ART class II restoration loss in primary molars: re-restoration or not?
Abstract: AIM: The purpose of this study was to find an answer as to what to do with Atraumatic Restorations (ART) failures: re-restore or leave the preparation further unfilled? STUDY DESIGN: Cross sectional study. METHODS: In 2006, 804 children in Kenya each had one proximal cavity treated using the ART approach. Out of the original group of 192 children, who had lost their restorations but still had the treated molars in situ, were selected for further study in 2008. The length of time that the restorations had been in situ was known while the colour, hardness and the extent of infected dentine was then evaluated and documented. STATISTICS: Analysis of the data obtained was conducted using SPSS 16.0. Chi Square tests were performed with the variables of hardness, colour and infected dentine, and a 5% confidence interval was used. The Spearman's Rank Correlation Coefficient was also calculated. RESULTS: The results showed that 66% of the molars that had lost restorations had hard dentine, 78% of the preparations showed dark dentine and 50.7% appeared to have no infected dentine. These percentages increased with the increase in the survival time of the restorations. CONCLUSIONS: It is not always necessary to re-restore primary molars after ART restoration loss. Further research is necessary to confirm these findings.

Key words: ART, remineralisation, Class II, developing countries
Article Type: Report
Subject: Pit and fissure sealants (Dentistry) (Usage)
Pit and fissure sealants (Dentistry) (Health aspects)
Molars (Health aspects)
Molars (Research)
Authors: Boon, C.P.J.M.
Visser, N.L.
Kemoli, A.M.
van Amerongen, W.E.
Pub Date: 10/01/2010
Publication: Name: European Archives of Paediatric Dentistry Publisher: European Academy of Paediatric Dentistry Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2010 European Academy of Paediatric Dentistry ISSN: 1818-6300
Issue: Date: Oct, 2010 Source Volume: 11 Source Issue: 5
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: Netherlands Geographic Code: 4EUNE Netherlands
Accession Number: 276898052
Full Text: Introduction

The Atraumatic Restorative Treatment (ART) approach was developed to meet the need for dental restorative treatment in deprived areas that was neither electricity nor running water [Frencken and Van Amerongen, 2008]. ART consists of caries removal exclusively using hand instruments combined with restoring a tooth and sealing the adjacent fissures with an adhesive restorative material [Frencken et al., 1996; Smales and Yip, 2002]. Glass ionomer cement (GIC) is the most commonly used restorative material with the ART technique. The advantages of using GIC are based on its properties such as the release of fluoride, chemical curing, simplicity, biocompatibility with the dental pulp, chemical adhesion to tooth substance and a good cost-effectiveness ratio [Phantumvanit et al., 1995].

In general, many ART restorations do fail, particularly as a result of their size or the way the materials are handled. The success rates for Class I and II restorations are different. Short- term studies (evaluation after one year) indicate success rates of approximately 80% to 95% for Class I and 30% to 75% for Class II restorations. Long term studies (evaluation after more than 2.5 years) showed success rates of approximately 43.4% to 86.1% for Class I and 12.2% to 82.1% for Class II restorations [Luo et al., 1999; Holmgren, 2001; Yip., 2002; Taifour et al., 2002a; Ersin et al., 2006; Roeleveld et al., 2006; Van Gemert- Schriks et al., 2007; Van de Hoef and Van Amerongen, 2007].

The causes of the failure can be multi-factorial. The influence of the operator's techniques are important such as inadequate caries removal, mixing of GIC, and the application of the restoration material are all factors that can influence the survival rate [Frencken et al., 1996; Mallow et al., 1998; Rahimtoola and Van Amerongen, 2002; Taifour et al., 2002b; Frencken et al., 2004; Van Gemert-Schriks et al., 2007]. Previous researchers in relation to Class II restorations have reported failure of these restorations being due to the mechanical strength of GIC that is not able to resist the occlusal forces exerted on them [Frencken and Holmgren, 1999]. Teeth that have lost their restorations show a difference in hardness and colour of the dentine remaining behind [Santiago et al., 2005; Pereira et al., 1998]. There is no answer, however, given to the question as to whether the failed ART restorations have to be repaired or re-restored.

