Dental caries in primary and permanent molars in 7-8-year-old schoolchildren evaluated with Caries Assessment Spectrum and Treatment (CAST) index

Dental caries remains a serious problem in many populations worldwide, with a marked increase of the prevalence in several countries during the last decade [1]. A continued surveillance of the dental epidemiological status is necessary. The Decayed, Missed and Fill teeth (DMFT) index, due to the recommendation of the World Health Organization [2] the most commonly used tool in the epidemiological surveys, has failed to meet the challenges of the 21th century. The detection of pre-cavitated lesions is a matter of importance in the populations with a low prevalence of cavities. The application of the International Caries Detection and Assessment System (ICDAS), in which three stages of enamel lesions are distinguished, may be a solution [3]. However, the system requires using compressed air to dry tooth surfaces and double checking of teeth so that the surveys are costly and time-consuming. For the populations with a high prevalence and a severe course of caries, tools like Pulpal Involvement-Ulceration-Fistula-Abscess (PUFA) and Pulpal Involvement-Roots-Sepsis (PRS) for the detection of consequences of untreated dental caries index were proposed [4, 5]. Although PUFA and PRS arouse a great interest, their disadvantage is that they cover only a part of the wide range of caries stages and they only complement the DMFT or ICDAS. From the practical point of view, the most advantageous solution in epidemiological surveys is to use a single index describing the full continuum of a disease.

Recently, an innovative instrument for the epidemiological studies named Caries Assessment Spectrum and Treatment (CAST) was introduced by Frencken at al. [6]. The details concerning the structure of the index were explained in a range of publications [7, 8]. A novelty of CAST is the recommendation to include teeth with dental fillings in the category of sound teeth, which is in line with the epidemiological concept of health. A special attention should be given to a modern way of assessing the face and content validity of the instrument by the RAND modified e-Delphi consensus method, with 56 researchers from 24 countries involved in the process [8]. CAST has been already validated in extensive in vitro and in vivo studies which have proven its high specificity, sensitivity and reliability in epidemiological surveys [9, 10]. However, the CAST index should be tested in other independent surveys in order to become established as a plausible instrument.

Some universal patterns in caries can be observed, for example: caries levels follow trend lines, there is a specific mathematical relationship between the mean DMFT and mean DMFS, and changes in mean DMFT scores for individuals and groups are not linear [11]. The regularities in caries distribution with regard to teeth, sites and groups of sites were also found [12‐16]. The three possible patterns were evaluated: random, aggregated and regular, however, the hypothesis that teeth were randomly infected by caries was rejected. Through the years, the concept of caries regular occurrence, e.g. a symmetrical prevalence with respect to the midline as well as between the upper and lower jaws, was so widely accepted that some measures assessed the level of caries by doubling the results obtained by the examination of a half of the dental arch [11]. However, Vannonberg et al. [16] found that, at the population level, caries had a tendency to symmetrical distribution, but at the individual subject level the cavities rather accumulated on one (left or right) side of the mouth. Batchelor and Sheiham [14] confirmed that a precise symmetry of caries did not occur, but there were groups of teeth with a similar susceptibility. A symmetrical occurrence of lesions resulted in a stepped model of disease levels, i.e. a decline or an increase of caries courses in pairs [11]. The same could be considered for the aggregated pattern. It means that the inclusion of proper preventive measures in the most susceptible groups of sites should result in a substantial caries reduction [14].

Most research studies on the caries pattern were based on a dental evaluation according to DMF [11, 12, 14, 15]. Honcala et al. [17] analyzed the caries distribution and correlation in primary and permanent molar teeth with regard to ICDAS. No reports on a caries pattern covering the full spectrum of the disease could be found in the literature. The aim of this study was to evaluate caries in primary and permanent molars of 7-8-year-old Polish children by the CAST index and to find whether there was any correlation between the caries stages in such teeth.

The unweighted kappa value for the intra-examiner reliability was established at 0.96 for the primary dentition and at 0.878 for permanent teeth. Figure 1 shows the percentage of children according to the highest CAST code per mouth, separately for primary and permanent dentition. With regard to deciduous teeth, a quarter of the subjects showed a pulpal involvement (code 6) and one fifth a dentine cavity (code 5) as the most serious caries stages. For permanent teeth, fissure sealants (34.9%) were most prevalent, followed by enamel lesions (26.4%). None of the children scored the categories 7 and 8 in permanent teeth. No tooth scored the category 9 either.

