A randomised controlled trial to measure the effects and costs of a dental caries prevention regime for young children attending primary care dental services: the Northern Ireland Caries Prevention In Practice (NIC-PIP) trial

Epidemiology of caries in young children

Although dental caries is a preventable disease, it is a persistent international public health problem. A 2009 review16 of the available epidemiological data on caries from a number of countries suggested that the prevalence of dental caries was increasing and that the increases are concentrated in lower socioeconomic groups, new immigrants and children.

In the UK, national child dental health surveys are undertaken every 10 years and have been carried out since 1973, although Northern Ireland only started participating in the programme from 1983 and Scotland did not participate in the latest survey published in 2013. The Children’s Dental Health in the UK National Survey 200317 reported that 43% of 5-year-olds had tooth decay. The dental examination undertaken in the 2003 survey was an assessment of the ‘obvious decay experience’ of children’s teeth, defined as teeth that, at the time of the examination, had decay into dentine (including teeth that were filled in the past but which needed further treatment), were filled or were missing because of decay. Prevalence varied from 41% in England to 52% in Wales and 61% in Northern Ireland (data for Scotland were not reported). The 2003 survey showed that little had changed since the 1993 National Survey,18 which reported a prevalence of 45%.

The Children’s Dental Health Survey 2013. Report 2: Dental Disease and Damage in Children England, Wales and Northern Ireland3 results were released in 2015. The diagnostic protocol for the survey was changed from that of 2003 to enable dental caries to be measured across a range of detection thresholds. The rationale for this change was to reflect the way in which the detection and management of tooth decay has evolved towards more preventive approaches to care, rather than just providing treatment for disease.

There were also changes in the consent process; in previous surveys, negative (opt-out) parental consent was obtained for the dental examination of children. However, in 2006 the Department of Health in England produced guidance that required positive written consent from parents for the dental examination of young children participating in epidemiological surveys. As a result, the consent procedures used in the 2013 survey for dental examinations of 5-year-olds required written positive (opt-in) consent to be collected from parents. Children could also opt out on the day of the examination. Dental caries is closely associated with deprivation,6 but providing consent for school-based surveys is also associated with deprivation19 and, therefore, is likely to result in an under-representation of those children with the most severe dental caries. The 2013 survey reports ‘obvious decay experience’ which includes untreated caries that has progressed into dentine and caries that has previously been subject to restorative treatment (fillings) or tooth extraction. This categorisation includes both cavitated and ‘visual’ decay into dentine, the latter term describing caries lesions in which dentinal decay can be visualised through the enamel but without frank cavitation.

In the 2003 survey, using an opt-out approach, 88% of Northern Irish children selected received a dental examination. However, even with the opt-in approach used in the 2013 survey 79% of Northern Irish children selected received a dental examination. The changes to the diagnostic protocol and consent procedures meant that caries data for 5-year-olds reported in the 2013 survey were not directly comparable to those of previous surveys and, consequently, trends in caries among 5-year-old children were not presented in the 2013 survey (i.e. cross-sectional data for 5-year-olds were presented in isolation). Nevertheless, there is reason to believe that caries levels in 5-year-olds in Northern Ireland fell substantially between the 2003 and 2013 surveys. In 2003, in Northern Ireland, 61% of 5-year-olds had obvious decay experience. In 2013, with slightly revised diagnostic criteria and a changed consent process, 40% of 5-year-olds had obvious decay experience in their primary teeth.

Since the early 1980s, the NHS in England and Wales has funded a programme of local child dental health surveys complementary to the national child dental health surveys. These surveys of 5-year-old children are undertaken more frequently than the national surveys and involve much larger numbers of participants, which allows for statistics to be reported at lower levels of geography. Data from NHS surveys showed a significant reduction in caries prevalence in 5-year-olds from 2005/6 to 2007/8. 20 However, this reduction could be attributed to a change in the parental consent procedure between the survey in 2005/6, that used negative (opt-out) consent, and a requirement for positive (opt-in) consent in the 2007/8 survey. If consent was influenced by socioeconomic status, poor children with higher levels of disease might have been less likely to participate in the survey than their more affluent peers. 19 A further survey of 5-year-olds was published in 201321 using the same opt-in consent process. Comparing the results for England, prevalence fell from 30.9% in 2007/8 to 27.9% in 2011/12. However, we cannot assume that disease risk of children for whom consent was not provided was the same in both surveys. Nevertheless, these data suggest that disease levels across the UK are falling.

The UK has an enviable library of epidemiological data sets describing trends in prevalence and severity of dental caries in various population subgroups over the last 40 years. All of these data sets are cross-sectional and there are few prospective studies available to provide an understanding of how the disease behaves longitudinally. A prospective cohort study,22 published in 2008, followed 739 children aged 3–6 years attending 50 dental practices in the north-west of England over a 3-year period. This study demonstrated a stark difference between children who present with and without the disease at their first visit to the dentist. Over the study period, 25% of children who were initially caries free developed caries active; by contrast, 72% of those with the disease at their initial visit developed further cavities. No matter what age a child developed the disease, it progressed at the same rapid rate. An important finding of this study was that more ‘cases’ (children with caries active) arose from the initially caries-free population (n = 155, 21% of the total study population) than from those who presented with the disease at their first visit to the dentist (n = 118, 16% of the total study population). The study also reported that restoration (filling) of primary teeth made no difference to the trajectory of the disease. This is an important finding, as it points to the failure of secondary prevention (restoration) and demonstrates the importance of primary prevention in general dental practice.

