Durability of amalgam in the restoration of class II cavities in primary molars: a systematic review of the literature.
|Article Type:||Clinical report|
(Care and treatment)
Dental caries (Patient outcomes)
Dental amalgams (Usage)
Dental amalgams (Research)
|Publication:||Name: European Archives of Paediatric Dentistry Publisher: European Academy of Paediatric Dentistry Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2007 European Academy of Paediatric Dentistry ISSN: 1818-6300|
|Issue:||Date: March, 2007 Source Volume: 8 Source Issue: 1|
|Topic:||Event Code: 310 Science & research|
|Geographic:||Geographic Scope: Australia Geographic Code: 8AUST Australia|
Aim: To review the available literature on the durability of amalgam when used to restore interproximal (class II) cavities in primary molars. Methods: The literature was searched using OVID Medline and EMBASE from 1966 to 2006. The search plan of the electronic databases included: "dental amalgam or amalgam or alloy" and "deciduous or primary or milk or first or baby or natal" or "tooth or teeth or dentition" and "permanent restorations or permanent fillings". Relevant prospective clinical studies were reviewed by 2 reviewers against a set of defined criteria. Papers were graded according to the number of criteria met as A = >90%, B1 = 75%, B2 = 50% and C = <50%. Results:: No study achieved a Grade A; four studies rated Grade B1 and thirteen B2. Seven studies were rated Grade C. A wide range of failure rates for amalgam were reported from 0 - 58%. However, in the context of a controlled clinical environment in a developed country the failure rate varied between 0 and 22%. Conclusion: Amalgam remains an appropriate choice of material for the restoration of the primary dentition. However, factors other than durability are increasingly influencing its use in clinical practice.
Key words: Dental amalgam, primary teeth, restorations
Dental amalgam (AM) has been the cornerstone of restorative dentistry for over 150 years. However, the introduction of aesthetic, adhesive restorative materials has provided clinicians and families with choices which have led to changes in the pattern of use of restorative materials particularly in the primary dentition. Concerns regarding possibly toxicity of mercury and AM coupled with environmental issues have led some authorities to recommend that it no longer be used in the primary dentition [NHMRC, 1999]. Whilst neither the British Society of Paediatric Dentistry [Rugg-Gunn et al., 2001] nor its equivalent the American Academy of Pediatric Dentistry [AAPD, 2005-6] support the move away from the use of AM, the reality appears to be that its use in the primary dentition is declining [Milsom et al., 2002; Tran and Messer, 2003; Roshan et al., 2003] although for some clinicians it remains the material for choice [Gordon et al., 2005].
For many dentists, AM is the benchmark against which alternatives should be compared, although it is certainly not a panacea as it is prone to fracture in thin section; this, together with the need to incorporate mechanical retention into the cavity design, often necessitates the removal of sound tooth tissue during cavity preparation. These factors when added to parental preference [Peretz and Ram, 2002] may lead one to choose an alternative restorative technique. Nevertheless, it is a relatively technique insensitive material particularly with respect to moisture contamination which may be advantageous when treating young patients. So what is the evidence that dental amalgam is a durable and effective restorative material to use in the primary dentition?
Much of the evidence available on its effectiveness in the primary dentition is derived from retrospective studies of patient records [Braff, 1975; Llewelyn, 1977; Dawson et al., 1981; Holland et al., 1986; Levering and Messer, 1988; Qvist et al., 1986; Papathanasiou et al., 1994]. Failure rates as high as 70 to 89% have been reported [Braff, 1975; Dawson et al., 1981] with longevity related to the age of the patient at the time of treatment [Holland et al., 1986; Levering and Messer, 1988].
Holland and co-workers  reviewed 1,139 AM restorations placed in primary molars in a dental hospital and analysed the data using a life survival analysis technique [Peto et al., 1977]. This technique allowed comparison of two treatment modalities over the duration of the study period rather than at specific time intervals thereby allowing subjects to be recruited at different times, producing a value known as the Median Survival Time [Mitchell and Walls, 1991]. The Median Survival Time (MST) reported for class II amalgam restorations placed in three year olds was 11 months which rose to 44 months in 7 to 8 year olds [Holland et al., 1986]. This relationship was supported by a study in the USA [Levering and Messer, 1988] which analysed the case histories of 1,898 class I and II amalgam restorations placed by dental students. On the basis of their findings the authors predicted that 51% of class II amalgams placed in children under the age of 4 would survive 5 years, whilst this prediction rose to 70% of those placed in children over this age.
In the past, clinicians have based their decision making on their own clinical experience and the results of these retrospective studies. However such studies are often, by nature, poorly controlled and rely on accurate record keeping. As a result they are associated with problems of incomplete and inaccurate data. With the increasing emphasis on evidence based clinical practice, clinical choices should be based upon on evidence from well designed, carefully controlled randomised prospective clinical trials. This paper aims to review, in a systematic fashion, the evidence surrounding the effectiveness of dental amalgam in children.
* dental amalgam,
* deciduous teeth,
* permanent dental restoration.
The search plan of the electronic databases was:
* dental amalgam or amalgam or alloy,
* deciduous or primary or milk or first or baby or natal tooth or teeth or dentition,
* permanent restorations or permanent filling.
The literature search was limited to articles published in the English Language and the study population to children up to the age of 18 years.
Inclusion criteria: Studies were included if they presented data from prospective clinical studies, involving the use of conventional AM in interproximal (class II) cavities in primary molars.
