A Cluster Randomised Controlled Trial of a Targeted Distributed Toothpaste and Toothbrushing Programme

Dental decay remains a global health concern with a significant proportion of children still suffering from this disease [1]. When decay becomes severe or a child is unable to cooperate with treatment under local anaesthetic, they may require dental extraction under general anaesthetic (DGA). Extraction of teeth under DGA is the most common reason that children aged 5–9 are admitted to hospital in the UK. Since 2015, over 25,000 DGA cases have been carried out every year in England, more than double the numbers admitted to hospital for the next most common reason of tonsillitis [2, 3]. It is not only the UK where high numbers of children require treatment under DGA, many countries show similar results but there are variations across locations in the provision and uptake of DGA [4]. DGA numbers in Australia remain one of the most common reasons for hospitalisation of children [5]. While the exact number of DGAs for the USA is unknown, it is estimated 100,000–250,000 paediatric dental sedations are performed annually in the United States [6]. Previous research [7] has suggested that when children experience severe tooth decay and during their wait for a DGA they can encounter pain (67%), sleepless nights (38%), and poorer school attendance (26%). Fluoride is orally health protective and has a crucial role in preventing caries [8]. There are numerous interventions which include the use of fluoride; from population-based initiatives such as water fluoridation, the application of fluoride varnish through dental practices or school programmes and encouraging patient’s behaviour to practice toothbrushing twice a day at home using fluoride toothpaste.

A variety of research has found that distributed toothpaste programmes in children are likely to reduce tooth decay. Davies et al. [9] found that 1,450 ppm toothpaste distribution provided a significant improvement in caries compared to a control group receiving 440 ppm. Ellwood et al. [10], found that the distribution of toothpaste at either low 440 ppm or high 1,450 ppm concentrations, to young children from deprived backgrounds, improved their tooth decay (as measured by “dmft”) and should in theory lower the risk of extractions. Whilst a further study [11] did not report a positive significant result when looking at the community wide effect on early childhood caries for the efficacy of a distributed toothpaste programme, similar benefits to the use of a high fluoride toothpaste (1,450 ppm) were highlighted. The incidence of caries has fallen dramatically in England since many of these studies took place [12] and is now concentrated in a subpopulation of the most disadvantaged children [13]. Therefore, community interventions that have worked previously worked might no longer be as effective when caries prevalence is reduced and concentrated in underserved groups.

It is important to note that, particularly given the cost of living crisis, oral health can be an area that really suffers. While toothbrush and toothpaste may seem an affordable item, this is not always the case. When budgets are tight hygiene essentials are reported as being at the bottom of the list, a YouGov survey showed that of a sample of respondents who stated they experienced hygiene poverty, 28% said they had gone without toothpaste, toothbrushes or essential dental products [14]. A recent survey by the BDA showed 81% of secondary schools had children who they said did not have access to toothpaste. Previous work has also shown access to toothpaste and toothbrushes can be a barrier [15].

Overall, it can be hypothesised that a distributed fluoride toothpaste programme to young, at risk children will confer benefits by reducing caries risk, and by extension the need for repeat extractions under DGA. This study addresses a gap in knowledge by targeting the intervention (distributed toothpaste and toothbrushing programme) at the most high-risk families. “High risk” is defined as previous family history of childhood extractions under DGA within a family; itself suggestive of severe dental caries and the need for a targeted intervention.

The aim of this research was to evaluate the effectiveness of a readily implementable targeted toothpaste and toothbrush scheme in reducing the incidence of referral for child DGAs within families where one or more children have undergone a DGA previously (high risk families).

Specifically, the study will

• ascertain the number of repeat DGA referrals for each reference child;

• evaluate the clinical effectiveness of the targeted distributed toothpaste (1,450 ppm) and toothbrushing programme in reducing individual referral for childhood DGAs within families identified as at risk using a cluster-randomised control trial (chosen both for practical reasons and to minimise contamination).

The study was a multicentre, recruiter/assessor-blinded, parallel, stratified, cluster-randomised controlled trial (RCT). Participants were recruited from seven hospitals in the UK. The family formed the cluster with a minimum cluster size of one, the child referred for DGA, with siblings as additional cluster members. Families were randomised at the cluster level in a 1:1 ratio. The primary individual-level outcome was referral for a DGA for caries issued between 6 and 24 months of follow-up and recorded using routinely recorded NHS information. Further details on the methods, recruitment sites, randomisation and statistical analysis can be found in the supplementary material (for all online suppl. material, see https://doi.org/10.1159/000539416). [16].