Table 2 presents the distribution of each CAST code in the evaluated molars. Figure 2 shows the distribution of molar teeth according to different disease stages defined as healthy dentition (codes 0–2), reversible premorbidity stage (code 3), teeth with morbidity (codes 4 and 5) and with serious morbidity (codes 6 and 7), and teeth with mortality (code 8). With regard to the permanent molars, caries was observed in 14.8% to 17.3% of the teeth; however, most lesions were scored at the non-cavitation level (code 3 and 4). About two-thirds of evaluated primary teeth were found to be healthy (codes 0–2). For both, first and second primary molars, caries was most often recorded at the stage of cavitated dentine lesion. A serious morbidity was found to be more prevalent in first, and then in second molars; also extractions were recorded 2–3 times more often in the first than in the second counterparts. The Mann–Whitney U test did not reveal any difference in the distribution of the CAST index in evaluated teeth according to age and sex.

Table 3 shows the results of the repeatedly performed Spearman correlation test where the p-values both without and with the Bonferroni correction are given. The analysis of distribution of the CAST codes in primary second and permanent first molars revealed a strong correlation between caries stages in counterpart teeth from the right and left sides of the oral cavity. Only for the lower first primary molars (84/74) the rank correlation coefficient was lower than 0.5. The correlation of the status of first and second primary teeth was stronger for the left than for the right side of the mouth; r was 0.627 and 0.472 in maxilla and 0.513 and 0.483 in mandible (p < 0.001), respectively. For the neighbouring primary and permanent molars r values were lower than 0.3, which meant a weak correlation. With regard to the teeth situated in opposite jaws the study revealed that the correlations were moderate - r between 0.33 and 0.49. All correlations except the one between second primary and first permanent molars from the left side in maxilla were statistically significant, however, with the Bonferroni correction, the 16/55 correlation was also found to be insignificant.

The contemporary concepts of caries indices are based on the idea of incorporation of all caries stages into one tool. Among many systems, the CAST index stands out with its simple hierarchical structure including the full spectrum of the disease, the categorization of the caries process according to its progression and a modern approach to filled teeth due to their inclusion in the category of sound teeth. CAST is a promising index for epidemiological research studies because the instrument allows obtaining more detailed data on caries prevalence and experience than DMF. Moreover, its use during a survey should be less costly and time-consuming compared to the use of ICDAS, however, such advantages need to be proven in further studies. So far, the reproducibility of CAST in clinical studies was assessed to be from substantial to almost perfect depending on the age of participants [10]. Our study, presenting good intra-examiner reliability in both sets of dentition, proved that CAST could be used in epidemiological studies. The lower kappa value for permanent molars than for primary dentition obtained by us is in accordance with the findings of de Souza et al. [10]. They stated that the level of reproducibility might be influenced by a low prevalence of particular CAST codes, e.g. in the permanent teeth soon after the eruption.

Our findings about a high prevalence of caries, particularly in deciduous teeth, in Polish children are in agreement with the previously reported data [18‐21]. 7-year-old children were involved in the Polish National Oral Health Survey in 2011 [20]; 8-year-olds were never a part of such survey. In 2011, only 10.2% of surveyed 7-year-olds in Poland had caries-free primary teeth. The mean d₃mft was estimated at 4.45 and mean D₃MFT at 0.42. In our study, the percentage of children found to be caries-free with regard to primary teeth was 7%, however, this proportion would increase to 15.6% if the enamel lesions were excluded from the analysis. For permanent teeth, the percentage of caries-free subjects would be greater for one quarter taking the level of cavitation into dentine as the disease threshold. We observed that the pulpal involvement, the category involving a cavity reaching the pulp or the presence of root fragments, was found to be the most serious stage in 26.4% (primary teeth) and 0.3% (permanent teeth) of the subjects. In 2011, as much as 41.9% of 7-year-old Polish children needed tooth extraction, and 16.4% of them should have had an endodontic treatment. The neglects in dental treatment with regard to the deciduous dentition have been observed worldwide [22‐26]. It was previously proven that the dmft level positively correlated with the number of teeth scored with the pufa (pulpal involvement-ulceration-fistula-abscess) index assessing the consequences of untreated dental caries [5].