Relationship with deprivation

A strong association between caries active in young children and deprivation has been reported consistently over the last 20 years in different countries. Inequalities in caries prevalence and experience have been demonstrated by poverty, race and ethnicity. In the USA, the National Center for Health Statistics reported findings of the National Health and Nutrition Examination Survey, 2009–2010, in 2012. 23 Among children aged 3–5 years, the prevalence of untreated caries was significantly higher in non-Hispanic black children (19.3%) and in Hispanic children (19.8%) than in non-Hispanic white children (11.3%). In this same age group, the percentage of untreated dental caries was significantly higher in children living at or below the federal poverty level (25.1%) than in children living above the poverty level (10.5%).

In the UK, both geographical and social gradients in caries prevalence and experience have been consistently observed. 24 The NHS 2011/12 survey21 reported a prevalence of 21% in the south-east of England (excluding London), compared with a prevalence of 35% in the north-west of England. The same survey reported a strong correlation between caries and deprivation score at lower tier local authority level. The Children’s Dental Health Survey 2013. Report 2: Dental Disease and Damage in Children England, Wales and Northern Ireland3 reported that 21% of 5-year-olds who were eligible for free school meals had severe or extensive tooth decay, compared with 11% of children who were not eligible for free school meals. In Northern Ireland, the percentage of 5-year-old children with severe or extensive dental decay showed a marked social gradient (Table 1): 38% of children in the most deprived quintile of deprivation [categorised using the 2010 Northern Ireland Multiple Deprivation Measure (MDM) quintiles25] were affected, compared with 10% in the least deprived quintile.

Impact of the disease

Once young children develop caries in their primary teeth, pain and extraction are common outcomes. A prospective cohort study conducted in the UK followed a population of 3- to 6-year-olds over a 3-year period. 4 Approximately one in five children with caries active presented with dental pain at an unscheduled visit at the dentist, compared with only 1 in 100 children who were caries free. In children with caries active, 1 in 10 had a primary molar tooth extracted each year. Dental extraction is the most common reason why young children have a GA. In 2013–14, in England, approximately 46,500 children and adolescents under 19 years of age were admitted to hospital with a primary diagnosis of dental caries. Most admissions were in the 5- to 9-year-old age group, among which group admissions showed a 14% increase between 2010–11 and 2013–14, from 22,574 to 25,812. The second highest number of admissions in 2013–14 were for tonsillitis, with approximately 11,500 cases, making dental caries the most common reason for children aged between 5 and 9 years being admitted to hospital. 27

In Northern Ireland, the significant reduction in dental disease identified between the 2003 and 2013 national surveys has been reflected in a reduction in the number of GAs for dental extractions. Data on the annual number of primary teeth extracted under GA in Northern Ireland provided by the Business Services Organisation of Northern Ireland are presented in Figure 1. From a peak of 33,686 teeth extracted under GA in 2004, the number fell to 22,056 in 2013. Figure 2 shows a similar decline in the number of children who had dental extractions under GA in Northern Ireland, from a peak of 8856 in 2004 to 5351 in 2013.

FIGURE 1.

The number of primary teeth extracted under GA in Northern Ireland 2002–13 (provided by Business Services Organisation of Northern Ireland, Business Services Organisation, Belfast, UK).

FIGURE 2.

Total number of children receiving a dental extraction GA in Northern Ireland 2002–13 (provided by Business Services Organisation of Northern Ireland, Business Services Organisation, Belfast, UK).

Extractions performed under GA have a negative impact on young children and their families,28 and there is a strong association between a history of dental extraction at a young age and the development of dental anxiety,29 which can continue to affect individuals in later life. 30

Children who develop caries active in early childhood are likely to have a high risk of the disease into adolescence. 31 In the Children’s Dental Health Survey 2013. Report 2: Dental Disease and Damage in Children England, Wales and Northern Ireland,3 15-year-old children were asked about the impact of oral disease on their daily lives. More than half (54%) of 15-year-olds who had severe or extensive decay had at least one problem resulting from oral health that affected their daily lives during the previous 3 months, compared with 44% of those with no severe or extensive decay. The problems most likely to be reported were embarrassment when smiling or laughing and difficulties with eating and cleaning their teeth.

Parents of 15-year-olds participating in the survey were asked whether or not the health of their child’s teeth and mouth had affected their family life during the past 6 months. Just over one-third (35%) of parents reported negative impacts resulting from oral health problems. The most frequent impacts were a parent having to take time off work (23%); the child needing more attention (15%); the parent feeling stressed or anxious (13%); and the parent feeling guilty (11%). Some 37% of parents of 15-year-olds with severe or extensive decay experience reported that their child had received a GA in the past as part of dental treatment, compared with 8% of parents whose children did not have severe disease.

Dental caries in young children also has a significant impact on NHS costs. It was estimated in 2014 that the NHS spends about £3.4B per year in England on primary dental care services. 32 The Health & Social Care Information Centre33 reported that in England 29.9 million patients were seen in the 24-month period ending June 2014, which included 7.9 million (26.4%) children. The same report identified that children received 10.6 million courses of treatment in 2013/14, just over one-quarter of the total number of courses of treatment provided for all patients (39.7 million). This activity consumes roughly one-quarter of the NHS dental budget, and is devoted primarily to the management of dental caries. These costs do not include GAs, as they are provided in hospital at a cost of about £700–800 per case.