Of a total of 2023 families assessed for eligibility, 961 families (1,671 children) were recruited. Shortly after randomisation, 2 families (2 children) in the control group withdrew from the study upon learning that they would not be receiving the intervention, and one family allocated to the intervention group was withdrawn by the study investigator having provided incorrect information. Participants were analysed in their originally assigned groups. Follow-up data were available for 477 families (837 children) in the control group and 478 families (825 children) in the intervention group. Eight participants in the intervention group were lost to follow-up as they relocated during the study, and therefore, the intervention could not be provided, and outcome not assessed. The number of withdrawals/loss to follow-up was minimal and similar in both trial arms (Fig. 1).

Descriptive demographic statistics for participants providing follow-up data are presented in Table 1. The age of children recruited into the study ranged from 3.0 to 10.9 years with a mean of 6.7 (SD 2.01) years. The majority of families were in the most deprived quintile for area-level deprivation (54% control, 61% intervention), with very few families in the least deprived quintile (6% control, 7% intervention). Most families comprised one or two children, with a median number of children of 2 (IQR 1–2).

At 24-month follow-up, 57 children (6.8%) had been referred for DGA in the control group and 76 children (9.2%) in the intervention. In the control group 20 (4%) of the 477 consenting children presenting for DGA were subsequently re-referred during the follow-up period; in the intervention group 29 (6%) of the 478 consenting children presenting for DGA were subsequently re-referred during the follow-up period. Overall, 49 (5%) children were re-referred for DGA during the follow-up period. At 48-month follow-up, these numbers had more than doubled with 153 children (18.3%) referred in the control group and 164 (19.9%) in the intervention group (Table 2).

A population-averaged analysis accounting for the clustering of children within families was undertaken in order to assess the impact of the intervention on individual participants. For the primary outcome of DGA referral up to 24 months, we undertook the following analyses: an unadjusted model, a model additionally adjusted for the stratification variable, and an adjusted model that included age at recruitment, and area-level deprivation. Results were reported in terms of relative (relative risk [RR]) and absolute effects (risk difference [RD]). All analyses were carried out in Stata using the xtgee procedure with robust standard errors.

Results at 24 months suggest that the data are not consistent with a clinically meaningful treatment effect of a benefit of the intervention on referral for DGA at 24 months (RD 0.024, 95% CI: −0.001–0.05, p = 0.06; RR 1.36, 95% CI: 0.98–1.89, p = 0.07; see Table 3). For the follow-up period of up to 48 months, the percentages referred for a DGA were higher but comparable across the trial arms (see Table 2). From the adjusted models the Risk Difference in the number of referrals was 0.008 (95% CI: −0.03–0.05), Risk Ratio 1.07 (95% CI: 0.88–1.32)

While previous research has indicated distributed toothpaste and toothbrush packs can reduce the prevalence and incidence of dental disease (dmft/DMFT), the effects on subsequent referral for dental general anaesthetic, typically required where there is extensive and severe decay or where a child is unable to tolerate treatment under local anaesthetic or inhalation sedation, has been infrequently studied. Results of the adjusted analyses suggest that the intervention of distributed toothpaste and toothbrushes were not consistent with a clinically meaningful reduction in referral for DGA (within either 24 or 48 months) when targeted at families who had already experienced a DGA.

An important finding from this research is that families of those comprising any child with at least one dental extraction under DGA, as a result of caries, should be the target for future oral health interventions. Over a 2 year period 8% of children within the study were referred for DGA and after 4 years 19% were referred. Families within the study experienced another DGA within a relatively short period of time. This shows that the TIGER TEETH (Targeted Intervention General anaesthetic Extraction Reduction) study’s target of focus was correct and thus future programmes should target the same characteristics of a population; families of children at high risk of caries due to having an extraction under GA.

The distribution of eligible recruited families was higher in more deprived areas compared to less deprived areas reflecting the fact that original DGA referrals were higher in children who lived in areas of high deprivation. Fifty-nine percent of children were recruited to the study from the most deprived IMD quintile compared to 6% from the least deprived IMD quintile. This is in line with results from previous studies that severe levels of disease are more prevalent in children living in more deprived areas [16‒18].