Ten Cate and Duinen [1995] found that teeth with GIC restorations exhibited hypermineralisation in the dentinal tissue bordering the filling. The (caries-like) lesions were hypermineralised, even under conditions of heavy plaque formation. The in vivo process under a GIC restoration is as yet unknown. The results of the research by Weerheijm et al. [1993] reported a reduction of microorganisms could be expected under the GIC restorations. The GIC can potentially have a hypermineralising influence on the carious dentine underneath the restoration. For hypermineralisation to take place calcium and phosphate ions are essential [ten Cate et al., 1995]. These ions could potentially originate from the serum of the dentine tubules, the micro leakage from the restoration, or might still be present in the carious tissue. It is probable that a combination of these factors is likely to determine the behaviour of carious dentine [Van Amerongen, 1996].

The purpose of the present study was to define the potential re-treatment need of failed ART restorations, such as to whether to re-restore or to just leave the preparation for a failed restoration unfilled. This would, of course, depend on the degree of arrestment of the caries. The hypothesis of this study was that the time span of the presence of GIC restorative material in a tooth-cavity has a positive influence on the arrestment of the caries.

Material and Methods

Pre study situation: In 2006 a longitudinal clinical study was started in Kenya to investigate the influencing factors of the survival rate of proximal ART-restorations placed in primary molars. This study took place in Matungulu/Kangundo divisions in Machakos, Kenya, which is located about 60km southeast of Nairobi. Patients in the age group between 6 and 8 years with at least one Class II cavity in a primary molar were selected [Kemoli and Van Amerongen, 2009].

A cohort of 804 children was randomly recruited into the study. An informed consent was obtained from the parents or guardians and ethical approval given by the local university Ethical Committee. The children were randomly placed into one of three treatment groups. Each group was treated with a different GIC restorative material. Within each group a division into two subgroups was made. Again, based on a random selection each patient was allocated to one of two isolation methods i.e. by rubber dam or cotton wool rolls [Kemoli and Van Amerongen, 2010]. Finally, these subgroups were randomly assigned to an experienced or inexperienced operator. Three experienced and four inexperienced operators participated in the study. Four experienced and four inexperienced assistants assisted them.

During the restorative stage, all cavities were prepared with hand instruments, aided by a caries detector-dye (private label based on acid red by ACTA, the Netherlands) in the removal of the carious material. The cleaned cavity, isolated with cotton wool rolls or rubber dam, was restored with Fuji IX (GC Europe), Ketac Molar Easymix (KME) or Ketac Molar Aplicap (KMA) (3M ESPE AG products). The child was asked not to eat within the first hour after restoration.

After placement of the restorations, evaluations were completed immediately after treatment (within 2 hours). Further evaluations and scoring of tooth condition were made after one week, one month, five months, one year, one and a half years and finally at 2 years.

Procedure for the study: At the final evaluation, (after two years post-placement of the proximal ART restorations), the children were re-examined. The oral health of these children was also determined and documented by scoring the plaque using the method described by Greene and Vermillion [Greene and Vermillion, 1964].

During the final evaluation of the main study, 779 children were examined. Children who had lost an ART-restoration and whose treated molar was still in situ were selected for the present sub-study group (N=192). Finally 64 children were afterwards excluded from the research because the data from the previous evaluation results were missing (13 molars) or because of inability to score the dentine due to pulp exposure resulting in a disappearance of the roof of the pulp chamber (51 molars). The remaining 128 patients were, therefore, included in the study.

Hardness, colour and measure of infection of the dentine. Food debris in the treated proximal cavities of the molars was carefully removed with a probe (Dentsply Maillefer) and the cavity surface of the molar was cleaned with one wet-cotton pellet and dried with two cotton pellets. This process exposed the state of a tooth where an ART restoration had been lost showing as coloured and possibly harden dentine (Fig 1.).