Different types of NHS contracts with different in-built financial incentives are in place in the four home nations. In England and Wales a cost and volume contract, paying contract holders to deliver an agreed amount of activity for an agreed price using units of dental activity (UDAs) as the contract currency, has been in place since 2006. In both Northern Ireland and Scotland, patients are registered with a dentist and the majority of fees paid to dentists for the care of children is through capitation payments. In Scotland, the average cost to the general dental service (GDS) of treating a child during 2013/14 was £66. The total GDS spend on child dental care for the year ending March 2014 was over £68M; this total was made up of approximately 40% for items of service and 60% for capitation. 34

Table 2 shows that in 2013, in Northern Ireland, where the trial was conducted, 73.2% of children were registered with a GDS dentist. Registration has been increasing annually since 2010. Table 2 shows that there is a very steep increase in registration between the age groups of 0–2 years (29.3% registered) and 3–5 years (74.8% registered).

Table 3 summarises the costs of NHS dental care provided by the GDS in 2013/14 and 2014/15. The total costs in 2014/15 were £17.3M, with approximately 75% of costs made up of capitation payments.

Evidence base for interventions to prevent caries

Technologies designed to prevent caries fall into three broad categories designed to:

1. affect the dynamic balance of demineralisation and remineralisation at the tooth surface to favour remineralisation primarily through the use of fluoride

2. decrease the volume, and frequency of consumption, of refined carbohydrates

3. seal the surface of the tooth to insulate it from acid attack.

Dental visiting and prevention

There is a strong association between dental visiting and caries in young children. Those children who attend the dentist regularly and asymptomatically are more likely to have lower levels of caries than peers who are irregular, symptomatic attenders. 7 A Cochrane review56 examined the effect of altering the recall interval for dental check-ups and oral health and health-care system costs. The review looked at different recall intervals for different types of dental check-up: (1) clinical examination only; (2) clinical examination plus scale and polish; (3) clinical examination plus preventative advice; and (4) clinical examination plus preventative advice plus scale and polish. The review included only one study57 of 185 children (aged 3–5-years) and young adults (aged 16–20 years) attending a public dental clinic in Norway. Participants were randomly chosen to have a clinical examination every 12 months or every 24 months and were followed up for 24 months. For 3- to 5-year-olds, there was a non-significant difference in mean dmfs increment of –0.90 (95% CI –1.96 to 0.16) in favour of a 12-month recall. The study was judged to be of very low quality and the authors could not make any conclusions about whether or not extending the time between dental check-ups reduces or increases the risk of tooth decay and/or costs.

The reasons for the association between dental caries and dental visiting patterns remain unclear. There is no evidence to demonstrate that this relationship is directly attributable to interventions provided by dental services, or if regular dental visiting is a marker for a set of caries risk-reducing behaviours adopted by parents of young children within the home, such as restricting sugar consumption and frequent use of fluoride-containing toothpaste. However, if part of the national strategy for caries reduction is to deliver evidence-based prevention within dental settings, then sufficient numbers of children have to attend the dentist and attend at the recommended frequency if the benefits are to be realised at both the individual and population levels.

Overall, the child dental health surveys3,17 suggest there has been little change in dental visiting patterns across England, Wales and Northern Ireland between 2003 and 2013. Dental registrations of children in Northern Ireland have increased in this period, but this appears to be a result of an increase in the registration interval from 15 months to 24 months rather than a change in attendance behaviour. According to parental responses in the national surveys, approximately 9 out of 10 children across all age groups, and countries, attended for a regular check-up in both 2003 and 2013. There has also been little change in the reported age of the first visit to the dentist since 2003, with around one-third of 5- to 8-year-old children having visited the dentist by the age of 2 years in 2013. Despite high-profile media coverage about problems in accessing NHS dentistry, more than 8 out of 10 parents, in both 2003 and 2013, reported that they had never experienced any difficulty finding an NHS dentist for their child. So although visiting patterns have remained stable, there has been a decline in disease, suggesting that a change in age of attendance and the volume and frequency of attendance are not responsible for the fall in caries. However, the quality and effectiveness of the preventative care delivered at dental visits may have improved. Research that informed the rationale and design of this study suggested that the preventative care provided by GDPs was ineffective and inequitable58 and that dentists were ill equipped in terms of their knowledge59 and how they present information to their patients60 to provide an effective preventative service. In response to the concerns about the quality of prevention in practice, the Department of Health in England developed and distributed DBOH13 to every dental practice in England. The advice for prevention of caries in children aged 0–6 years provided by the third edition of DBOH is replicated in Table 4, and the classification used to grade the evidence is provided in Table 5.

The guidance recommends prevention for all children including those who present caries free. The levels of evidence for each intervention are provided; however, this evidence is based largely on explanatory studies and there is an implicit assumption in the guidance that providing advice, or a professional intervention, will replicate the effects of explanatory studies. The effectiveness of these interventions relies on changing parents’ and health professionals’ behaviour in a sustainable way. The translation of advice to a change in behaviour is tenuous for both patients61 and health-care professionals. 62 A large amount of policy and commissioning effort has been put into DBOH, and there is a concern that this effort may not result in significant improvement at population or practice levels. DBOH is currently not an NHS contractual obligation for dentists in England; however, the Health & Social Care Information Centre63 reported that in 2014/15 children’s courses of treatment that included fluoride-containing varnish application rose 24.6% to 3.4 million from the previous year. The total number of courses of treatment provided to children in 2014/15 was 11 million, 70% of which were for band 1 (check-ups) alone. So although provision of fluoride-containing varnish has been increasing, only a relatively small proportion of courses of treatment provided to children involve reported application of fluoride-containing varnish.