Exclusion Criteria: Studies were excluded if they were retrospective review of case notes; case reports, laboratory, animal or questionnaire based studies, assessed safety or toxicity of AM, involved the use of an unconventional/experimental AM alloy or used AM in an unconventional manner eg. Bonded AM. Occlusal restorations were also excluded from this study as they did not represent the major clinical challenge in managing caries in children.
Assessment criteria: These were based on those of Curzon and Toumba,  with modifications relevant to the restorations under review. These modifications were made on discussion with the authors of the systematic review of resin modified GICs published simultaneously herein [Chadwick and Evans, 2007]. Only studies with a minimum 2 year follow up were included. Half marks were given for incomplete reporting; if patients were randomly allocated, but details of randomisation were not supplied (criteria 13); taking but not standardising radiographs (criteria 14); use of consensus assessments from more than one examiner were used but no details reported of any calibration process (criterion 17). Finally, given that it is impossible for examiners to be blinded to the use of AM, a 1/2 mark was given if the examiners were independent from the operators. Where only one operator was involved criterion 11 was given a whole mark as calibration of operator became unnecessary and it was assumed that the operator was adequately trained and not a novice. As in the GIC review sponsorship was recorded but not scored [Chadwick and Evans, 2007]. Thus there were 22 criteria listed in Table 1.
Grading: All included papers were graded according to the number of points scored using the modified criteria as follows:
Grade A = 90% or greater (20/22)
Grade B1 = 75% (16.5 to 19.5/22)
Grade B2 = 50% (11 to 16/22)
Grade C = <50% (less than 11/22)
The literature search identified 380 papers in OVID--Medline and 44 papers in EMBASE, however all of the papers listed in EMBASE were also found in Medline. After reviewing the abstracts of the 380 identified papers 325 were discarded leaving 46 papers written in English. These papers were evaluated independently by the two authors according to the assessment criteria; following this a further 22 studies were discarded: 9 were retrospective chart reviews; 10 failed to report 24 month review data; 2 involved assessment of restored but extracted or exfoliated teeth; and one only included occlusal (class I) restorations. The remaining 24 papers were independently scored by both authors using the agreed criteria. Where differences existed the papers were reviewed again and a consensus reached. A further calibration exercise was carried out with a subset of the papers that were also included in the systematic review of glass ionomer cements [Chadwick and Evans, 2007].
No study achieved a Grade A, 4 studies rated Grade B1 and 13 rated Grade B2. Seven studies failed to reach a minimum score of 11/22 and were therefore rated C [Morris et al., 1979; Nelson et al., 1980; Petersson et al., 1985; Roberts et al., 1985; Roberts and Sherriff, 1990; Hickel and Voss, 1990; Chu et al., 1996]. Table 2 summarises the criteria not achieved by these grade C studies. In general these studies failed to provide an adequate description of the type of participants, the inclusion and exclusion criteria and caries status. In addition no study reported on the training or calibration of the examiners (who in many cases were also the operators). Whilst these studies will not be discussed further in detail it is worth noting one study because its outcomes are often quoted as benchmark data [Roberts and Sherriff, 1990]. This was a large study involving a single specialist paediatric dentist in a private practice setting. It is useful in that it provides the longest ever follow up on a large number of restorations. The failure rate for the class II AM restorations over the 10 period of the study was extremely low (11.6%) with a MST in excess of 7.5 years. Despite these strengths it only scored 8/22 in this systematic review process. Inadequate information was provided on the inclusion and exclusion criteria, choice of tooth, randomisation of restorative materials and the operator was also the examiner. As such it missed criteria 3 - 8, 11, 13, 14, 16-18 and 21, with 1/2 marks for 9 (on the assumption that all patients referred to the practice were involved) and 12.
Of the studies that achieved a grade B1 the highest score (18.5/22) was a field study in Syria of the Atraumatic Restorative Technique (ART) [Taifour et al., 2002]. In this study AM was used as a control material for two conventional glass ionomer cements (GIC). A large number of multi-surface restorations (425 AM and 610 GIC) were placed by 8 operators in Syria using the principles of minimal intervention. In addition the background fluoride status was not reported on, radiographs not included (criteria 14) and half marks were given to criteria 11, 13 and 16. Using clearly reported and well recognised assessment criteria [Frencken et al., 1996] this study showed that just under 58% of class II AMs failed over the 3 year study period in comparison to 51% for the GICs (the difference not being statistically significant, p>0.05). There was also no significant difference in the cumulative 3-year survival for class II AMs (42.9%) and the GICs (48.7%). However, the GIC did perform significantly better than AM in the occlusal restorations. There were also marked differences in outcomes between the 8 operators despite similar training.
The remaining 3 grade B1 studies were all carried out under more controlled clinical environments, involved much smaller numbers of restorations and all used modified USPHS assessment criteria [Cvar and Ryge, 1971]. In their randomized clinical trial (RCT) Duggal and co-workers placed 78 pairs of restorations in a split mouth design to compare the durability of compomer with a conventional AM in two settings; a general dental practice and a hospital paediatric practice [Duggal et al., 2002]. This carefully controlled study scored 17.5/22 omitting only criteria 3, 7, 16 and 21 and scoring a 1/2 mark for criteria 14 as they used radiographs routinely pre-operatively but only used when clinically indicated post-operatively. Whilst local analgesia (LA) was used routinely, rubber dam (RD) was not but the children were screened for behavioural characteristics. Sixty (77%) of the original 78 pairs of restorations were available at 24 months and the effectiveness of the two materials was reported in terms of recurrent caries, wear of material, marginal integrity and surface texture. This study did not clearly report on the actual failure rates rather in terms of complete loss of a restoration, with 4 (6.6%) of the AM and 2 (3.3%) of the compomer restorations being lost. However, 9 (11%) of the AMs were reported to have recurrent caries at 24 months and 20 (33.3%) of the AMs scored B or greater for marginal integrity (which was significantly greater than the compomer, p< 0.05).