The hardness of the dentine was scored by scraping with a probe using a 3 point scale: 1 = hard, 2 = medium, 3 = soft. [Kidd, 1993]. A colour scale, to determine the colour of the dentine and the measure of infection, was made out of 12 wooden chopsticks by creating three sticks in each of the defined four dentine colours: cream-white, light brown, brown and dark brown (Talens Amsterdam Acryl) (Fig 2). Subsequently, the tips of the three sticks within each colour group were dipped in three different solutions of caries detector dye private label based on acid red by ACTA, (the Netherlands) diluted with water: in a ratio of 1:40, 1:20 and one solution consisting of only caries detector. To make sure the colours would not fade and the sticks could be cleaned with alcohol, the sticks were covered with a thin layer of varnish (Sikkens, the Netherlands).


The colour of the dentine (as exemplified in Fig 3) was scored at a four-point scale [Weerheijm et al., 1993]. 1 = cream-white, 2 = light brown, 3 = brown, 4 = dark brown. To score the measure of infection of the dentine, caries detector on a cotton pellet was applied for 5-10 seconds in the preparation. The dentine was rinsed with one wet cotton pellet and dried with two cotton pellets. The three sticks with the corresponding colour of the dentine that were prepared with the three concentrations of caries detector were used to score the measure of infection, based on the following criteria: 1=none or limited infected, 2= moderate infected, 3= infected. (1:40, 1:20, only caries detector).


The two first authors, both dentists, who had trained in the examination technique, together evaluated the colour, the hardness and the measure of infection, by coming to three consensus scores for each child. During the observations the dentists did not know the retention time of the restoration.

Statistical analysis: Analysis of the data obtained was conducted using SPSS 16.0. Chi Square tests were performed with the variables, and a 5% confidence interval was used. Also, the Spearman's Rank Correlation Coefficient was calculated.


The results shown in Table 1 where the time a tooth was without a restoration is recorded. Correspondingly the restorations in 74 molars failed during the first 6 months (57.8%) and a tooth was therefore without a restoration for between 12-18 months. This occurred for 28 molars between 6-12 months and 26 molars between 12-18 months. Of the molars that had lost their restoration within the first 0-6 months, 59.5% had scored hard dentine, 81.1% scored dark dentine and 45.9% scored non-infected dentine. When a restoration had remained in place for 12-18 months then 92.3% scored hard dentine, 73.1% scored dark dentin and 61.5% scored non-infected dentine. No abcesses or draining fistulas were found during the examinations.

Data analysis indicated that there were only very small differences between the four dentine colour variables. Therefore, the decision was made to combine the cream-white and light brown categories into 'light dentine' and the brown and dark brown categories into 'dark dentine'. When the analysis was repeated using the new category, there was a significant correlation between the dentine hardness and the retention time of the GIC material (Spearman: 0.22, p<0.05). When the period of time that a cavity was without a restoration increased a greater amount of soften dentine was scored.

Of those children that participated in the study, 55.8% were found to have a moderate level of oral hygiene (6-12 parts) and 41.9% scored poor on this criterion (>12 parts). No correlation was found between oral hygiene and hardness, colour and infected dentine and between oral hygiene and retention time (p=0.277, p=0.169, p=0.570 and p=0.837 respectively). The hardness of the dentine was also not correlated with the restoration material that had been used, experience of the operator and assistant. (Spearman, p=0.360, p=0.604 and p=0.595 respectively).


Of the restored molars 51 could not be scored due to pulp exposure resulting in the disappearance of the roof of the pulp chamber. No abcesses or draining fistulas were found. The inability to score these molars could influence the results in a positive way, because these molars could have been scored as soft dentine.

A considerable number of the children in the study group were over 9 years old and already were in the late mixed dentition stage. This could have also influenced the results in a positive way because it was unknown whether the teeth were absent because of extraction due to caries or had been exfoliated. Based on the relatively higher age and mixed dentition factor, the DMFT would not reflect a proper indication of the real caries prevalence.

In the case of the scoring of the plaque no correlation was found between oral hygiene on the one hand and the hardness, colour, infected dentine and retention time on the other hand. This showed that oral hygiene apparently had no influence on the results. Scraping with a probe scored hardness of dentine and this measurement was subjective, but given the circumstances it was the only way to determine hardness of the dentine. It was not possible to positively conclude that remineralisation had occurred. The hardness of the dentine can be caused by the long retention time of the restoration or by the short time the restoration had been out of the tooth. However, the hardness of the dentine seemed to be a better standard to use than colour or the presence of the infected dentine.