There is also a concern that DBOH could increase inequalities in caries levels between the rich and poor in society, as children from the most deprived backgrounds, with the greatest risk of dental caries, are less likely to attend the dentist from an early age, and are less likely to attend on a regular basis, than children from more affluent backgrounds. The Children’s Dental Health Survey 2013 Report 1: Attitudes, Behaviours and Children’s Dental Health England, Wales and Northern Ireland, 201364 recorded that around 90% of parents reported that their 5-year-old child attends the dentist on a regular basis. Table 6 compares parentally reported visiting patterns in the different countries.

Although these statistics may be subject to response bias, it seems that dental services can reach the majority of the child population. Other interventions are required to reach the children who do not attend the dentist on a regular basis. If prevention in practice is seen as a key means of improving population health in young children, we need to understand the cost-effectiveness of interventions recommended by DBOH. Estimating the cost-effectiveness of this approach to prevention is particularly important in the current financial climate of the NHS, as prevention delivered in dental practice is a potentially expensive option as a result of high staff costs. 15

Situation in 2008

The planning for this trial took place in 2008 and the application was approved in 2009. From the inception of the trial to its completion, policy has progressed with the intent to have prevention as the principal focus of NHS primary care dental services. More significantly, national epidemiological surveys show that there have been reductions in the prevalence and severity of the disease in the population of young children over the last 10 years, particularly in Northern Ireland where the fall in disease in 5-year-old children has been marked. Therefore, the trial has been conducted against a background of falling population disease levels. However, the NHS contractual arrangements under which dentists in Northern Ireland operate have remained stable during the period of the trial. During the conduct of the trial, policy-makers in all four home nations have sought to elicit a change in emphasis of NHS dental services to focus primarily on prevention rather than treatment. New NHS dental contracts with a system of remuneration based largely on capitation, aimed at supporting prevention, are being evaluated in England, Wales and Northern Ireland.

Caries in 5-year-olds: a priority

Prevention of dental caries in young children is a policy priority across the UK. The Primary Dental Care Strategy for Northern Ireland,65 published in 2006, placed a strong emphasis on prevention of caries in general practice and the subsequent An Oral Health Strategy for Northern Ireland,66 in 2007, sets targets for reduction in the caries levels of 5-year-olds. In England, tooth decay in children aged 5 years is an indicator in the Public Health Outcomes Framework67 and caries prevention in young children figures prominently in Public Health England’s document, Local Authorities Improving Oral Health: Commissioning Better Oral Health for Children and Young People. An Evidence-Informed Toolkit for Local Authorities. 5

In Scotland, Route Map to the 2020 Vision for Health and Social Care in Scotland68 identified preventative measures on alcohol, tobacco, dental health, physical activity and early detection of cancer as a particular focus. The priority for prevention of dental caries in young children is reflected in a national HEAT (Health Improvement, Efficiency, Access to treatment, Treatment) target: ‘At least 60% of 3 and 4 year old children in each Scottish Index of Multiple Deprivation (SIMD) quintile to receive at least two applications of fluoride varnish (FV) per year by March 2014’. 69 This forms part of the Quality Measurement Framework to measure progress of the National Quality Strategy to realise the 2020 vision. A national approach to prevention in Scotland started in 2011 with the establishment of the Childsmile programme, designed to improve the oral health of children in Scotland and reduce social inequalities both in dental health and access to dental services. There are a number of elements to the programme.

• Childsmile Core programme: every child is provided with a dental pack containing a toothbrush and a tube of toothpaste containing 1000 p.p.m. fluoride on at least six occasions by the age of 5 years. Children also receive a free-flow feeder cup by 1 year of age. In addition, every 3- to 4-year-old child attending nursery (whether it is a local authority, voluntary or private nursery) is offered free, daily, supervised toothbrushing.

• Childsmile practice: a network of Dental Health Support Workers facilitate regular attendance at dental practices of children from the age of 6 months to receive preventative care including twice-yearly fluoride-containing varnish applications from 2 years of age.

• Childsmile Nursery and School: educational establishments with the highest proportion of children living in the most-deprived local quintile, as defined by the Scottish Index of Multiple Deprivation, are targeted for provision of additional twice-yearly fluoride-containing varnish applications within the nursery and school setting.

Childsmile is a composite national programme. It is an evaluation programme, which has largely been confined to assessment of process. 70 Serial cross-sectional studies71 in Glasgow, an early adopter of the Childsmile programme, show reductions in dental caries in 3-year-old children, but without a control population causal inference is difficult and no data have been published on the cost-effectiveness of the programme.

In Wales, Together for Health: A National Oral Health Plan for Wales (2013–18)72 was published in 2013. A central element of the plan is ‘Designed to Smile’,73 and this national programme has two main elements:

1. A preventative programme for nursery/primary school children: this involves the delivery of school-/nursery-based toothbrushing and fluoride-containing varnish programmes for children aged 3–5 years, helping establish good habits from an early age. In addition, children aged 6–11 years will receive a fissure sealant programme as well as preventative advice on how to look after their oral health.

2. A preventative programme for children from birth to 3 years of age. The aim of this programme is to give good consistent advice to parents, to provide toothbrushes and toothpaste, and to encourage going to the dentist.