A more recent study also compared a compomer with AM and also scored 17.5, missing criteria 3, 5 and 18 [Kavvadia et al., 2004]. In this 24 month study 75 pairs of restorations, randomly assigned in a split mouth design were placed by 2 experienced paediatric dentists in private specialist practices. This is one of the few studies to refer to the background fluoride levels in that they reported that every child was seen 6 monthly for review and topical fluoride applications. This scored a 1/2 mark for criterion 21. In addition the study scored 1/2 marks for criteria 16 and 17 as the examiners were reported to have been independently calibrated but no further details were provided. As in the Duggal et al. study  the overall failure rate for AM (2%) was no different to that of the compomer, however in this study the AM restorations showed less deterioration in marginal integrity and anatomic form over the 24 month follow up period suggesting potentially greater longevity.
Kotsanos and Dionysopoulos  reported a 24 month randomised clinical trial (RCT) comparing a compomer with an AM control. This study scored highly (18/22) omitting criteria 3 and 5 only and scoring 1/2 marks for 13, 17, 18 and 21. Interestingly the prime outcome of interest was not the durability of the restorative material itself but the caries preventive effect on the adjacent tooth surface. Therefore little information was reported on the restorations, i.e. no USPHS criteria. It was noted, however, that only 1 AM and 1 GIC actually failed during the course of the study giving rise to a failure rate of 1/41 = 2.4%. There was no statistically significant difference in the caries preventive effect between the surfaces adjacent to the GIC and those next to the amalgam (p = 0.813).
Thirteen studies rated B2 (Table 4) of which three were scored 16 [Welbury et al., 1991; Qvist et al., 1997; Qvist et al., 2004] although 2 of these are actually the same study reported twice, at 3 and then 8 years. On both occasions the same criteria were omitted (3, 14, 16-18 and 21). The study defined failure simply as the need for further treatment but did include a summary of the reasons for the re-treatment. This has been scored as a 1/2 mark by other reviewers [Chadwick and Evans, 2007] but it was considered sufficient to score fully in the present study. The study involved 14 dentists working in the Danish public dental service who were all trained in the contemporary diagnosis and management of dental caries. Whilst RD dam was not used, LA and nitrous oxide relative analgesia (as appropriate) were and the clinical technique described fully. However, the operators were also the examiners and there was no report of any formal training in the objective application of neither assessment criteria nor examiner calibration. At both 3 and 8 years the class II AM restorations performed significantly better than the GICs, with only a 22% failure rate at 96 months (compared with 46% for the GICs). The MST for the class II AMs was greater than 7.8 years, limited only by the duration of the study and the fact that the majority of the restored teeth had exfoliated naturally by the end of the study.
In comparison to the community based clinical trial, Welbury and co-authors  also scored 16/22 for their well designed RCT using a split mouth design involving only 2 experienced operators over a period of five years. Whilst both operators were also examiners, an examiner calibration study was reported. However, this study omitted criteria 3, 7, 11, 14, 16 and 21. The study involved 119 pairs of restorations with 8 being occlusal restorations (separating the data pertaining to occlusal restorations from that of the class II restorations was not always possible). The overall failure rate for the AM restorations was 20% and the MST was 41 months being significantly better than the GIC outcomes. The earlier report of the same study [Walls et al., 1988] only scored 13/22 as, in addition to omitting the same criteria, it also failed to report on criteria 5, 17 and 18.
A 36 month study scored 15/22 [Marks et al., 1999] in which the cavities were minimal (being diagnosed radiographically) and the clinical technique carefully controlled with RD isolation and LA. This study missed criteria 3, 5, 10, 11 and 21. Half marks were given to criteria 13 and 16 - 18 due to incomplete reporting regarding independence and calibration of clinical assessments. Of the initial 30 pairs of restorations only 17 remained at the end of the 3-year review period. The overall failure rate was very low for both materials with only 2 (6.6%) of the initial 30 AM restorations known to have failed. One was due to pulpal pathology which potentially should not be attributed to the restorative material itself, rather it is likely to be a misdiagnosis of pulp status at the time of initial treatment. The other was due to recurrent caries. Again the AM restorations performed inferiorly in terms of marginal integrity after 36 months compared with the compomer (P = -0.028) but this failed to reach clinical significance.
Under our review process the study reported by Fuks in 2000 scored 14.5/22 with criteria 3, 5, 7, 16, 18 and 21 being omitted. Half marks were given to criteria 13, 14, 16, 17 and 22. In this hospital based study all restorations were placed by the same operator and a degree of independence was introduced in the post-operative examinations. Radiographs were taken of many but not all of the restorations post-operatively. Although the authors reported that none of the AM restorations failed it is important to note that of the initial 40 restorations that were placed, only 2 were reviewed at 36 months. Furthermore it is difficult to establish from the data what proportion of restorations was examined (clinically and radiographically) and at what time frames.
