Introduction

Managing the dental care of young children is an important part of the role of family dentists. The findings of recent research1,2,3,4,5 in the United Kingdom have triggered a debate on how best to manage the dental care of young children.6,7 The debate has primarily focused on the restorative care of young children with carious primary teeth, however, additional concerns have recently been raised about the approach to preventive care taken by general dental practitioners, which seems to be ad hoc and not based on best evidence.8,9,10

Failure of practice based preventive regimes for children results in the development of caries. Once young children develop the disease, there is a high risk of adverse outcomes in terms of pain, infection and extraction.2,5,11 For those children with caries and access to regular primary dental care, the traditional model of dental care has involved repeated episodes of restorative treatment designed to maintain the integrity of teeth until exfoliation but a relatively unstructured approach to the more important need to prevent the disease. However, GDPs are faced with uncertainty about the strength of the evidence base supporting this approach.

Whilst questions have emerged about providing restorative treatment to children with carious primary teeth, there is a measure of certainty about the effectiveness of practice based preventive regimes for which there is a strong evidence base.12,13 If the preventive armamentarium available to the dental practitioner can be deployed effectively, then children are likely to develop less caries in the first place and the debate about the optimal approach to restorative care of the carious primary dentition might then be viewed from a somewhat different perspective.

In order for dentists to offer timely and effective preventive and restorative treatment to their child patients, we have to understand how dental caries in the primary dentition behaves over time, particularly in populations that attend general dental practice. The aim of this study therefore was to describe the progression of dental caries, as assessed by GDPs, in the primary molar teeth of young children who attended general dental practices in the North West of England and received their dental care funded by the National Health Service (NHS).

Methods

A practice-based observational prospective cohort study was undertaken. The population under investigation were children aged approximately 3-6 years attending 50 general dental practices in the North West of England within a defined recruitment period, whose parents provided consent for them to be included in the study. Children were recruited to the study between 22.1.2000 and 22.11.2001 and followed over a three year period. Recruitment was conducted in two phases; in the first instance all dentists in four health districts, Bury & Rochdale, Salford & Trafford, North Cheshire and South Cheshire were invited to participate in the study and 50 practices responded positively to the invitation. The second phase involved recruitment of children attending the practices. This was approached in two ways. Children in the age group of interest, who were registered with the practice, were identified from the practice databases and invited to participate in the study and attend for a baseline examination. In addition, any new patients of the correct age who attended the practice during the recruitment period were invited to participate.

The key outcome measure recorded in the study was the development of a new visible cavity (caries into dentine) and recorded in the dental record by the child's dentist in any of the previously caries free primary molar teeth. The presence or absence of caries was recorded at each dental attendance over the three year follow up period. Data were collected from participants clinical records. The dentists followed a protocol for recording data in the clinical case notes of each subject in a standardised way. The dentists were not trained and calibrated to diagnose caries according to a diagnostic protocol. They were asked to record all new cavitated lesions and provide care for each subject according to their usual practice. The clinical records of all subjects were photocopied each year in the dental practice by a research assistant and a trained dental nurse transcribed data from the clinical records to a data abstraction form; the data were then entered into a computer for analyses.

In the analyses the incidence of new cavities was recorded at both the person and tooth levels. Incidence rates for the development of new caries into dentine lesions (cavitation) were calculated for subjects who:

  • Were caries free at recruitment

  • Had caries at recruitment

  • Were caries free at recruitment but developed caries during follow up.

Incidence rates were compiled by first identifying person years of follow up. For those who were caries free for the duration of the study, person years of follow up began at a child's first examination and finished at their last attendance date at the dental practice during the follow up period. For those children who had caries at recruitment or developed the disease during the follow up period, person years of follow up ended when the first new cavity was recorded in the clinical record. Person level incidence rates in each of the three categories of the study cohort (caries free at recruitment, caries at recruitment or caries free at recruitment but developed caries during follow up) were calculated by dividing the total person years of follow up by the number of children that developed a new cavity. At the tooth level two analyses were performed; the first compared the number of carious primary molar teeth in each child at recruitment with the number at the end of the follow up period. This analysis was restricted to children with more than two years of follow up. The second tooth level analysis compared the incidence of new cavity development for each primary molar tooth. Incidence rates were calculated for each primary molar tooth and the rates combined to provide the incidence of new cavities developing in children who were caries free at recruitment and those who had caries at recruitment. A rate ratio, with 95% confidence intervals, was constructed to compare the rates in the caries free and carious populations. This process was repeated to provide the incidence of new cavities developing in children who, although caries free at recruitment, subsequently contracted the disease.