Similar to the situation in Scotland, evaluation has been hampered by the lack of a counterfactual and monitoring reports have focused largely on process evaluation. 74

In Northern Ireland, funding has been provided to the community dental services (CDSs) to run fluoride-containing toothpaste schemes, which started in 2005. There are three categories of scheme, with children aged 3–5 years (inclusive) being the targeted age group:

1. postal schemes: fluoride toothpaste is posted out in a pack, along with a toothbrush and instructions on use, to children from deprived families

2. supervised tooth brushing schemes: children in pre-school settings were overseen by staff in a daily brushing routine with fluoride toothpaste

3. pre-school distribution schemes: children attending pre-schools in selected areas receive toothbrushes, toothpaste and instructions at school for home use.

In total, across all three types of scheme, approximately 22,000 children were/are involved each year, which equates to about one-third of all children in Northern Ireland in the 3- to 5-year-old cohort. A personal correspondence with the dental lead for the NHS in Northern Ireland confirmed that this programme is now well established and has involved consistent numbers of children each year from 2005 to 2015 (Mr Michael Donaldson, Consultant in Dental Public Health. Health and Social Care Board of Northern Ireland, 2015, personal communication).

Therefore, prevention of dental caries in early childhood is a priority for all four home nations. In each country evidence-based preventative programmes using fluoride-containing varnish and fluoride-containing toothpaste are delivered in various settings, including general dental practice. However, there have been no pragmatic trials that investigate health and cost outcomes of these interventions. Serial cross-sectional surveys show reductions in disease and are encouraging, but the evidence produced is weak with a high risk of bias. This illustrates the need to undertake high-quality pragmatic trials to establish the cost-effectiveness of these interventions, particularly at a time when the NHS across the UK is seeking to invest in prevention by redesigning dental services.

Redesigning NHS dental services to support prevention

Internationally there has been a consensus that dental services should be reoriented to prioritise prevention. The Liverpool Declaration: Promoting Oral Health in the 21st Century75 was produced by the eighth World Congress on Preventive Dentistry, organised jointly by the International Association for Dental Research, the World Health Organization, the European Association of Dental Public Health and the British Association for the Study of Community Dentistry. The declaration called for a number of areas to be strengthened by 2020, including ‘countries should ensure access to primary oral health care with emphasis on prevention and health promotion’.

In England and Wales, locally commissioned NHS dental contracts were introduced in April 2006. One of the reasons the English Department of Health changed the NHS dental contract at that time was to encourage prevention. However, the 2006 contract in England was heavily criticised by the dental profession and NHS managers76 and by the House of Commons Health Select Committee77 for offering little incentive for dentists to provide preventative care. Indeed, one of the recommendations of the House of Commons Health Select Committee report was that ‘the Department of Health undertake research to determine the extent to which the provision of preventive advice is being given and its cost-effectiveness’.

In England piloting of new contracts started in 2011, based on a registration and capitation remuneration model with additional financial incentives for quality. Central to the pilot contract was an information technology-supported system of care pathways, based on the outcomes of a standardised risk assessment with a strong focus on prevention. The pilots have not been subject to a robust academic evaluation testing a priori hypotheses. The reports of the pilots have been mainly descriptive in nature78 and the evaluation of the standardised oral health assessment, which forms the cornerstone of the new contract, was limited to professional and social acceptability. The ability of the associated risk algorithms to correctly classify patients and predict future disease development was not tested, only a count of the number of additional appointments made for those considered to be at most risk was reported. During the evaluation, changes were made to the data collected, and there were concerns about the quality of the information being generated, for example only a small sample (n = 10) of the 70 pilot practices were chosen to record tooth-level data.

The Department of Health in England developed a Dental Quality and Outcomes Framework (DQOF),79 which was planned to be included in the pilots. The DQOF includes caries in 5-year-old children as a quality indicator: ‘Decayed teeth (dt) aged 5 years old and under, reduction in number of carious teeth/child’. 79 The rationale for use of this indicator was to ‘monitor the primary dental care team’s adoption of evidenced informed preventative advice and intervention and their impact on oral health’. 79 However, the DQOF was not included in the pilots of the new English NHS dental contract and is also untested. This makes conclusions about the impact of the English pilots difficult to infer. In January 2015, the Department of Health announced dental contact reform prototypes,80 acknowledging the frustrations with the pace of reform and reiterated a commitment to prevention. The remuneration mechanism for prototypes is to be based on a blend of capitation, activity and quality. Remuneration will be based on minimum activity and capitation targets, adjusted for performance against the DQOF. The timescales for the start and completion of prototypes are unclear and the phrase ‘evolution not revolution’ was mentioned.

In Wales, piloting of a new dental contract started in 2011. The National Oral Health Plan for Wales72 referenced the fact that the Department of Health was testing a new prevention-oriented GDS contract. Two pilots were selected, one basing payment on capitation and quality to incentivise practices to maintain and increase patient numbers and to promote prevention. The second was a children and young people’s pilot for 0- to 17-year-olds, which aimed to incentivise preventative care for 0- to 17-year-olds and was designed to complement ‘Designed to Smile’. The pilots were completed on 31 March 2015 and at the time of writing the evaluation has yet to be published. 81

In Northern Ireland, the Department of Health Social Services and Public Safety and the Health & Social Care Board of Northern Ireland have started pilots for a new dental contract. The first wave of pilots to test a potential capitation-based model for primary dental care started on 13 November 2014 and has paved the way for a larger intake of practices in wave 2, which commenced in August 2015. The wave 2 pilot will run for a period of 1 year and the practices involved will switch from the current fee-for-service system of remuneration to capitation and then back to fee for service at the end of the year. This change in the contract, like those in England and Wales, is designed to support prevention. The evaluation of the new contract pilots in Northern Ireland will be subject to an independent evaluation funded by a NIHR project grant [URL: www.nets.nihr.ac.uk/projects/hsdr/141912 (accessed 26 August 2016)].