Four studies scored 14/22 but on closer review, whilst well designed, the study by Honkala and co-workers  only provided data on 9 class II AM restorations (26 class I AM and an additional 48 ART restorations). Scoring 14/22 this study failed to report on criteria 3, 7, 14, 16, 18, 20 and 21. A 1/2 mark was given to criterion 11 as the nature of the training of the operators with respect to the placement of the AM restorations was not clearly reported. Similarly, criterion 17 was also given a 1/2 mark as only one examiner was available for the final assessment. Whilst none of the class II AM restorations in this study
failed, the very small sample (9) makes this information of limited value.
Small sample size plagues many of the studies set in the more controlled environments of a hospital paediatric unit or specialist practice. Mass and co-workers  also scored 14/22 with criteria 3, 5, 7, 11, 16, 18, 21 missing and with 1/2 marks for inadequate reporting of criteria 13 and 17. Under carefully controlled conditions of a private dental practice 44 AM and 63 compomer restorations were placed under RD. The use of LA was not recorded and the operator also examined the restorations clinically over a 36 months period. Radiographs were taken regularly and these were reviewed by independent examiners. No clinical failures were reported but it appeared that only 4 AM (and 6 compomers) were available at 36 months. Even at 24 months only 18 compomer and 14 AM restorations were reviewed.
A slightly larger final sample was reported by Oldenberg and co-workers  with 18 out of the original 34 restorations available for review at 24 months. This study scored 14/22 and missed criteria 3, 5, 7, 11, 14, 16, 20 and 21 reporting that 'in most instances the child was treated by the same operator ...' and also that the two examiners were trained in the use of the modified USPHS system. However, the overlap between the operator (s) and examiners was not clear. The failure rate was reported as being 8.8% based on 3 failures out of the initial 34 restorations placed.
Tonn and co-workers  scored surprisingly high (14/22) for quite an old study comparing an early chemically curing composite resin with AM in primary molars. They missed criteria 3, 5, 7, 14, 16 and 21 with 1/2 marks for criteria 17, 18 and 20. Whilst the composite restorations did not perform very well, only 8 AMs required replacement over a 24 month period, a 7.6% failure rate based upon the initial 105 (8/105) restorations placed. However, at 24 months only 76 were available for review but even using this figure the failure rate was low at 10% (8/76). This study provided some basic comparisons of rates of deterioration between composite resins and AM across all criteria (anatomic form, marginal integrity and caries). AM was significantly superior but only with regard to anatomic form.
Barr-Agholme and co-workers used AM as the control material in their study of a composite resin restorative material in a high caries population (initial doffs = 7.3 +/- 4.3) hospital paediatric practice, scoring 13.5/22. However, as in a more recent study [Kavvadia et al., 2004] a 1/2 mark was given for criterion 21 because 6-monthly fluoride varnish applications were reported. Half marks were also given to criteria 10, 12 and 13 because, while the number of operators was reported, inadequate information was provided on their experience, the use of LA or RD and the randomisation process. The presentation of the results in this study is highly confusing; 21 AM restorations were reportedly not evaluated at 24 months because of recurrent caries (n=7), fracture (n=7) or censored (exfoliated or failed to attend, n =7). The reported outcome for AM was based upon the residual 34 (55-21) restorations with 23/34 (68%) rated as acceptable (rating alpha or bravo using the USPHS criteria) which implied a 32% failure rate (11/34). But it would be appropriate to include as 'failures' the 14 restorations that were reported with recurrent caries (7) and fracture (7) prior to the 24 month review. Recalculating the outcome based on this assumption would make the total number of AMs assessed as 48 (34 + 14) and the overall failure rate 11 + 14 = 25/48, which was 52% over the 2 year period. Unfortunately there is insufficient information reported to clarify these issues.
Donly and co-workers  reviewed 40 class II AMs over 36 months as part of an RCT to evaluating the effectiveness of a compomer . This study scored 13/22 with criteria 3, 5, 7, 14, 16-18 and 21 being missed. Further 1/2 marks were given for criteria 10 and 13. At the 36 months review only 19 of the initial 40 restorations were available for review. The failure rate (taken to be the re-treatment rate) was reported as 29.4% for the AM restorations, however, whether this figure was based upon the original 40 restorations (of which 10 were not available for review even 6 months after placement) or on the final 19 restorations is unclear.
There was a much higher retention rate in Ostlund's study [Ostland et al, 1992] with 23 out of the initial 25 class II AMs being available for review after 36 months. Two operators (who were also the examiners) placed all the restorations under standard conditions which included RD for the composite resins and GIC restorations but not for the AM restorations. However, this study only scored 12.5/22 because it missed criteria 3, 5, 7, 11, 13, 14, 16, 18, 20 and 21. A 1/2 mark was given for criterion 17. Due in all likelihood to the young age of the children, only 2/25 teeth exfoliated (with the restorations reportedly intact) but the age of exfoliation was not reported. In total only 2 AMs (8%) required replacement (in comparison with 4 composite resins and 15 GICs). Whilst no statistical analysis was reported the difference in outcome particularly with respect to the GIC is notable.
