the gingiva.
Heavy pressure through the contact can be reduced by moving the floss in a bucco-lingual direction (sawing) with slight apical pressure when moving through the contact area
The floss is introduced initially at the distal end of a last tooth in an arch and tightly adapted to the curved surface of the tooth, like bending a “C” around the tooth.
Next the floss is introduced to the next adjacent tooth with the same motion, removing plaque from the mesial and distal surfaces of each tooth, moving around the arch until reaching the distal of the most posterior tooth on the contralateral side of the arch
Moderate lateral force is used, and the floss is moved in a short forward-backward motion as well as upwards and downwards motion as if one were drying their back with a towel. This helps to ensure contact with the mesial or distal surfaces and helps to prevent injuring the supracrestal gingival attachment.
1. Levi PA, Rudy RJ, Jeong YN, et al. Non-surgical control of periodontal diseases: A comprehensive handbook. Berlin: Springer; 2016. 234p.
Reproduced with permission from the authors.
Control Group
Participants assigned to Group B were instructed to continue with their usual oral hygiene practices. Group B was not given any instruction in microbial biofilm removal techniques, nor were they advised on how frequently biofilm removal should be performed. Group B participants were examined in the same way as Group A at each subsequent visit.
Procedure
Modification of the gingival bleeding index, indicating the presence or absence of bleeding in the interdental gingival crevices, was utilized to obtain an interproximal BoP score13 at each visit for both groups. The examiner gently inserted the probe interproximally at the most apical portion of the gingival crevice from line angle to line angle and from buccal to lingual or lingual to buccal interproximally. After probing the facial or lingual of each arch, the presence or absence of BoP on the mesiofacial and lingual or distofacial and lingual was assessed and recorded. Bleeding on probing was only recorded interproximally; cervical bleeding was not evaluated nor recorded. The number of proximal bleeding sites was calculated and then divided by the total number of proximal sites by 100 to calculate the percentage of sites of BoP for each participant. The calibrated examiners were blinded to the groups during each follow-up visit. Subjects were examined and data collected at baseline (visit 1), two weeks (visit 2), four weeks (visit 3) and eight weeks (visit 4). The rationale for the examination intervals was twofold. Instructing Group A with the AHVFT at an interval of two weeks helped assure that the participants were performing the flossing technique correctly; doubling that interval would be effective in determining the results of the AHVFT on gingival healing, as both groups exhibited at least 20% interproximal gingivitis, evidenced by BoP, at the outset of the study.
Flossing efficacy in Group A was evaluated at each visit. To be considered proficient, a subject had to demonstrate flossing on each tooth. The floss had to be moved in short buccal-lingual motion as well as corono-apical motion and be well-adapted to the interproximal surfaces and the line angles of the teeth (AHVFT).
Statistical Analysis
Descriptive statistics (means, medians, standard deviations, and inter-quartile ranges [IQRs] for continuous variables; counts and percentages for categorical variables) were calculated. Groups A and B were compared at each follow-up visit in terms of the percentage of sites with BoP; these comparisons were performed using a generalized linear mixed model (GLMM) in conjunction with contrasts. Generalized linear mixed models are a flexible class of models that account for the presence of multiple observations (here, multiple visits) per subject, as well as allowing for non-normally distributed data. The particular GLMM utilized can be selected based on the Akaike information criterion (AIC).14 For the data analyzed in this study, the lognormal distribution was selected based on the AIC. The analysis was performed using the PROC GLIMMIX procedure in a statistical software program (SAS 9.4; SAS Institute, Cary, NC, USA). This procedure allows for missing data under the assumption of missing completely at random or missing at random. In between-group analyses, Group A and Group B were compared in terms of percentage of sites with BoP at follow-up visits 2, 3, and 4, while adjusting for age, gender, and baseline percentage of sites with BoP. A global (omnibus) test was conducted to compare the groups based on the data from all follow-up visits simultaneously, adjusting for the covariates. Once the global test was found to be significant, comparisons of the groups at visits 2, 3, and 4 could be conducted individually using contrasts.
In within-group analyses, the percentage of sites with BoP were compared across the different visits with separate GLMM analyses conducted for Groups A and B. Post-hoc tests following a significant within-group global test were conducted to determine which visits exhibited significant differences for the given group, using the Bonferroni correction to account for multiple comparisons. In the post-hoc tests, all visits were compared to one another within each group.