Results

A total of 739 children were recruited to the cohort. Their age range was 2.8-6.2 years at recruitment. Four children were aged below 3 years and three were older than 6 years. These children were included in the 3 and 5-year-old cohorts respectively. Therefore, at recruitment, the numbers in each birth cohort were 3 years 207, 4 years 367, and 5 years 165. Table 1 shows the numbers and proportions of children who had caries in their primary molar teeth at recruitment. In total, 620 (84%) children were caries free at recruitment and 119 (16%) had caries. Within each birth cohort there was a considerable range in the number of carious teeth recorded, but the proportion of children caries free was lowest in the oldest birth cohort. All 620 children who were caries free at recruitment were included in the analysis of the incidence of developing a first cavity in a primary molar tooth, but 31 of these children did not attend the dentist after recruitment and therefore contributed no person years or events (developing a first cavity) to the analysis. Of the remaining 589 subjects 157 (26.7%) developed caries in their primary molar teeth. The age specific incidence of developing the first cavity in previously caries free primary molar teeth is shown in Table 2. These data show that the incidence of developing caries increases with age, for example for children aged between 4-4.99 the incidence of the first carious lesion was 9.5 per 100 person years and for children aged between 7 and 7.99 the incidence was 19.6 per 100 person years.

Table 1 Number and proportion of children in each birth cohort categorised according to the number of carious primary molar teeth recorded at recruitment
Full size table
Table 2 Age specific incidence of developing a first cavity in children recorded as caries free at recruitment
Full size table

During the three year follow up period, 42 subjects did not attend after their initial examination. The dental attendance patterns of the remaining children meant that 131 children had less than two years of follow up. Table 3 describes the number and percentage of children in the 566 children with more than two years of follow up who developed a new cavity in their previously caries free primary molar teeth according to the number of carious teeth they had at recruitment. Of the 486 children who were caries free at recruitment, 27 percent (N = 132) developed the disease. By contrast, in the 80 children with caries at recruitment (one child had eight carious primary molars at recruitment), 57 (72%) developed new cavities in their primary molar teeth. It is interesting to note that high proportions of children who had caries on recruitment developed new cavities irrespective of whether they had one tooth or multiple teeth affected by the disease.

Table 3 Number and percentage of children developing new cavities in previously caries-free teeth according to the number of carious primary molar teeth recorded at recruitment in 565 children who were followed for at least two years
Full size table

The incidence of developing a new cavity in each individual primary molar tooth for subjects who were caries free, had caries, or were initially caries free at recruitment but contracted the disease during follow up is presented in Table 4. In children with caries at recruitment, a total of 168 caries free teeth developed a new cavity in 1195.7 tooth years; a crude rate of 14.05 (95% confidence interval 12.0 to 16.0) per 100 tooth years. This means that generally speaking, approximately 1 in 7 children in this group will develop a new cavity each year. In contrast, for those children who were recorded as being caries free at recruitment a total of 265 teeth developed caries in 10,997.2 tooth years; a crude rate of 2.41 (95% confidence interval 2.1 to 2.7) per 100 tooth years. Therefore 1 in 42 of these children is likely to develop a new cavity each year. The rate ratio of caries development in caries free teeth in children with caries at recruitment and children who were caries free at recruitment was 5.83 (95% confidence interval 4.77 to 7.10). For those children who were initially caries free at recruitment but developed the disease during follow up, 103 caries free teeth developed caries in 892.6 tooth years; a rate ratio of 11.54 (95% confidence interval 9.4 to 14.0) per 100 tooth years.

Table 4 Specific incidence rates for developing new carious lesions in each primary molar teeth in children who had caries and were caries free at recruitment and children who developed caries during follow up
Full size table

Discussion

This study presents data on progression of dental caries in the primary molar teeth in a cohort of children attending general dental practices. The study concentrated on the primary molar teeth as previous studies demonstrate that pain and infection are much more likely to result from caries in the primary molar teeth than anterior primary teeth.1 The study relied on dentists recording in the clinical notes primary molar teeth that were cavitated, which the dentists subsequently observed and/or treated. Although the dentists involved in the study were not calibrated in their diagnoses we do not believe interexaminer variation in diagnosis had a large influence on the results of the study, for a number of reasons. First, interexaminer variability is important in conventional dental epidemiological studies, for example to measure the prevalence of caries in a population. However, this study was primary care-based and is more concerned with the real-life consequences of patients contracting the disease and its progression from tooth to tooth, both for patients and for GDPs who have to manage their care. Therefore the diagnostic decisions and recording of disease by GDPs directly affect the care received by patients, unlike the diagnostic decisions made by dental epidemiologists conducting surveys.