Given the importance of AM as the benchmark against which all newer restorative materials are compared, the lack of any Grade A studies is very disappointing. The failure rate of class II AM restorations in primary molars ranged widely from 0% - 58%. The highest failure rate reported (58%) actually came from one of the highest quality studies (18.5/22) however the result needs to be viewed in context and cautiously [Taifour et al., 2002]. This study was a field trial of ART in which AM was used as the control. The operators (8) were all trained in ART which does not involve the use of rotary instruments. The AM restorations were reportedly placed using '..the drill. .' but no further details were available and the use of LA was reportedly uncommon. It may not be appropriate to extrapolate these results to the more developed general dental practice. Furthermore, in another study in which AM was used in a more conventional manner (with LA and rotary instrumentation in a paediatric dental clinic), and compared with the ART, it performed well over the 24 month period with no failures [Honkala et al., 2003]. However, close examination of this study reveals that only 9 class II AM restorations were actually placed.
Perhaps the most clinically realistic study, which also rated highly (16/22), is that by Qvist and co-workers  set in the Danish Public Dental Service. The size of the study (N >1,000), the number of operators (14) and the duration of follow up (96 months) allow for more credible extrapolation to the general dental environment. The MST for all types of AM restoration in primary molars exceeded 7.5 years whilst the equivalent value for the GIC was significantly less at 42 months (p<0.001). The aim of this large study was to evaluate the potential impact of a ban on AM on the public dental service and so the outcomes reported included not only longevity of all restorations but also the impact of material type on future treatment need for enamel surfaces in contact with the restorations. When this factor was analysed 30% of the surfaces in contact with an AM required further treatment compared with only 16% for those in contact with the GIC. Despite this the authors did not support the use of GICs as an alternative to AM due to its extremely high failure rate (46%).
Under the more controlled conditions of an RCT the failure rate for AM in class II cavities in primary molars appears to be very low. Figures from 2% [Kavvadia et al., 2004, Kotsanos and Dionysopoulos, 2004] to 6.6% [Marks et al., 1999; Duggal et al., 2002] to just over 20% [Walls et al., 1988, Welbury et al., 1991] have been reported. Given the carefully controlled conditions under which the restorations are placed in these studies, these results potentially represent the optimal outcome for AM.
Three issues arise from this systematic review: first only 2 studies ran for longer than 36 months [Welbury et al., 1991; Qvist et al., 2004]. Given that data from the early retrospective studies (and indeed the Roberts and Sherriff study) suggest that the durability of AM exceeds 36 months, prospective studies trials that run for less than this are likely to under represent the outcomes for AM. Fortunately both the longer studies scored at the upper level in terms of quality of evidence (16/22). Furthermore, the two studies present data from very different settings, one being a community based study whilst the other was a carefully controlled RCT in a specialist hospital setting. As such it is not unreasonable to consider the failure rate for class II AM in the primary dentition as being around 20%, given that both studies showed similar failure rates (20.2%, 22.0%). The very low failure rates (2 - 6.6%) reported in the three RCT's that scored B1 whilst encouraging and positive are likely to be examples of where the short duration of the study has not allowed the full potential of the restorative to be evaluated [Duggal et al., 2002, Kavvadia et al., 2004, Kotsanos and Dionysopoulos, 2004].
The second issue is the high drop out rate in many studies and the difficulties this placed on interpretation of the results. For example; in two studies 4 out 44 and 2 out of 24 AM restorations were available for review at 36 months [Mass et al., 1999, Fuks et al., 2000]. In another study 40 pairs of restorations were placed and 19 were reviewed at 3 years [Donly et al., 1999]. However, the results are presented in such a way that it is difficult to establish the failure rate; 29.4% of the AM restorations were described as 'replacement' and a further 14.29% as 'fractured'. What is not clear is exactly what these terms mean and whether the 'fracture' group were included in the 'replacement' group. So called censored (exfoliated or lost to follow-up) restorations are also handled in a variety of ways. In one study it is assumed that all censored restorations remained intact which effectively optimises the outcomes [Oldenburg et al., 1987]. At 24 months 16 of the initial 34 AM restorations had either exfoliated (15) or were lost (1). The failure rate is reported as being 8.8% which is 3 out of the initial 34 restorations placed. The worst case scenario could be a failure rate of 3 out of the remaining 18 restorations or 16.6%.
Finally, the third issue is the absence of studies specifically designed to evaluate AM rather than studies in which this material has been used as the control. This situation arises from a number of factors including the historical use of AM in clinical practice, decreasing popularity and concerns over safety as well as the influence of industrial sponsorship on research programs. However, as a result the outcome criteria used in many studies may well be valid and sensitive when applied to adhesive/aesthetic materials but may not be so pertinent to AM itself. Many studies use modifications of the USPHS criteria for the clinical assessments and report on changes in marginal integrity, anatomic form etc [Cvar and Ryge, 1971]. However when completing this review, we found it more difficult to draw clinically relevant conclusions from findings of deterioration in marginal integrity or anatomic form for AM compared with an adhesive material and clinical relevance, especially when the follow up period was short.
A systematic review of the durability of amalgam, when used to restore interproximal (class II) cavities in primary molars, suggests that it can be expected to survive a minimum of 3.5 years but potentially in excess of 7 years. Given the limited lifespan of most primary molars, amalgam remains an appropriate treatment option for the management of caries in children. Nevertheless, the future use of amalgam in paediatric dentistry is likely to be influenced not by evidence of its effectiveness and durability but rather by public concerns over toxicity, safety and aesthetics. Therefore, despite the disappointing lack of high quality studies it is questionable whether amalgam should be considered as a restorative material option in paediatric dentistry in RCT's in the future.