This summary demonstrates that prevention of caries in the primary dentition of young children is a public health priority for the four home nations. In all four nations there is activity to reorientate dental services to focus on prevention. The Children’s Dental Health Survey 2013 Report 1: Attitudes, Behaviours and Children’s Dental Health England, Wales and Northern Ireland, 201364 reported that 90% of 5-year-olds in England, Wales and Northern Ireland had visited a dentist in the last 12 months. Therefore, interventions delivered in dental practice, in which over 90% of NHS dental resource resides, should be able to reach a high proportion of the child population. Although there is national guidance13,55,82 on providing evidence-based preventative care, primarily using fluoride interventions, to young children in general practice, the guidance is based on evidence that has some significant gaps, particularly how evidence from largely explanatory trials performs in the real world, taking into account compliance of practitioners and patients. There is also little understanding of the economics of prevention delivered in a practice setting. Dentists are highly paid health-care professionals; the Health & Social Care Information Centre15 reported that the average NHS earnings for providing-performer dentists in England and Wales was £114,100 in 2012/13 and the average for performer-only dentists was £60,800. There have been attempts to use skill mix to reduce staff costs for prevention, a good example being use of the extended duties dental nurse in the Scottish Childsmile programme, but these programmes are staff intensive and may not be as cost-effective as community-based prevention programmes such as water fluoridation or distributed fluoride toothpaste programmes in reducing population disease levels. There is a need to understand the costs and effects of preventative programmes delivered in general practice through well-designed and adequately funded, pragmatic, randomised controlled trials, particularly in the context of falling population disease levels.

Once young children develop caries active in their primary teeth, they are very likely to experience pain or have an extraction over a 3-year period. 4 In addition, the majority of preventative care in dental practice is directed to children who initially present caries free, and it is believed that the majority of 5-year-old children who have the disease emerge from this population. 22 As adverse outcomes are common in children once they develop the disease and the majority of NHS resources are directed to those who initially present caries free to keep them in this state, the rationale for this trial was to develop the evidence base for preventative care delivered in general dental practice, with a focus on primary prevention and attempting to keep children caries free.

Dentists cannot prevent caries from starting in children who at their first visit already have the disease. These children with the disease should be considered as a separate population from those who are caries free; their dental care needs are quite different and are complicated by the effects of restorative treatment. Another Health Technology Assessment programme-funded trial is investigating the management of active disease in this population. 11

If the technologies tested in this trial are shown to be effective at preventing caries and/or reducing costs it will reassure policy-makers that the investment in prevention and the reorientation of dental services is justified. If the intervention is shown not to be a cost-effective use of resources, policy-makers may wish to consider the merits of community-based prevention interventions.

Aim

The aim of the study was to measure the effects and costs of a composite fluoride intervention designed to prevent caries in young children attending dental services.

Objectives

The objectives of the study were to compare in children, aged 2–3 years who were caries free at baseline, the effectiveness of a varnish containing 22,600 p.p.m. fluoride, a toothpaste containing 1450 p.p.m. fluoride and standardised health education, provided twice a year in general dental practice, as a ‘preventative package’ versus standardised health education provided twice a year alone in:

• reducing the conversion of children from caries-free to caries-active states in the primary dentition

• reducing the number of carious surfaces (caries into dentine) in the primary dentition in children who convert from caries-free to caries-active states

• reducing the number of episodes of pain and/or extraction of primary teeth.

The cost-effectiveness of the preventative package relative to standardised health education alone was also evaluated.

Stage 1: recruiting practices

A flyer was sent out to all practices across Northern Ireland to participate in the trial. An open meeting available to all practices in Northern Ireland hosted by the Chief Dental Officer was initially held to explain the aims of the trial and what would be expected of practices if they participated in the trial. This was followed by a working meeting of representatives of all of the local dental committees in Northern Ireland to discuss the practical aspects of the trial, for example how to minimise disruption and what would be a fair financial package to reimburse practices for disruption and loss of earnings as a result of hosting the trial. A reimbursement package for practices was agreed with the Health & Social Care Board of Northern Ireland to cover the additional costs to practices from participating in the trial. This included:

• a £1000 initial payment to each dentist participating in the trial to cover out-of-pocket expenses for time taken for GCP- and trial-specific training and earnings lost because of the recruitment process

• a £25 fee for each visit to provide the intervention and complete the case report form (CRF) as per protocol.

Practices were selected based on their size (registered population of 2- to 3-year-old children), location (to provide a similar and management number of practices for each CDS examining team in each NHS trust area), willingness to participate in the study and an agreement to comply with the protocol and GCP requirements of the trial.

Stage 2: recruiting participants

Participants were children aged 2–3 years who attend the selected GDS practices.

Children were eligible to participate in the study if they fulfilled the following criteria.

Intervention group

The intervention was a composite fluoride intervention comprising two elements:

1. A fluoride-containing varnish (at a fluoride concentration of 22,600 p.p.m.) in the form of Duraphat^® (Colgate-Palmolive Ltd, Guildford, UK), provided in its normal commercial packaging. Duraphat (used off label) is classed by the Medicines and Healthcare products Regulatory Agency as an investigative medicinal product and, therefore, its use in this CTIMP had to comply with relevant UK regulations. 85 A participating dentist applied the fluoride-containing varnish to all of the dried primary teeth of the children at two visits to the dental surgery each year, at approximately 6-monthly intervals (± 4 weeks). One drop of varnish was applied to the primary teeth in each arch (two drops in total) using a standardised brush applicator. After application, parents were advised not to brush their children’s teeth for 24 hours.