Groups A and B were also compared in terms of demographic and baseline variables. The chi-square test was used to compare these groups in terms of gender. The independent-samples t-test was used to compare the groups in terms of baseline percentage of sites with plaque, baseline percentage of sites with BoP, and age, with the latter two variables log-transformed due to non-normality. The significance level was set at α=.05, except for tests employing the Bonferroni correction, in which the significance level was α=.05/6≈.008.
RESULTS
Thirty-six participants were recruited for the study and were randomly divided into Group A (n=18) and Group B (n=18). The progression through study enrollment, allocation, follow-up, and analysis is shown in Figure 2. Demographic and baseline plaque data are shown in Table I; there were no statistically significant differences between groups. Seventeen participants in each group completed all four visits. One participant in Group A completed the first three visits but did not attend visit 4 and one participant in Group B completed the baseline visit but did not attend any subsequent visits.
Consolidated Standards of Reporting Trials (CONSORT) flow diagram
Baseline demographics and plaque levels (n=36)
Bleeding Scores
Descriptive statistics for the percentage of sites with BoP (i.e., BoP scores) stratified by group and visit are shown in Table II. Although Group A’s mean BoP score at baseline was slightly higher than that of Group B, this difference was not statistically significant (p=.521). On the other hand, the global test within the GLMM analysis showed that BoP scores for Group A at follow-up visits (using data from all visits) were significantly lower than those of Group B (p<.001) when adjusting for age, gender, visit, and baseline percentage of sites with BoP. Furthermore, when comparing the BoP scores for Group A and B at each individual follow-up visit, mean scores for Group A were lower than Group B at every visit (Table II). Specifically, when considering the absolute difference between groups, the mean BoP score for Group A was 7.0% lower than that of Group B at visit 2 (8.6% vs. 15.6%), 5.1% lower at visit 3 (7.2% vs. 12.3%), and 10.1% lower at visit 4 (7.9% vs. 18.0%). Each of these comparisons was statistically significant (p=.015, p=.009, and p<.001, respectively).
Group bleeding on probing scores over time (percentages)
Regarding within-group comparisons via the GLMM analysis, the global test comparing BoP scores across visits was statistically significant for Group A (p<.001). In post-hoc tests, the difference between baseline and the other three visits was significant when applying the Bonferroni correction (p<.008), with higher BoP at baseline than at the other visits. Differences between visits 2, 3 and 4 were not significant (p>.008). This pattern of significance was also found when comparing BoP scores across visits within Group B: the global test was significant (p<.001), BoP was significantly higher at baseline than at visits 2, 3 and 4 (p<.008), and differences between visits 2, 3 and 4 were not significant (p>.008). Therefore, in each of the groups, BoP scores were significantly higher at baseline than at visit 2, 3, and 4, but there were no significant differences between visits 2 and 3, visits 2 and 4, or visits 3 and 4 (Table II).
Flossing Proficiency
The performance of the flossing technique was evaluated for Group A at each follow-up visit. At visit 2, more than one half (58.8%)of the participants demonstrated the AHVFT correctly. At visit 3, 76.5% demonstrated the correct technique, and at visit 4, 88.2% demonstrated the AHVFT correctly. Most (82.3%) of Group B participants self-reported flossing at least once a day throughout the study period but none received instructions in performing the AHVFT or any other flossing method.
DISCUSSION
Participants who had received instruction in the daily use of dental floss using the AHVFT had significantly fewer BoP sites than participants who did not receive any flossing instruction and were not required to floss on a daily basis in this study. Bleeding on probing was used as the indicator of gingival inflammation. Because inflammation can be present in the absence of clinically visual signs, particularly in individuals with a thick phenotype, BoP indices are considered an earlier sign of interdental inflammation than other visible signs.15 Inflammation of the gingiva has systemic effects regardless of whether it presents as gingivitis or periodontitis. Controlling gingival inflammation through evidence-based biofilm removing techniques is of importance to oral health and overall health and well-being. Clinicians must be familiar with effective techniques for biofilm removal, such as the AHVFT, particularly for interproximal areas. In addition, chairside instruction in biofilm removal techniques was shown to have a positive impact on an individual’s ability to proficiently use dental floss.