AAPD. Guideline on pediatric restoration dentistry. Pediatr Dent. 2005-6;Reference Manual: 122 - 129.
Barr-Agholme M, Oden A, Dahllof G, Modeer T. A two-year clinical study of light-cured composite and amalgam restorations in primary molars. Dent Mater. 1991;7: 230-3.
Braff MH. A comparison between stainless steel crowns and multi-surface amalgams in primary molars. J Dent Child. 1975;42: 474-8.
Chadwick BL, Evans DJP. Restoration of class II cavities in primary teeth with conventional and resin modified glass ionomer cements: An evidence based assessment of the literature. Eur Arch Paed Dent. 2007; 1:13-20.
Chu CH, King NM, Lee AM, Yiu CK, Wei SH. A pilot study of the marginal adaptation and surface morphology of glass-cermet cements. Quintessence Int. 1996;27: 493-501.
Curzon ME, Toumba KJ. Restoration of primary teeth: clinical criteria for assessment of the literature. Eur Arch Paed Dent. 2006;7: 48 - 52.
Cvar JF, Ryge G. Criterion for the clinical evaluation of dental restorative materials. 1971;USPHS Publication No 790-244.
Dawson LR, Simon JF, Jr., Taylor PP. Use of amalgam and stainless steel restorations for primary molars. J Dent Child. 1981;48: 420-2.
Donly KJ, Segura A, Kanellis M, Erickson RL. Clinical performance and caries inhibition of resin-modified glass ionomer cement and amalgam restorations. J Am Dent Assoc. 1999;130: 1459-66.
Duggal MS, Toumba KJ, Sharma NK. Clinical performance of a compomer and amalgam for the interproximal restoration of primary molars: a 24-month evaluation. Br Dent J. 2002;193: 339-42.
Frencken JE, Makoni F, Sithole WD. Atraumatic restorative treatment and glass-ionomer sealants in a school oral health programme in Zimbabwe: evaluation after 1 year. Caries Res. 1996;30: 428-33.
Fuks AB, Araujo FB, Osorio LB, Hadani PE, Pinto AS. Clinical and radiographic assessment of Class II esthetic restorations in primary molars. Pediatr Dent. 2000;22: 479-85.
Gordon M, Gorfil C, Segal S, Mass E. Treatment policies among Israeli specialists in paediatric dentistry. Europ J Paediatr Dent. 2005;6: 73-8.
Hickel R, Voss A. A comparison of glass cermet cement and amalgam restorations in primary molars. J Dent Child. 1990;57: 184-8.
Holland IS, Walls AW, Wallwork MA, Murray JJ. The longevity of amalgam restorations in deciduous molars. Br Dent J. 1986;161: 255-8.
Honkala E, Behbehani J, Ibricevic H, Kerosuo E, Al-Jame G. The atraumatic restorative treatment (ART) approach to restoring primary teeth in a standard dental clinic. Int J Paediatr Dent. 2003;13: 172-9.
Kavvadia K, Kakaboura A, Vanderas AP, Papagiannoulis L. Clinical evaluation of a compomer and an amalgam primary teeth class II restorations: a 2-year comparative study. Pediatr Dent. 2004;26: 245-50.
Kotsanos N, Dionysopoulos P. Lack of effect of fluoride releasing resin modified glass ionomer restorations on the contacting surface of adjacent primary molars: a clinical prospective study. Europ J Paediatr Dent. 2004;5: 136 - 142.
Levering NJ, Messer LB. The durability of primary molar restorations: I. Observations and predictions of success of amalgams. Pediatr Dent. 1988;10: 74-80.
Llewelyn DR. A pilot study of 230 restorations in children's mouths. Proc Br Paedod Soc. 1977;7: 19-21.
Marks LA, Weerheijm KL, van Amerongen WE, Groen HJ, Martens LC. Dyract versus Tytin Class II restorations in primary molars: 36 months evaluation. Caries Res. 1999;33: 387-92.
Mass E, Gordon M, Fuks AB. Assessment of compomer proximal restorations in primary molars: a retrospective study in children. J Dent Child. 1999;66: 93-7.
Milsom KM, Tickle M, Blinkhorn A. The prescription and relative outcomes of different materials used in general dental practice in the northwest region of England to restore the primary dentition. J Dent. 2002;30: 77 - 82.
Mitchell L, Walls AWG. Life survival analysis. Dent Update. 1991;18: 125 - 128.
Morris ME, Barkin PR, Soelberg KB, Weis RW. Complex primary molar restorations using a composite resins. (A 42-month study using amalgam as comparative material). J Calif Dent Assoc. 1979;7: 39-42.
Nelson GV, Osborne JW, Gale EN, Norman RD, Phillips RW. A three-year clinical evaluation of composite resin and a high copper amalgam in posterior primary teeth. J Dent Child. 1980;47: 414-8.
NHMRC (1999) Dental amalgam and mercury in dentistry. National Health and Medical Research Council, Canberra, Australia.
Oldenburg TR, Vann WF, Jr., Dilley DC. Comparison of composite and amalgam in posterior teeth of children. Dent Mater. 1987;3: 182-6.
Ostlund J, Moller K, Koch G. Amalgam, composite resin and glass ionomer cement in Class II restorations in primary molars-a three year clinical evaluation. Swed Dent J 1992;16:81-86.