2. Participating dentists and their staff were trained to apply the varnish in accordance with the product brochure and practices were provided with an illustrated fluoride-containing varnish application guide describing the process of application. The UK summary of product characteristics86 was also made available to the dentists. The varnish was dispensed by the pharmacy department at the Belfast Health and Social Care (HSC) Trust. The temperature of the varnish was monitored during distribution and storage using maximum and minimum thermometers to ensure the varnish used in the trial complied with the guidance in the product brochure.

3. The second element of the fluoride intervention comprised a free toothbrush and a free 50-ml tube of toothpaste containing 1450 p.p.m. fluoride. This element was provided to intervention group children twice a year along with the fluoride-containing varnish. The toothpaste was Colgate^® Cavity Protection (Colgate-Palmolive Ltd, Guildford, UK), which was provided in its normal commercial packaging. Parents of participating children under 3 years of age were advised to use a smear of toothpaste, and those whose children were over 3 years were advised to use a pea-sized blob of toothpaste when brushing their teeth. Photographs of a smear and a pea-size blob were included in a standardised dental health education sheet (see Appendix 1). It was stressed to parents that an adult must supervise the child when they brushed their teeth.

Control group

Parents of children allocated to the control group were invited to bring their children for a dental check-up at 6-monthly intervals. At these visits the children received the same standardised dental health education as the children in the intervention group. The control group children did not receive any professionally applied or NHS service-provided fluoride interventions.

The trial visits were integrated into the usual 6-monthly dental check-up appointment of all children in both intervention and control groups over the 3-year follow-up period of the trial.

The date of visits for both intervention and control groups, and the date of each application of fluoride varnish, were recorded for each participant in the intervention group by the dentist (local investigator) on the CRF. The CRF identified the batch number of fluoride varnish used for each application. Empty or expired tubes of varnish were collected and retained. Participants who did not attend for their check-up appointments were sent out reminder letters.

Primary outcome measures (measured at 3 years)

• Conversion of caries-free children to caries-active (caries into dentine) children. The diagnostic threshold for caries used in the trial was caries into dentine. We used the same diagnostic protocol and same examination processes and procedures that are used in the NHS child dental health surveys. 20,21 The clinical examination processes and procedures and the caries data collection form used in the CRF is included in Appendix 2. For the purposes of this study the term ‘caries free’ was used to denote a child whom the examiners (trained and calibrated to the diagnostic protocol used in the trial) judged to have no carious lesions into dentine. The term ‘caries active’ was used to denote a child that had at least one tooth with caries into dentine. Caries in the enamel were not assessed or recorded.

Secondary outcome measures (measured at 3 years)

• Number of carious surfaces (caries into dentine in primary teeth) that develop in children who convert from caries free to caries active.

• Number of teeth extracted in children who are caries active.

• Number of episodes of pain.

• Costs of dental care plus other health-care costs, as well as parental costs incurred as a result of visits to the dentist over the 3-year follow-up period.

• Adverse reactions (ARs) and serious adverse events (SAEs).

Table 7 identifies the variables used in the statistical analyses and how these were measured and collected.

All children who converted from caries free to caries active received dental treatment, for example fillings or extractions, in the usual way as prescribed by their dentist. All children who converted from caries free to caries active continued to receive the trial interventions (both intervention and control groups) for the duration of the trial.

Primary outcome

The primary analysis compared the proportion of children in the two groups who converted from caries free to caries active over the 3 years using a binary logistic regression model (‘logistic’ procedure in Stata). The primary analysis was adjusted for age and socioeconomic status (as measured by MDM 2010 quintiles). 25 The assumption of logit of outcome changing linearly with each unit change in age was tested by categorising age into equal intervals and rerunning the model checking for linearity. The likelihood ratio goodness-of-fit test was used to indicate model appropriateness and the Wald statistic was used to test the significance of the difference between groups.

We also report two other analyses: first, an unadjusted analysis and, second, an analysis adjusting for practice as well as age and MDM 2010 quintile. The latter analysis was a secondary analysis, as we did not consider the clinical placement of the varnish (or distribution of toothpaste) to be practice specific. This analysis used the Huber–White approach within Stata [vce(cluster)] to deal with potential practice clustering effects (also known as sandwich estimator and robust estimator of variance). This technique relaxes the assumption of independence of the observations and can produce the ‘correct’ standard errors even if the observations are correlated.

We undertook a subgroup analysis for children from deprived/non-deprived areas. This dichotomy was achieved by comparing children in the two most-deprived quintiles of the MDM 2010, based on their home postcode, with those categorised into the three least deprived quintiles of deprivation. This was formally investigated by means of an ‘interaction test’ of the null hypothesis: that the relative efficacy of the two interventions was the same in children in deprived and non-deprived areas. It should be noted, however, that the trial was not formally powered to detect socioeconomic status interaction effects; consequently, we expected to observe an interaction as being statistically significant only if this was very large.

Secondary outcomes

Primary outcome: conversion from caries-free to caries-active

The caries data for three binary outcomes are shown in Table 16. The primary outcome, the percentage of children who converted to caries active over the trial, was 34% in the intervention group and 39% in the control group. The primary analysis for estimating differences between caries-free and caries-active children with respect to study group, fitting a logistic model adjusted for age and socioeconomic status measured by MDM 2010 quintiles, was not statistically significant (OR 0.81, 95% CI 0.64 to 1.04; p = 0.11) (Table 17). This analysis was repeated categorising age into equal intervals. The likelihood ratio test indicated no difference between the fit of the two models [likelihood ratio chi-squared (3 degrees of freedom) = 3.84; p = 0.28].

