Composite Failure

Background. Failure of dental restorations is amajor concern in dental practice. Replacement of failedrestorations constitutes the majority of operative work.Clinicians should be aware of the longevity of, and likely reasons for the failure of, direct posterior restora-tions. In a long-term, randomized clinical trial, the authors compared the longevity of amalgam and composite.Subjects, Methods and Materials. The authors randomly assignedone-half of the 472 subjects, whose age ranged from 8 through 12 years, toreceive amalgam restorations in posterior teeth and the other one-half toreceive resin-based composite restorations. Study dentists saw subjectsannually to conduct follow-up oral examinations and take bitewing radi-ographs. Restorations needing replacement were failures. The dentistsrecorded differential reasons for restoration failure.Results. Subjects received a total of 1,748 restorations at baseline, whichthe authors followed for up to seven years. Overall, 10.1 percent of the base-line restorations failed. The survival rate of the amalgam restorations was94.4 percent; that of composite restorations was 85.5 percent. Annual failurerates ranged from 0.16 to 2.83 percent for amalgam restorations and from0.94 to 9.43 percent for composite restorations. Secondary caries was themain reason for failure in both materials. Risk of secondary caries was 3.5 times greater in the composite group.Conclusion. Amalgam restorations performed better than did compositerestorations. The difference in performance was accentuated in large resto-rations and in those with more than three surfaces involved.Clinical Implications. Use of amalgam appears to be preferable to use ofcomposites in multisurface restorations of large posterior teeth if longevity isthe primary criterion in material selection.Key Words. Amalgam; composite; randomized controlled clinical trials;dental restoration failure.JADA 2007;138(6):775-83.

The performance of dentalrestorations is influencedby several factors,including the restorativematerials used,1-3 the

clinicians level of experience,4 thetype of tooth,5,6 the tooths positionin the dental arch,7,8 the restora-tions design,9 the restorationssize,6 the number of restored sur-faces10,11 and the patients age.4,11

Failure occurs when a restorationreaches a level of degradation thatprecludes proper performanceeither for esthetic or functional rea-sons or because of inability to pre-vent new disease.

Failure of dental restorations isof major concern in dental practice.It has been estimated that thereplacement of failed restorationsconstitutes about 60 percent of alloperative work.12 Survival andfailure rates may be used as mea-sures of clinical performance. Thereason why a restoration fails alsois important, because it points to aspecific weakness of the restoration-tooth system.

The two direct dental restorativematerials most commonly usedtoday are silver-mercury amalgamand resin-based composites. Amal-gam is not suitable for visible resto-

A B S T R A C T

Dr. Bernardo is an associate professor of community and preventive dentistry, Faculdade de Medicina Dentria, Universidade de Lisboa, Portugal.Mr. Lus is a faculty member, Faculdade de Medicina Dentria, Universidade de Lisboa, Portugal. Dr. Martin is an associate professor, Departments of Oral Medicine, Dental Public Health Sciences, and Epidemiology, University of Washington, Health SciencesBuilding, 1958 N.E. Pacific St., Room B316, Seattle, Wash. 98195-6370, e-mail [email protected]. Address reprint requests to Dr. Martin.Dr. Leroux is an associate professor, Department of Dental Public Health Sciences and Department of Biostatistics, University of Washington, Seattle.Ms. Rue is a research scientist, Department of Dental Public Health Sciences and Department of Biostatistics, University of Washington, Seattle.Dr. Leito is a cathedratical professor, Institute of Health Sciences, Portuguese Catholic University, Lisbon, Portugal.Dr. DeRouen is a professor, Department of Dental Public Health Sciences and Department of Biostatistics, and the executive associate dean for research and academic affairs, University of Washington, Seattle.

Survival and reasons for failure of amalgamversus composite posterior restorationsplaced in a randomized clinical trialMario Bernardo, DMD, PhD; Henrique Luis, MS; Michael D. Martin, DMD, MSD, MPH, MA, PhD;Brian G. Leroux, MSc, PhD; Tessa Rue, MS; Jorge Leito, MD; Timothy A. DeRouen, PhD

C L I N I C A L P R A C T I C E

JADA, Vol. 138 http://jada.ada.org June 2007 775

ARTICLE

2

JA D A

CO

NT

IN

UI N G E D

UC

A

TIO

N

Copyright 2007 American Dental Association. All rights reserved.

rations in anterior teeth for esthetic reasons, butit still is used widely for posterior restorations. Inrecent years, the use of resin-based composites forthe restoration of posterior permanent teeth hasincreased significantly, although they are moretechnique-sensitive to place and more costly.13 Thereasons for this situation have to do with thebetter esthetic properties of the composites, andwith the general concerns about the use of metalsin the mouth. There is some evidence that thelongevity of composite restorations is less thanthat of amalgam restorations in similar circum-stances.1.3 It is important to consider the impact ofthe increasing use of composites in posterior teethand for clinicians to be aware of the longevity ofthese materials and likely reasons for theirfailure.

The Casa Pia Study of the Health Effects ofDental Amalgams in Children was a randomizedclinical trial designed to assess the safety of low-level mercury exposure attributable to dentalamalgam restorations.14 It began in 1996 as a col-laborative project between the University ofWashington, Seattle; the University of Lisbon,Portugal; and the National Institute of Dental andCraniofacial Research, and it recently concludedwith publication of its main findings.15 Asapproved by the institutional review boards of theUniversity of Washington and the University ofLisbon Faculty of Dental Medicine, this studyenrolled 507 children and provided comprehensivedental care for each of them for a period of sevenyears.