Papathanasiou AG, Curzon ME, Fairpo CG. The influence of restorative material on the survival rate of restorations in primary molars. Pediatr Dent. 1994;16: 282-8.
Peretz B, Ram D. Restorative material for children's teeth: preferences of parents and children. J Dent Child. 2002;69: 243 - 248.
Petersson LG, Rasmusson CG, Hagborg S, Isacsson P. Fluoride supplemented and non gamma 2 amalgam. A comparative clinical study into the primary and permanent dentition in children. Swed Dent J. 1985;9: 49-53.
Peto R, Pike MC, Armitage P. Design and analysis of randomised clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer. 1977;35: 1 - 39.
Qvist V, Thylstrup A, Mjor IA. Restorative treatment pattern and longevity of amalgam restorations in Denmark. Acta Odontol Scand. 1986;44: 343-9.
Qvist V, Laurberg L, Poulsen A, Teglers PT. Longevity and cariostatic effects of everyday conventional glass-ionomer and amalgam restorations in primary teeth: three-year results. J Dent Res. 1997;76: 1387-96.
Qvist V, Laurberg L, Poulsen A, Teglers PT. Eight-year study on conventional glass ionomer and amalgam restorations in primary teeth. Acta Odontol Scand. 2004;62: 37-45.
Roberts JF, Sherriff M. The fate and survival of amalgam and preformed crown molar restorations placed in a specialist paediatric dental practice.[erratum appears in Br Dent J 1990 Nov 10;169(9):285.
Roberts MW, Broring CL, Moffa JP. Two-year clinical evaluation of a proprietary composite resin for the restoration of primary posterior teeth. Pediatr Dent. 1985;7: 14-8.
Roshan D, Curzon ME, Fairpo CG. Changes in dentists' attitudes and practice in paediatric dentistry. Europ J Paediatr Dent. 2003;4: 21-7.
Rugg-Gunn AJ, Welbury RR, Toumba J, British Society of Paediatric D. British Society of Paediatric Dentistry: a policy document on the use of amalgam in paediatric dentistry. Int J Paediatr Dent. 2001;11: 233-8.
Taifour D, Frencken JE, Beiruti N, van't Hof MA, Truin GJ. Effectiveness of glass-ionomer (ART) and amalgam restorations in the deciduous dentition: results after 3 years. Caries Res. 2002;36: 437-44.
Tonn EM, Ryge G, Chambers DW. A two-year clinical study of a carvable composite resin used as class II restorations in primary molars. J Dent Child. 1980;47: 405-13.
Tran LA, Messer LB. Clinicians' choices of restorative materials for children. Aust Dent J. 2003;48: 221-32.
Walls AW, Murray JJ, McCabe JF. The use of glass polyalkenoate (ionomer) cements in the deciduous dentition. Br Dent J. 1988;165: 13-7.
Welbury RR, Walls AW, Murray JJ, McCabe JF. The 5-year results of a clinical trial comparing a glass polyalkenoate (ionomer) cement restoration with an amalgam restoration. Br Dent J. 1991;170: 177-81.
N.M. Kilpatrick *, A. Neumann *.
* Department of Dentistry, Royal Children's Hospital and Oral Health Research Unit, Murdoch Children's Research Institute, Melbourne, Victoria, Australia.
Postal address: A. Professor N Kilpatrick, Department of Dentistry, Royal Children's Hospital, Flemington Road, Parkville, Victoria, 3052 Australia..
Table 1 Criteria used to assess quality of paper. 1 Published in peer reviewed journal 2 Prospective study 3 Power calculation (study size determination) 4 Inclusion criteria listed 5 Exclusion criteria listed 6 Tooth selection criteria given 7 Caries status recorded dmfs/dmft 8 Criteria for groups given (age, sex, race) 9 Sample stratification or convenience sample recorded 10 Details on operators given (number, experience, setting) 11 Training, calibration etc of operators 12 Details of technique recorded (LA, isolation etc) 13 Assignment of subjects by an acceptable system, randomisation detailed. 14 Pre-op and post-op radiographs standardised 15 Post-op assessment criteria listed 16 Post operative assessment blinded 17 Post-op assessment examiner(s) calibrated 18 Kappa scores or equivalent for post-op examiners 19 Time to failure/replacement to at least 24 months 20 Appropriate statistical tests used 21 Fluoride background is reported 22 Outcome report is based on results Table 2 Summary of C grade studies on the use of amalgam in the primary dentition. Study Score Criteria Missing 1/2 mark Nelson et al 1980 10/22 3-8,11, 14, 16, 13, 17 18, 21 Morris et al 1979 9.5/22 3, 5, 7, 8, 10, 11 13 14, 16-18, 20, 21 Hickel and Voss 1991 9.5/22 3-7, 10,11, 14, 12 16-18, 21 Chu et al 1996 9.5/22 3, 5, 7,10, 11, 13, 6 14, 16-18, 20, 21 Roberts et al 1995 8/22 3-11, 14, 16-18, 21 Roberts and Sheriff 1990 8/22 3-8, 11, 13, 14, 9, 12 16-18, 21 Petersson et al 1985 7.5/22 3, 5-9, 11, 12, 14, 