According to the clinical criteria, re-restoration of primary molars may not be necessary in all cases when restored with ART This study has shown that the focus should be on the survival of the restored tooth rather than on the survival of the restoration (as in many previous studies). The treatment can be called a success when the tooth has remained painless until exfoliation. Further research is necessary to confirm and specify these findings. In future research it is essential to create a control group in which only caries will be removed using the ART technique while the molars are left unfilled. As a result, the influence of glass ionomer cements on the dentine can be demonstrated more effectively.


Amaral MT, Guedes-Pinto AC, Chevitarese O Effects of a glass-ionomer cement on the remineralisation of occlusal caries- an in situ study. Braz Oral Res. Apr-Jun; 2006;20(2):91-6.

Axelsson P. An introduction to risk prediction and preventive dentistry. Quintessence Publishing Co Inc, Karlstad, Sweden. 2000.pp 77-97.

Benelli EM, Serra MC, Rodrigues Jr AL, Cury JA. In situ Anticariogenic Potential of Glass Ionomer Cement. Caries Res; 1993; 27:280-284.

Ersin NK, Candan U, Aykut A, et al. A clinical evaluation of resin-based composite and glass ionomer cement restorations placed in primary teeth using the ART approach. Results at 24 months. J Amer Dent Assoc, 2006 Vol. 137, November, 1529-1536.

Frencken JE, Pilot T, Songpaisan Y, et al. Atraumatic Restorative Treatment (ART): Rationale, technique, en Development. Journal of Public Health Dent 1996; Vol. 56, No. 3, Special Issue.

Frencken JE, Homgren CJ. Atraumatic Restorative Treatment for dental caries. STI book Bv, Nijmegen. 1999.pp 76-77.

Frencken JE, van 't Hof MA, van Amerongen WE, et al. Effectiveness of single surface ART restorations in the permanent dentition: a meta-analysis. J Dent Res 2004;83:120-123.

Frencken JE, van Amerongen WE in Fejerskov O, Kidd E. Dental Caries. Blackwell Munksgaard, Oxford. 2008. Chapter 23, pp 427-440.

Gao W, Smales RJ, Yip HK. Demineralisation and remineralisation of dentine caries and the role of glass-ionomer cements. Int Dent J. 2000; Feb;50(1):51-6.

Greene JC, Vermillion JR. The simplified oral hygiene index. J Am Dent Assoc. Jan 1964; 68:7-13.

Hatibovic-Kofman S, Suljak JP, Koch G. Remineralization of natural carious lesions with a glass ionomer cement. Swed Dent J 1997; 21(1-2):11-7.

Kemoli AM, van Amerongen WE. Influence of the cavity-size on the survival rate of proximal ART restorations in primary molars. Int J Paediatr Dent. 2009; Nov;19(6):423-30.

Kemoli AM, van Amerongen WE, Opinya GN. Short Communication: Influence of different isolation methods on the survival of proximal ART restorations in primary molars after two years. Eur Arch Paediatr Dent. 2010; Jun;11(3):136-9.

Kidd EAM, Joyston-Bechal S, Beighton D. Microbiological Validation of Assessments of Caries Activity during Cavity Preparation. Caries Res 1993;27:402-408.

Lo ECM, Holmgren CJ. Provision of atraumatic restorative treatment (ART) restorations to Chinese pre-school children. A 30-month evaluation. Int J Paediatr Dent 2001; 11:3-10.

Luo Y, Wei SH, Fan MW, et al. Clinical investigation of a high-strength glass ionomer restorative used with the ART technique in Wuhan, China: One-year results. Chin J Dent Res 1999; 2:73-78

Mallow PK, Durward CS, Klaipo M. Restoration of permanent teeth in young rural children in Cambodia using the ART technique and Fuji II Glass Ionomer. Int J. Paediatr Dent 1998; 8:35-40.

Penning Ch , Van Amerongen JP, De Kloet HJ, et al. Caries lesions (Carieslaesies), Prelum uitgevers, Houten, 2007.pp 323-347.