The unadjusted model was similar to the estimate for the adjusted model with an OR of 0.81 (95% CI 0.63 to 1.04; p = 0.09), as was the adjusted model including the clustering of the dental practices, with an OR of 0.81 (95% CI 0.64 to 1.04; p = 0.10). Results for this model are shown in Appendix 5, Table 45.

It can be seen in Table 17 that the proportion of caries-active children was lower in the least deprived MDM quintiles. We undertook a subgroup analysis for children in deprived/affluent areas, as described in Statistical methods including methods for additional analyses. The ‘interaction’ test proved to be non-significant (p = 0.36), although this test does have low power (see Appendix 5, Table 40).

Secondary outcome: decayed, missing, filled tooth surfaces in primary dentition in children with caries active

Data for the discrete variables are presented in Table 18 only for children who were caries active. The secondary outcome was the number of carious surfaces in children who converted to being caries active. The mean number of carious surfaces among the 187 caries-active children in the intervention group was 7.2, compared with 9.6 among the 213 caries-active children in the control group. The adjusted multiple linear regression analysis indicates that this difference is statistically significant with a mean difference of –2.29 carious surfaces (95% CI –3.96 to –0.63 carious surfaces; p = 0.007) (Table 19).

Secondary outcome: number of extracted teeth in caries-active children

In the intervention group, 11.2% of caries-active children had teeth extracted over the 3-year period, compared with 13.1% of caries-active children in the control group (see Table 16), the mean percentage difference being 1.9% (95% CI –4.5% to 8.3%). The mean number of extracted teeth was 0.45 in the intervention group and 0.46 in the control group (see Table 18). A logistic regression model adjusted for age and MDM quintile was not statistically significant (OR 0.84, 95% CI 0.45 to 1.54; p = 0.56) (see Appendix 5, Table 47). The negative binomial model for the number of extracted teeth, which indicated significant overdispersion, was also not statistically significant (regression coefficient –0.03, 95% CI –0.88 to 0.82; p = 0.95) (see Appendix 5, Table 48).

Secondary outcome: number of episodes of pain

There were differences in the proportion of children with pain and the mean number of episodes per child between children who were or were not of caries-active status. The regression models therefore included caries status at follow-up as a covariate. There was no difference in the number of episodes of pain or proportion of children with toothache between the study groups over the 36 months (OR 0.95, 95% CI 0.69 to 1.30; p = 0.74) (see Table 16 and Appendix 5, Table 49). Forty-one per cent of caries-active children had toothache, compared with 9% of children who were caries free; this difference was statistically significant (OR 7.1, 95% CI 5.1 to 9.9; p < 0.001).

As it was difficult to determine single discrete episodes of pain (which went up to 17 episodes), this was capped for each child at a maximum of six over the 36-month period (this affected the scores of eight children). Among caries-active children, the mean number of episodes of pain was 0.85 in the intervention group compared with 1.08 in the control group. For all children, the negative binomial model, adjusted for caries status, for the number of episodes of pain, which indicated significant overdispersion, was also not statistically significant (regression coefficient –0.03, 95% CI –0.32 to 0.25; p = 0.81) (see Appendix 5, Table 50). There was a significant difference in the proportion of children with toothache between those who became caries active (164/400, 41.0%) and those who remained caries free (62/696, 8.9%) (OR 7.1, 95% CI 5.1 to 9.9; p < 0.0001).

Although not identified as outcomes for this study in the protocol, data for other caries indices such as dmft are presented for all children and those caries active in Appendix 5, Table 53. These data may be helpful in comparing the results from this trial with other studies. There was a statistically significant difference (p = 0.0013) in dmft index between the groups when all children are compared; mean dmft index was 1.15 [standard deviation (SD) 2.18] in the intervention group and 1.64 (SD 2.71) in the control group, indicating a relative reduction (prevented fraction) in disease of 30%.

Aim

To measure the effects and costs of a composite fluoride intervention designed to prevent caries in young children attending dental services.

Objectives

To compare, in children aged between 2 and 3 years who were caries free at baseline, the effectiveness of a varnish containing 22,600 p.p.m. of fluoride, a toothpaste containing 1450 p.p.m. of fluoride and standardised health education, provided twice a year in general dental practice, as a ‘preventative package’ compared with standardised health education provided twice a year alone in:

• reducing the conversion of children from caries-free to caries-active states in the primary dentition

• reducing the number of carious surfaces (caries into dentine) in the primary dentition in children who convert from caries-free to caries-active states

• reducing the number of episodes of pain and/or extraction of primary teeth.

The cost-effectiveness of the preventative package relative to standardised health education alone was also evaluated.

Objective 1

We did not demonstrate that the intervention prevented the conversion of children from caries-free to caries-active states.

Objective 2

We found a statistically significant difference between the intervention and control groups in the mean number of carious surfaces (caries into dentine) in children who converted from caries-free to caries-active states. Children who received the preventative package had on average 2.43 (95% CI –0.77 to 4.08) fewer tooth surfaces affected by decay than children in the control group.

Objective 3

We did not demonstrate that the intervention prevented episodes of pain or extraction of primary teeth.

Objective 4

The intervention was unlikely to be cost-effective in terms of either keeping children caries free or avoiding episodes of pain. The mean cost per carious surface avoided after 3 years was estimated at £251, with a wide 95% CI (£79.52 to £454.39).