13 16-18, 20, 21 Table 3 Summary of the studies of class II amalgam restorations in primary molars ranked B1. Author Score Design Subject No. CI II Year Ages (yrs) Restns (at end of study) * Taifour 18.5/22 Clinical 6-7 425 (78%) et al. 2002 Trial Kotsanos 18/22 RCT 4-7 83 pairs 2004 Split (41) mouth Duggal 17.5/22 RCT 4-7 78 pairs et al. 2002 (60) Split mouth Kavvadia 17.5/22 RCT 6-9 75 pairs et al. 2004 Split (57) mouth Author Examiners Study Failure Comments Year length Rate (mnths) Taifour 3 * 36 57.1% An ART field trial et al. 2002 vs amalgam control. Generally no LA ART assessment criteria Sponsorship reported Kotsanos 1 * 24 2.4% Specialist paediatric 2004 1/41 practice LA and rubber dam. Outcome focused on caries on adjacent surface Duggal 3 24 6.6% Both general and et al. 2002 hospital practice. LA but no rubber dam. Sponsorship reported. Kavvadia 3 * 24 2% Specialist paediatric et al. 2004 practice. Sponsorship reported. Amalgam deteriorated less than compomer. * examiners were not the same as the operators. Table 4 Summary of the studies on the use of class II amalgam restorations in primary molars ranked B2. No. of [Rest.sup.ns] Author Subject (No. at end Year Score Design Ages (yrs) of study) * Welbury 16/22 RCT 5-11 119 pairs et al. 1991 Split of [Rest.sup.ns] mouth (51 pairs of [Rest.sup.ns]) Qvist et al. 16/22 RCT 3-13 456 2004 (10%) Qvist et al. 16/22 RCT 3-13 456 1997 (353) Marks et al. 15/22 RCT 4-9 30 1999 Split (17) mouth Fuks et al. 14.5/22 RCT 8-10 24 2000 (2) Honkala et al. 14/22 Clinical 2-9 9 2003 Trial (9) Mass et al. 14/22 RCT 3-11 44 1999 (4) Oldenberg 14/22 RCT > 7 34 et al. 1987 (18) Tonn et al. 14/22 RCT 3-8 105 pairs 1980 of [Rest.sup.ns] (76 pairs of [Rest.sup.ns]) Barr-Agholme 13.5/22 RCT Mean = 6.4 55 et al. 1991 (34) Donly et al. 13/22 RCT 6-9 40 1999 Split (19) mouth Walls et al. 13/22 RCT 5-11 58 pairs 1988 Split of [Rest.sup.ns] mouth (51 pairs of [Rest.sup.ns]) Ostlund et al. 12.5/22 RCT 4-6 25 1992 (23) Study Author length Failure Year Examiners (mnths) Rate Welbury 2 * 60 20.2% et al. 1991 (24/119) Qvist et al. 14 96 22% 2004 (102/456) Qvist et al. 14 36 18% 1997 (82/456) Marks et al. 1 36 6.6% 1999 (2/30) Fuks et al. 3** 36 None 2000 Honkala et al. 2 24 none 2003 Mass et al. 1 (clinical) 36 none 1999 2 (radiographic) ** Oldenberg 2 24 8.8% et al. 1987 (3/34) Tonn et al. 2 Trained 24 5.7% 1980 (6/105) Barr-Agholme 2 24 Difficult et al. 1991 to interpret Donly et al. Unclear 36 Difficult 1999 to interpret Walls et al. 1 24 17.2% 1988 (10/58) Ostlund et al. 2 36 8% 1992 (2/25) Author Year Comments Welbury Analysis includes the 8 pairs et al. 1991 of class I restorations GIC failure rate = 32.8% (39/119) MST = 41.4 months Sponsorship declared Qvist et al. Operators = examiners 2004 GIC Cl II failure rate = 46% (177/384) MST = >7.8 yrs Sponsorship declared Qvist et al. Operators = examiners 1997 Examiners were trained GIC Cl II failure rate = 42% (61/384) Sponsorship declared Marks et al. Controlled hospital practice with 1999 LA and rubber dam GIC Cl II failure rate = 3.33% (1/30) Minimal cavities Fuks et al. Single operator. 2000 Only 2 amalgam [rest.sup.ns] were available at 36 months Honkala et al. LA used for the amalgams. 2003 Focus on GICs in ART rather than amalgam. GIC (ART Cl II) failure rate = none Mass et al. Only 4 amalgams available at 36 months. 1999 Very little outcome data presented Oldenberg 15/34 teeth exfoliated before et al. 1987 the end of the study. Outcome reported based on best case scenario. Failure rate of restorations for composite (H120) = 3.9% (2/51) Tonn et al. Unclear if one of the examiners 1980 was also the operator Rubber dam was used but LA not reported. Failure rate for composite = 12.4% (13/105) Barr-Agholme Composite appeared to survive et al. 1991 better than amalgam Donly et al. Very limited information on outcomes 1999 Amalgam replacement rate = 29.41% GI replacement rate = 26.66% Mean time to failure = 26.2 mnths +/- 16.3 mnths Walls et al. Analysis includes 8 pairs of class I restorations 1988 Limited information on failure of the control material (Ketac-Fil). No significant difference between amalgam and control. Early report of Welbury study Ostlund et al. Public dental clinic, LA but no rubber dam 1992 Only 2 amalgams lost to follow up Failure rate for composite resin = 16% (4/25) Failure rate for GIC = 60% (15/25) * examiners were not the same as the operators. * some blindness and calibration introduced.
|Gale Copyright:||Copyright 2007 Gale, Cengage Learning. All rights reserved.|