We decided to primarily concentrate on the correlations between the status of molars because of the considerable dynamics of front teeth exchange in children at the age of 7–8 years. The exclusion of incisors and canines from the analysis allowed us to keep the homogeneity of the study population [27]. A similar approach was adopted by other authors [17, 27]. We observed that the percentage of teeth with enamel lesions was at a similar level for second primary and first permanent molars, but with regard to first primary molars the prevalence of code 3 was lower. The tendency that cavitated lesions were more prevalent in primary than in permanent molars was very clear. The presented results are in accordance with the study of Honcala et al. [17] on Estonian children aged 7 and 8 years who assessed molar teeth by the ICDAS criteria. The enamel lesions visible on wet teeth (ICDAS code 2), located on occulsal surfaces of first permanent molars (up to 17% of the teeth) were most prevalent in their study. The highest percentage of teeth with dentine lesions in Estonian children was observed for lower second primary molars. The most probable explanation is that at the age of 7-8-years the factors causing dental caries act too short to induce the development of deep cavities in permanent teeth. Primary teeth are also more prone to a faster lesion progression from enamel to dentine and then to the development of pulpitis due to a lower thickness and a relatively larger pulp chamber in comparison to permanent teeth [28]. According to Sheiham and Sabbah [11], the rate of progression of caries through the enamel and into the dentine should determine the frequency of dental recalls.

We found that a large proportion of first and second primary molars with caries lesions at different stages of progression remained untreated. In the present study, the percentage of molars with a serious morbidity (involvement of pulp and tooth surrounding tissues) was especially high for first primary molars, and these teeth also showed the highest tooth mortality (CAST code 8). This observation is in contrast to many previous studies where second primary molars were reported to be more affected than first ones [25, 29, 30].

The high prevalence of enamel caries in the permanent molars, reaching 15.5% for tooth 36, puts the evaluated Polish children in a group of a high risk of future caries occurrence. The presence of pre-cavitated lesions is a predisposing factor for the cavity development [31]. Masood et al. [32] suggested that obligatory screenings should be started at the age of 6 years and followed for a certain period of time in order to select individuals with the highest caries risk. Another issue increasing the risk of caries in this population is a low percentage of pit and fissure sealants; sealants in permanent teeth were the highest score per mouth for one third of evaluated children (34.9%) only. Occlusal surfaces of permanent molars and buccal pits of lower molars are most prone to the development of caries lesions [14]. Sealants are strongly recommended in the high risk populations [33, 34], however, dentists differ in clinical decisions on the indications to seal pits and fissures, on the chosen technique and on the material used as a sealant [35]. The third factor predisposing to caries development in this group is a bad condition of primary teeth. Caries in primary molars is a well known predictive factor for the development of cavities in the permanent dentition, particularly in the first molar teeth [36, 37]. Steiner at al. [36] found that a low number of sound primary molars at the age of 7 and 8 years constituted the best and most consistent predictor of a high caries increment in the permanent dentition. According to Gray at al. [36], the presence of three or more deciduous molars at the age of five was the best predictor of caries experience in the first permanent molars at the age of 7 years. Skeie et al. [37] established the level of two surfaces with caries in primary second molars at the age of 5 years as a predictor of caries development during the period of next 5 years.

The present study is one of few reports on the caries pattern distinguishing the disease stages. The symmetry of caries distribution was not previously assessed with regard to the full spectrum of severity of the caries process. We found a strong correlation between the conditions of contralateral molars measured by the CAST index in both, primary and permanent, dentitions. A certain degree of symmetry in caries distribution was also observed for teeth in upper and lower jaws, which is in accordance with the previous reports [11, 15]. In our study, the correlations between CAST categories found in neighbouring primary molars were stronger for the left side of the mouth, both in a maxilla and a mandible, which might prove the theory about the accumulative caries pattern [12]. The correlations between caries stages found in first permanent and second primary molars were weak, similarly to the findings of Honcala et al. [17] about the distribution of ICDAS codes on particular surfaces of neighbouring primary and permanent molars. The probable reason is a low number of permanent teeth categorized with CAST codes 5 to 8. A prospective study is needed to assess whether the poor status of deciduous molars influenced the condition of permanent teeth in this population. The lack of such evaluation is undoubtedly one of the limitations of the present report. However, the CAST index is a relatively new research tool and it was impossible to conduct a longitudinal study.

The awareness and understanding of the regularities in the caries pattern should make clinical practitioners more accurate during a dental examination [16]. The practical implication of the described caries pattern is not only that the prevalence of caries lesion in one tooth can be treated as a predictive factor of the caries presence in other teeth from the group of similar susceptibility, but also that a severe caries in one tooth increases the risk of the development of deep cavities and further caries consequences in other teeth.

The strongest correlation in the evaluated population was found for the distribution of caries stages in primary molars on the left side of the mouth. The study proved the usefulness of the CAST index in epidemiological surveys.