All of the lesions recorded were recorded by the dentists as cavities in the clinical notes, so were by definition caries into dentine and of clinical significance. This recording of late stage disease means that inter-examiner variability among the multiple dentists involved in the study is likely to be minimal, as the level of diagnosis was at a gross level. One could argue that some dentists would routinely use bitewing radiographs and identify interproximal caries at an earlier stage than other colleagues. Previous studies have reported that bitewing radiographs are used infrequently by GDPs in the care of young children, for example in a questionnaire survey of GDPs in Scotland14 only 17% of practitioners would consider taking bitewing radiographs in children under 6 years, and the majority of dentists (61%) reported that they would first consider taking radiographs when children were 6-11-years-old. In our study only two (0.3%) children had bitewings taken during the study, therefore virtually all dentists were diagnosing caries visually at the level of cavitation. Finally all of the dentists provided care for the children under the same NHS capitation-based contract; a type of contract which has been criticised in the past for encouraging under-treatment,15,16 so a treatment effect, due to over-zealous restorative care, was also unlikely. For these reasons we believe that inter-dentist variation in diagnosis had a minimal effect on the results and the implications of this study.

This study describes the way that the clinically recorded disease behaves over time in those children who presented caries free and those who had already contracted the disease prior to their first attendance. The findings are important, as they provide information to help general practitioners manage the care of their young patients, who present with the disease and who are caries free. It is important to recognise that the majority of children attending dental practice present as caries free and remain in that state during the early years of childhood. This finding is consistent with other studies in the literature, which demonstrate low levels of caries in regular dental attenders.17 The study also found that caries incidence increases with age. Again this finding was expected, as serial cross-sectional caries studies have shown higher caries experience and prevalence in older populations.17,18 The most important result for clinical practice and for informing the development of strategies for the care of children was that, although the majority of children are caries free on their first attendance at a dental practice, dentists should not assume that all of their caries free patients will remain in this state.

There was a 5-6 times difference in the risk of developing new cavities between those subjects who were caries free and those with caries at their first visit to the dentist. The likelihood of developing new cavities in remaining healthy teeth was not predicted by the extent of caries at study entry, as the risk of developing a new cavity was similar in children with a single carious tooth and children with multiple carious teeth. When those children who were initially caries free but developed the disease during the follow up period were examined, their risk of developing new cavities (once they had contracted the disease) did not differ significantly from those who had caries on recruitment. These two findings strongly suggest that once children contract the disease it progresses at the same rapid rate. It seems that the key development is the emergence of the first carious cavity and once this state has been reached, further cavity development in previously healthy teeth is highly likely.

These findings have profound implications for the management of the dental care of young children in general practice. The difference in risk of developing new cavities between caries free children and those with caries is large, and it means that these two populations should be managed differently. For those children with caries, the risk of developing pain and having extractions is very high2,11 and due to the rapid rate of progression of the disease simply restoring carious teeth will have little effect in slowing the progression of the disease or reducing the risk of adverse outcomes.2,11 Currently we have little hard evidence to tell us if, once started, the progression of disease can be slowed and, most importantly, if the risk of adverse outcomes such as pain, sepsis and extraction can be reduced. This is because trials measuring the effects of preventive interventions have not reported results separately for caries free children and for those with the disease. We need to consider children with the disease in isolation and trials of intensive preventive care are required to determine if the progression and impact of the disease can be reduced.

Once the disease starts, at whatever stage in early childhood, it progresses at a similar rate leading to a very high risk of pain and extraction.11 This means that GDPs cannot ignore prevention in caries free children, who make up the majority of their young patients. High priority must be given to ensure that caries free children remain in this state. One approach is to try to identify the caries free children who are at high risk of developing caries. However, current methods of predicting the caries risk of individual young children are imprecise.19 Until we have tools which can accurately predict which children will develop the disease, we must provide effective preventive care for all children who present caries free. This whole-population approach to prevention is advocated by the recently published Delivering better oral health: an evidence-based toolkit.20 The toolkit sets out an evidence-based approach to prevention for GDPs to follow in the care of their young patients, focusing on interventions of proven effectiveness such as fluoride varnish and family strength fluoride toothpaste. This whole population approach is justified given the low cost of effective interventions and the almost inevitable and severe consequences children face if they contract the disease.

The cohort examined in this primary care based study were children attending dental practices in the North West of the UK and was therefore not representative of all young children in the UK. Similar studies in diverse populations are required to see if the incidence rates described here represent a norm or are specific to this population. In conclusion, this study found that in a population of children that regularly attend a dentist caries incidence increases with age. Once a child contracts the disease it develops very rapidly, producing a 5-6 times difference in the incidence of new cavities between caries free children and children with caries, irrespective of when a child developed the disease. As a consequence children with the disease and without the disease should be considered as two different populations, this has implications for the care strategies applied to each population.