Pereira PNR, Inokoshi S, Yamada T, et al. Microhardness of in vitro caries inhibition zone adjacent to conventional and resin-modified glass ionomer cements. Dent Mater 1998; 14:179-185, June.

Phantumvanit P, Songpaisan Y, Frencken JE, et al. Atraumatic restorative treatment (ART): Evaluation after 2 years. 1995.P.53 WCPD'95 Official Program.

Rahimtoola S, van Amerongen WE. Comparison of two tooth saving preparation techniques for one surface cavities. J Dent Child 2002; 69:16-26.

Roeleveld AC, van Amerongen WE, Mandari GJ. Influence of residual caries and cervical gaps on the survival rate of Class II glass ionomer restorations. Eur Arch Paediatr Dent. 2006; Jun;7(2): 85-91.

Santiago BM, Ventin DA, Primo LG, et al. Microhardness of dentine underlying ART restorations in primary molars: an in vivo pilot study. Br Dent J. 2005; 199 No. 2 July 23.

Smales RJ, Yip HK. The Atraumatic Restorative Treatment (ART) approach for the management of dental caries. Quintessence Int. 2002; 33,(6) 427-32.

Taifour D, Frencken JE, Beiruti N, et al. Effectiveness of the ART approach in a child population in Syria. (Doctoral thesis, December 16, 2002a) University of Njmegen.

Taifour D, Frencken JE, Beiruti N, et al. Effectiveness of Glass-ionomer sealents in newly erupted first molars after 5 years: a pilot study. Community Dent Oral Epidemiol 2002b; 31:314-319.

ten Cate JM, van Duinen RN. Hypermineralization of Dentinal lesions adjacent to Glass-ionomer Cement restorations. J Dent Res 1995; 74(6): 1266-1271, June.

ten Cate JM, Buijs MJ, Damen JJM. The effects of GIC restorations on enamel and dentin demineralization en remineralization. Adv Dent Res 1995; 9(4): 384-388, December.

van Amerongen WE. Dental caries under glass ionomer restorations. J Public Health Dent 1996; 56(3): 150-4.

van Duinen RN, van Strijp AJ, ten Cate JM. Dentin remineralisation induced by glass ionomers. Ned Tijdschr Tandheelkd 1992; May; 99(5):187-9.

van Gemert-Schriks MCM, van Amerongen WE, ten Cate JM, et al. Three year survival of single- and two-surface ART restorations in a high-caries child population. Clin Oral Invest 2007; 11:337-343.

van de Hoef N, van Amerongen WE. Influence of local anaesthesia on the quality of Class II glass ionomer restorations. Int J Paediatr Dent 2007; 17:239-247.

Weerheijm KL, de Soet JJ, van Amerongen WE, et al The effect of glass-ionomer cement on carious dentine: An in vivo study. Caries Res 1993; 27:417-423.

Yip HK. Comparison of ART and conventional cavity preparations for glass-ionomer cement restorations in primary molars: 12-month results. Quintessence Int 2002; 33:17-21.

C.P.J.M. Boon, N.L. Visser, A.M. Kemoli, W.E. van Amerongen, Dept. Cariology Endodontology Pedodontology, Academic Centre for Dentistry Amsterdam (ACTA), The Netherlands.

Postal address: Dr. C.P.J.M. Boon, Dept. Cariology Endodontology Pedodontology, ACTA, Gustav Mahlerlaan 3004, 1081 LA Amsterdam, The Netherlands

Table 1. Characteristics of dentine at different times
in a study on lost restorations after using ART.

                           % Time the preparation was
                              without a restoration

                    N    0-6 mths   6-12 mths   12-18 mths

N                  128      26         28           74
soft                4      3.8%        --          4.1%
-middle            39      3.8%       39.3%       36.5%
hard               85     92.3%       60.7%       59.5%
light              28     26.9%        25%        18.9%
dark               100    73.1%        75%        81.1%
Infected dentine
none or limited    65     61.5%       53.6%       45.9%
moderate           36     23.1%       28.6%       29.7%
infected           27     15.4%       17.9%       24.3%
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.