Implementation of a behavioural change intervention to enhance oral health behaviours in elite athletes: a feasibility study

Patient and public involvement statement

We designed the intervention and study protocol with input from experts and an elite athlete focus group. We have provided a final report to each team.

Development of the intervention

Table 1 presents a summary of the BCW activities that we followed to develop the intervention.9

The most important behavioural factor to mitigate the risks, of both dental caries and periodontal diseases, is routinely performed oral hygiene with fluoride.11 Where caries risk is increased, high strength prescription fluoride toothpaste (PFT) is indicated.5 Therefore, we identified two key target behaviours as effective plaque removal, including interdentally and increasing fluoride availability with PFT 2600 ppm. Two key motivations for change were emphasised:

1. Performance enhancement: poor oral health impairs performance.

2. Appearance: poor oral health can affect appearance.

Thus, oral health was not necessarily the end goal but an approach to achieving high value motivators.

The final intervention

The final intervention was on two levels: level 1—athletes and support team together viewed one 10 min presentation focused on building motivation to improve oral health, and three 90 s information films focused on increasing oral health knowledge and skills to perform optimum oral health behaviour. Level 2—athletes alone received oral health screening, personalised advice and an oral health toolkit containing a manual toothbrush (GlaxoSmithKline (GSK), Sensodyne 3.5), PFT 2800 (Colgate Duraphat) and flosspicks (Procter & Gamble, Oral B Glide) to enhance COM.

Study design and setting

We conducted this pragmatic, repeated measures study at the training venues of one professional and two olympic athlete teams. Participation in the study was entirely voluntary and with no obligation and we obtained Informed written consent from each athlete. The study ran for three field periods of July–October 2018, August–December 2018 and February–April 2019 and we collected data at four time points:

1. Prestudy visit (−1 week):

   • Gain participant consent.

   • Collect baseline questionnaire data.

2. Study visit 1 (intervention):

   • Presentation to the whole team including athletes and medical support staff.

   • Oral health screening, clinical assessment of gingival health, provision of oral health report and oral health kit to athletes.

   • Athlete evaluation of the presentations and films.

3. Study visit 2 (4–8 weeks):

   • Collect follow-up questionnaire data.

   • Clinical assessment of gingival health.

4. Study visit 3 (12–18 weeks):

   • Collect follow-up questionnaire data.

   • Clinical assessment of gingival health.

   • Athlete evaluation of the oral health kit.

Outcome measures

The primary outcomes were: (1) athlete oral health knowledge, oral health behaviours, athlete-reported performance impacts. We measured knowledge with a questionnaire previously used in a study to assess oral health knowledge between dental healthcare professionals and the public.12 This questionnaire comprised eight questions with options of true, false or do not know. Each correct answer scored 1, therefore, the maximum possible score was 8. We asked the athletes about their use of toothbrush, fluoride toothpaste and interdental cleaning aids with items from a questionnaire used in a previous study.4 We used an oral health modification of the Oslo Sports Trauma Research Centre (OSTRC) overuse injury questionnaire13 (maximum possible score 100) to measure impact on performance in sport.14 14

Secondary outcomes included: gingival inflammation as a short term, objective measure of oral health, athlete evaluation of the presentation and the oral health kit and adherence to toothbrushing measured using a remote sensor, ‘Brushlink’. Gingival inflammation (bleeding) was measured as a bleeding on probing percentage score (maximum possible score 100%).

Baseline oral health was measured during oral health screening with standard clinical indices of dental caries, periodontal health and erosive tooth wear, as described in a previous paper.1 The same dentist administered the questionnaires, delivered the intervention and conducted the clinical measurements at each study visit.

Sample size and statistical analysis

As this was an exploratory study, we did not calculate a minimum sample size. However, for studies using self-reported outcome measures, an acceptable minimum sample size is generally considered to be 50,15 therefore, we aimed to recruit 55 athletes to allow for a drop-out rate of 10%. Data were coded for entry into a spreadsheet by a person independent of the study then imported to a standard statistical package (IBM SPSS Statistics for Windows, V.22.0) for analysis. Distribution of the data was assessed and continuous variables described as median with IQR and range. Counts and percentages summarise the categorical data. Non-parametric tests included the X² for categorical data and the Wilcoxon signed-rank test for continuous data. Significance values for all hypothesis tests were set at 0.05.

Characteristics of the group

Table 2 provides a summary of the demographics and baseline oral health of the group. Thirty (48.4%) were professional athletes and 32 (51.6%) were from Olympic training squads.

Primary outcomes

Table 3 provides a summary of the primary outcomes at baseline, visit 2 (4–6 weeks) and visit 3 (12–18 weeks). There was a statistically significant improvement in athlete knowledge (p<0.001), athlete-reported use of PFT (p<0.001), interdental cleaning (p=0.013) and mean OSTRC score (p<0.001) from baseline to visit 3. At each visit, a different athlete recorded a very high OSTRC score, due to an acute oral health problem. Most non-zero scores were in the minor range (0–30) and the majority of athletes scored 0 (no impact).