Results from this study demonstrated that with repeated chairside instruction 88.2% of the experimental group mastered the AHVFT. Since flossing can dislodge both supra and subgingival interproximal microbial biofilm when the floss is properly adapted to the proximal tooth surfaces, it may also assist in preventing dental caries, as well as gingival, periodontal, and peri-implant diseases when used on a daily basis.16,17
At the final visit (8 weeks) the reduction of bleeding sites from baseline, considering all the interproximal sites with or without BoP, was 18.4% for participants in the experimental group. When this final percentage was divided by the initial percentage of bleeding sites (26.3%), the relative reduction of BoP was 70%. This reduction of BoP is similar to a previous study that reported reductions (67-69%) in BoP18 In the control group, the relative improvement of 30% in BoP, might be attributed to the Hawthorne effect where individuals in a study alter their normal behaviors when being observed. The Hawthorne effect could also have affected the participants in the experimental group. The reduction in BoP was also less when comparing visits 3 and 4 where there was a longer time interval between visits than the two-week interval between visits 1 and 2.19
Toothbrushing techniques and habits were not discussed with either group. This study only examined a specific dental flossing method (AHVFT) and its effect on interproximal BoP. As toothbrushing has not been shown to remove interproximal subgingival microbial biofilm, it was not evaluated as part of this study. Results of this study demonstrating the efficacy of instructions in the regular use of the AHVFT, combined with periodic feedback were not consistent with the results from the previous meta-analysis that demonstrated weak scientific evidence to support the use of dental floss.7
In this present study, the reduction in BoP in Group A was related to the chairside flossing instructions and the participants’ ability to learn and correctly perform the AHVFT on a daily basis. A previous meta-analysis demonstrated a small yet positive effect of flossing in reducing BoP, and reported that unsupervised flossing can be of minimal benefit.20 One-on-one flossing instructions and the demonstration of individual proficiency in the recommended technique was shown to significantly reduce gingival inflammation in the current study.
Strengths of this study include the level of evidence provided in a single blinded, randomized controlled trial and the variables examined. Results from this study emphasize that flossing without individualized instruction including feedback is of minimal benefit and that there is value to individualized instructions for daily flossing using the AHVFT. The primary outcome for this study was the reduction of BoP, a key variable in the diagnosis of periodontal diseases. The current findings suggest that achieving the correct technique of flossing can require multiple visits and continued instruction and reinforcement by dental professionals. The improvement of the subjects with the reduction of BoP over the study period provides measurable evidence that patients are capable of learning the AHVFT technique over time with individual instruction and coaching by dental professionals.
Limitations of this study include the binary bleeding score, evaluating only the presence or absence of BoP and did not measure the degree of intensity of the bleeding per site. The study population age was older than 18; children younger than 18 who also need to be taught how to use dental floss were not included. The sample population was dentally oriented, as subjects were recruited from individuals in a dental school setting, including dental students and dental assistants. This may explain why 82.3% of the participants in the control group reported flossing at least once a day at baseline. Another limitation was the type of periodontal probe used in evaluating BoP. An automated pressure sensitive periodontal probe might have allowed for more standardized data. Also, the participants were not classified based on the type of gingivitis (incipient, mild, moderate, or severe) which could have impacted the results.
CONCLUSION
Proficiency in the AHVFT was demonstrated by the majority of participants in a randomized control clinical trial. Individualized instructions and periodic reinforcement by dental professionals in the daily use of dental floss using the AHVFT can significantly reduce interproximal gingival inflammation leading to improved gingival health. Future multicenter studies with longer follow-ups, and a more diverse sample population are recommended to confirm the current findings.
ACKNOWLEDGEMENTS
The authors are grateful to Akshat Lalvani, BDS, Peng-Hao Huang, BDS, and Thomas Whitmer, DMD for their assistance as examiners and their support in conducting this clinical trial. The authors also appreciate the help of Drs. Britta Magnuson and Peter Arsenault for their assistance with the design of the study.
Footnotes
NDHRA priority area, Client level: Oral health care (health promotion: treatments, behaviors, products).
DISCLOSURE
This study was funded by the Department of Periodontology at Tufts University School of Dental Medicine, Boston MA, USA and Sunstar Americas, Inc. The dental floss (Sunstar GUM Butlerweave, 60-yard, mint waxed) was provided by Sunstar Americas, Inc., Schaumburg, IL, USA.
- Received October 14, 2022.
- Accepted March 13, 2023.
- Copyright © 2023 The American Dental Hygienists’ Association