The study design, a cluster RCT, comprising a large sample which was successfully recruited and retained throughout the study was a strength of this research and gives credibility to the results. The RCT design also reduced the risk of bias in the assessment of the effectiveness of the Tiger Teeth study’s distributed toothpaste programme. Allocation concealment was implemented but participants were unable to be blinded to the study groups once they received or did not receive the intervention. Data on referral patterns were collected from an electronic referral system. This reduces detection bias as this would not lead to systematic differences between the control and intervention groups in how outcomes were determined. With regards to attrition bias, the loss to follow-up were minor, and involved four families (eight children) emigrating during the course of the study in the intervention arm and two families withdrawing in the control group when they discovered they were not in the intervention group.

In addition, targeting a high-risk population who had already experienced decay, using a simple affordable intervention was important to research, as if it had proved successful the intervention could be easily implemented. The outcome is important from both a policy and patient perspective. Referrals for DGA can result in an extended wait for patients, often in pain, impacting on school attendance and sleep [7].

While the data were captured from hospitals across Greater Manchester, in the North West of England the results are applicable nationwide. Greater Manchester is one of the most diverse areas in the UK and, therefore, represents a variety of characteristic from the whole population. The need for dental general anaesthetic extraction for children is an issue both nationally and internationally, with children often waiting for extended periods of time to receive treatment. Therefore, the results provide valuable insight into this issue and are generalisable beyond those who took part.

The research team were not able to record how the intervention was used in the family home. The study did not employ any form of questionnaire to be completed by families beyond the information recorded at recruitment, given the poor completion of longitudinal follow-up questionnaires [19].

Reporting studies which are unable to observe a statistically significant benefit can be challenging, but the results of this RCT provide important information and has been the first to specifically target children in this manner; using an index child who has experienced a DGA along with their siblings to monitor referrals for DGA.

There are several possible explanations as to why the interventions may not have resulted in clinically meaningful reduced referrals into secondary care. Behaviour change is a complex area and there are multiple reasons why someone may or may not take up and maintain a behaviour over time. While families were supplied with sufficient amounts of toothpastes and toothbrushes to ensure everyone in the study received one every 3 months (and, therefore, resource was not an issue), it was unknown whether families used the toothpastes or if they were used enough. Brushing may have only been conducted occasionally and the time of brushing will have made a difference. Brushing in the morning and consuming a high sugar diet throughout the rest of the day will increase the risk of caries. While bright, visual, stimulating imagery was used in intervention packaging to appeal to children and to enhance motivation indirectly as much as possible, the intervention may not have impacted the motivation of parents and carers. When looking at changing behaviour a model called COM-B [20] describes three variables which need to be present for behaviour to occur. These are capability, opportunity, and motivation. The intervention provided both the opportunity and capability through regular oral health packs which contained information on the importance of brushing twice a day with fluoridated toothpaste. While increased opportunity and capability can impact on motivation, this component may need to be the focus in the future. There have been numerous interventions which aim to enhance motivation to improve oral health practices. One area which has received considerable recent attention is motivational interviewing (MI). MI is a behavioural intervention where a conversation is used to determine a specific goal, it is a “guiding” type of approach rather than a “directing” type of approach. The results from MI research aimed at changing the oral health, behaviours and attitudes of parents, caregivers and young children have been equivocal. Some studies have shown clear benefits with MI in reducing the number of surfaces affected by childhood caries [21] or development of new caries for those who had previously received a DGA [22]. Other studies have been unable to demonstrate a benefit of MI with populations at high risk from dental decay. A recent commentary of MI on the prevention of early childhood caries highlighted the challenges associated with using MI [23‒26], including timing to carry out the intervention, training and retaining appropriate staff, resources, etc. Therefore, while studies have shown MI has the potential to make an impact, there are challenges in implementing and maintaining such interventions. MI maybe beneficial in a programme of interventions and alternatives should be explored to understand the best combination of interventions which could impact in this area.

The outcome measure for this study was the incidence of referrals; this was chosen as an important outcome as treatment within secondary care is usually the result of extensive decay and, therefore, preventing this would indicate improved oral health and also reduce the burden on the NHS. This was an important reason why “referral” was chosen as an outcome rather than “caries increment.” While a referral is a reliable and accurate method to ensure incidence was recorded there are some problems in using this as an outcome which should be noted. As the outcome is based on a referral there is no way to discover if a procedure actually took place. A referral was chosen as opposed to actual attendance for a DGA given the prolonged wait list for DGAs, which has been exacerbated due to the COVID-19 pandemic. DGAs had to be postponed/reduced during COVID and since this time additional services have been put in place in an attempt to treat patients under a different pathway given the extended wait times for DGA within the geographical area studied. A DGA referral will usually take place following a dental examination by a GDP. There are a number of reasons a GDP may decide to refer on for a possible dental general anaesthetic to be carried out most often this is due to the level of decay seen and the ability of the child to cooperate with treatment; therefore, many DGA referrals occur in young children who have extreme decay which has progressed over time [27, 28]. While these are potential weaknesses in outcome measure, the RCT employed diminishes the impact as these should be evenly distributed between both the control and intervention groups.