Secondary outcomes

The median bleeding score (table 3) decreased from 10 at baseline to 7.5 at visit 2 then increased to 10.5 at visit 3. Overall there was no significant change in the bleeding score (p=0.952).

Fifty-eight athletes completed a written evaluation of the presentation at study visit one and 55 athletes completed an evaluation of the utility of the oral health kit at study visit three. Table 4 provides a summary of the athlete feedback. The presentation appears to have provided a good level of motivation for the athletes to adhere to the recommended oral hygiene behaviours in order to avoid negative impacts from poor oral health. In addition, most athletes intend to continue to use the PFT.

Key findings

We implemented this pragmatic, repeated measures study in three separate elite athlete groups and a high proportion (88.7%) of athletes completed the study at 3 months despite their intensive training activities. Overall, this intervention was associated with improvements in athlete knowledge, enhanced oral health behaviour and athlete-reported performance impacts. Athletes said they were highly motivated to avoid negative impacts from poor oral health and 80% said they intended to continue to use PFT.

Strengths of the study

This study demonstrated high levels of participant retention and adherence to the intervention. We developed this study within a research framework recommended for use in the field of sports and exercise medicine; the KTS model for implementation of intervention.7 It was conducted exclusively in an elite athlete setting and was informed by data collected from a representative sample of Olympic and professional athletes1 4 to help understand the context of oral health in elite athletes.16 Our knowledge transfer group comprised behavioural change experts, dental experts and key stakeholders in elite sport including athletes and members of the support team. User engagement played a pivotal role in the development, implementation and evaluation of this project. The intervention design was based on the BCW, a contemporary behavioural change method for characterising and designing behavioural change interventions10 and the final ‘product’ developed using the COM-B model of behaviour that was incorporated into the KTS as an additional stage (table 5). The study, therefore, benefited from a clear and structured approach and this methodology could be applicable to other health behavioural change targets in sports medicine.

In relation to oral health the preventive interventions conformed with best available evidence contained in the guidance, Delivering Better Oral Health.5

Oral health promotion by dental professionals has been shown to have a positive effect on recipient knowledge, behaviour and gingival health.17 Therefore the 10 min presentation was delivered by a dentist (JG) and the three short (90 s) information films featured an Olympic athlete as a model. Information included the following elements: emphasise the benefits of adherence to the recommended oral hygiene routine, provide information on susceptibility to oral disease, encourage athlete to plan where, when and how the behaviour will occur and emphasise habit formation. During the oral health screening, feedback to each athlete provided a further opportunity to enhance motivation by emphasising cognitions likely to increase the recommended behaviour.18

The baseline oral health score for the athlete groups in this study were lower than the levels of oral diseases measured in our previous epidemiological study.1 The reason for this is not clear, however, we were not able to conduct representative sampling due to the requirements of this study and availability of athletes. The lower levels of disease could be a random chance phenomenon or reflect selective bias if more athletes with better oral health volunteered.

Limitations of the study

There were, however, some limitations to this study, the most important being the lack of a control group. We had originally planned for a randomised controlled trial, but teams did not wish to involve athletes in a control group. Coaches and elite team managers did not consider attendance for repeated monitoring of a non-intervention group an appropriate use of athletes’ time. Another is the reliance for several of the outcome measures on self-report.

Blinding of the examiner or the participants was not possible. Like many behavioural change interventions, there was a short-term improvement in knowledge and behaviours but no measurable health outcome changes. This study employed a number of elements and future research could aim to identify which are most effective from an athlete perspective. The intervention in this study was aimed at the whole group. Future research could further investigate the use of specific behavioural change techniques during feedback to individual athletes. This was a short-term study, conducted with limited resources, however, it provides a basis for a larger scale implementation initiative.

Implications for well-being and performance

In elite sport, performance is everything16 and there can be differences in perspective between a performance point of view and a medical viewpoint.19 20 High-performance sport is a time-pressured environment where immediate results can often supersede pursuit of the most effective evidence-based practice.21 We hypothesised that ‘loss aversion’ would be a more potent motivator (‘M’ from COM-B) than oral health improvement. The intervention was designed to target this aspect as well as another target which became apparent during the initial athlete engagement phase, that is, appearance. We conducted a poststudy evaluation which found that the factor which was most widely reported to have motivated the athletes (93%) to adhere to the intervention was avoidance of inflammation from oral disease elsewhere in the body. The use of short (10 min) intervention strands seemed most implementable in high-performance sport as they have been used widely for lifestyle interventions such as tobacco cessation. Elite athletes are a highly disciplined group and this intervention built on established oral healthcare routines, requiring little additional behavioural change from the athletes, demonstrating similarity with other models that place emphasis on an intervention being easy, attractive, social and timely. This study demonstrates that The KTS plus the BCW provides a framework on which to base future implementation research.