As caries is multifactorial, toothbrushing is not the only element which will impact on oral health. Diet is a key factor and is again a complex area in which to alter behaviour of individuals and families. Sugar consumption is the cause of tooth decay, not lack of fluoride. Fluoride has a preventive effect against caries; however, sugar has the potential to overwhelm the effect of fluoride. When consumption of sugar is higher than 3% of total energy intake, caries will steadily accumulate throughout life, even in populations who are widely exposed to fluoride [29]. A 2019 cross-sectional study conducted in the UK also showed that deprivation level was associated with increases in consumption of six of the HFSS (high in fat, salt, and sugar) products including energy drinks in young children [30]. Recent research has demonstrated a diminishing caries prevalence which is now largely concentrated in the most disadvantaged groups [12, 13]. Whilst fluoride interventions have historically shown good evidence of effectiveness more recent studies have indicated these interventions are no longer having the same effect; with suggestions that caries is resulting from excessive sugar consumption that fluoride is unable to combat at these levels. Interventions to prevent extreme decay and subsequent referral for DGA may need to concentrate on sugar consumption [31‒33]. In addition an important contextual factor to consider is the provision of dental services for children. There continues to be a high need for dental services for children, within England and beyond. Access to these services are not always consistent across different locations. A recent survey of dentists in the North West of England found that 66% of respondents felt the current provision of paediatric dental services was inadequate [34]. Children who are unable to access regular, timely dental care could result in an increased number who will require more extensive treatment and potentially a subsequent referral into secondary care if their dental decay is left untreated. While accessing dental care is important to support and encourage prevention and for early treatment if caries is already established, given that a main cause of caries is sugar this needs to be the predominate focus for future population level interventions.

Targeted distributed toothpaste programmes alone do not appear, in this particular population, at this particular time to result in a reduction of referrals for a DGA. The majority of the children with a DGA recruited into the study were from the most deprived areas stressing continued health inequalities. This reality suggests the need for more complex interventions to sustain the dentition of young children. These could include interventions such as MI or other intensive prevention provided within a program of treatments, such as additional oral hygiene sessions, as recent research has shown this intervention has some success in reducing plaque and gingival bleeding (although not a difference in caries given the short time span of the study) for children who had previous had treatment under GA [35].

The team would like to thank Ishtiaq Hussain for his work sending out the distributed toothpaste/toothbrush packs and the Clinical Research Network and research support at each hospital who carried out the recruitment for the study.

This study protocol was reviewed and approved by Preston Research Ethics Committee (REC), approval number 16/NW/0057. Written informed consent was obtained from parents/legal guardians for all participants under 18 years old to participate in the study.

All authors have completed the ICMJE uniform disclosure at http://www.icmje.org/disclosure-of-interest/. Iain A Pretty reports funding from Colgate Palmolive, grants from NIHR, and editorial board member payments from Wiley-Blackwell. Michaela Goodwin and Tanya Walsh report funding from Colgate Palmolive and grants from NIHR.

This study was funded through an unrestricted grant from Colgate Palmolive as part of the work carried out by the Dental Health Unit, a unique collaboration between Colgate Palmolive and the University of Manchester. Funder had no role in study design, data collection, analysis, or decision to submit the paper.

I.A.P. contributed to conception, design, data interpretation, and critically revised the manuscript. M.G. contributed to conception, design, running the study, data acquisition and interpretation, and statistical analyses and drafted and critically revised the manuscript. L.M. contributed to running the study and data acquisition and critically revised the manuscript. T.W. contributed to design, data interpretation, and statistical analyses and critically revised the manuscript. K.A. contributed to the statistical analyses and drafted and critically revised the manuscript.

Trial registration: https://doi.org/10.1186/ISRCTN72962272.

The Tiger Teeth study investigators are committed to furthering research by sharing where possible de-identified individual participant data. The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants and participants were not expressly asked about sharing their data for further study. To apply for access to the data please contact the corresponding author (M.G.).