Because one-half of the subjects received onlycomposites and the other one-half only amalgamsfor posterior restorations, this study provided theopportunity to compare the survival and the rea-sons for failure of posterior amalgam and com-posite restorations in a randomized, controlledclinical trial with seven years of follow-up.

SUBJECTS, METHODS AND MATERIALS

The study sample consisted of 472 children bornfrom 1986 through 1989. We obtained consentfrom parents or guardians as well as assent fromthe children for participation in the study. Thechildren attended seven different schools inLisbon, all belonging to the same school system.In addition to age, eligibility criteria included dat least one carious lesion in a permanent posterior tooth;dno prior exposure to dental amalgam;durinary mercury concentration of less than 10

micrograms per liter;dblood lead concentration of less than 15 g perdeciliter;dan IQ score of at least 67 on the Comprehen-sive Test of Nonverbal Intelligence;dno interfering health conditions.

The subjects included in the trial ranged in agefrom 8 through 12 years. Forty-three subjectswere aged 8 years, 122 subjects were aged 9years, 156 subjects were aged 10 years, 136 sub-jects were aged 11 years, and 15 subjects wereaged 12 years.

We randomly assigned subjects to one of twotreatment groups for restoration of posterior per-manent teeth: one-half of the children receivedonly amalgam restorations, and the other one-half received only composite restorations. Onlyresin-based composite and amalgam restorationsof permanent posterior teeth were considered forthe purposes of this study, although any anteriorteeth needing restoration were treated (with com-posite in both groups). The reason for placementof all the restorations was primary caries.

Figure 1 shows the composition of the treat-ment groups and the number of restorations doneat baseline within each group. For purposes of thecomparison of restoration failures presented here,we chose to include only the restorations placedat baseline. Restorations placed at baseline weredone under the same initial conditions and wereobserved for the same period, allowing for directcomparisons between restorative materials.

During follow-up, we instituted oral hygieneand prevention programs to decrease diseaserates. This meant that restorations placed duringfollow-up were done in oral environments alteredfrom those at baseline, which could in turn makelongevity of newer restorations different fromthat of those placed at baseline.

All dental care was provided at the Universityof Lisbon Faculty of Dental Medicine, usingexisting standards of care common to both theUnited States and Portugal. We chose thematerials used in the study, Dispersalloy(Caulk/Dentsply, Milford, Del.) and Z100 MP +Scotchbond Multi-Purpose (3M ESPE, St. Paul,Minn.), to be representative of those most com-monly in use at the time the study began; theystill are representative of materials in use today.


776 JADA, Vol. 138 http://jada.ada.org June 2007

ABBREVIATION KEY. mAFR: Mean annual failurerates.


Study dentists placed restorations using rubberdam isolation whenever possible, and they usedthe materials according to the manufacturersinstructions. All subjects in both groups receivedthe same preventive measures in all respectsthroughout the study.

The restorative procedures were standardized,and because of the group assignment, clinicaldecisions were limited. This meant that dentistscould not decide which material to use in eachspecific case. Fourteen dentists with varyinglevels of practice experience placed the restora-tions. One dentist (M.B.) was involved in thetreatment planning and assessment of all thesubjects.

The study dentists saw the subjects annuallyfor follow-up oral examinations and bitewingradiographs, at which time they performed com-plete dental charting and noted any new treat-ment needs. We considered restorations needingreplacement to be failures. The study dentistscarefully considered and recorded differential rea-sons for restoration failure as they occurred,starting during the second year of follow-up. Wesubsequently classified failures occurring beforethat point through review of the clinical record.

We noted several tooth and restoration charac-teristics to further investigate their relationshipwith failure. These characteristics included thearch (maxillary or mandibular), the type of tooth(premolar or molar), the number of restored sur-faces and the size of the restoration (small,medium or large). For each restored surface, weconsidered the restoration small if the propor-tion of the area restored was less than one-quarter of the total surface area, medium if thesame proportion ranged from one-quarter to one-half, and large if the restored area occupied one-half or more than one-half of the entire surface.The score attributed to the entire restoration con-sisted of the maximum size considering all therestored surfaces of each tooth.

We calculated mean annual failure rates mAFR = 1 7 1 x

according to the formula in which x expresses thetotal failure rate at seven years.16 For the calcula-tion of relative risks (RR) and confidence inter-vals (CI) for them, we fit a Poisson distributionmodel and obtained the P values from a Waldtest. To adjust the RR for the effects of covariates,we used Poisson regression models. We used gen-eralized estimation equation methods to accountfor correlation between restorations placed in the

same person. The failure rates we used to calcu-late the RR consisted of the ratio of the number ofevents (frequency of failures) to the number ofrestoration-years. We followed restorations fromthe time of placement to the last examination atwhich each restoration was found either to besound or to have failed. The time to failure for thetwo kinds of restorations was displayed throughKaplan-Meier survival curves.

RESULTS

The study dentists placed 1,748 posterior restora-tions during the baseline phase of the Casa Piastudy, which we followed for a period of up to sevenyears. Table 1 shows the number of restorationsplaced by restorative material, tooth and restorationcharacteristics.

Overall, 177 (10.1 percent) restorations failedduring the course of the study. The survival rate ofthe amalgam restorations was 94.4 percent at sevenyears (Table 2). The survival rate for composite res-torations was 85.5 percent. Amalgam restorationswith only one surface or of small size had thehighest survival rates, of 98.8 percent and 98.9 per-cent, respectively. We found that among theamalgam restorations, large restorations and resto-rations with three or more restored surfaces had thelowest survival rates. Survival rates of the com-posite restorations followed the same trend



ChildrenRandomized

N = 507

Children With BaselineRestorations and Follow-up

n = 472*

Children inComposite Group

n = 233

Children inAmalgam Group

n = 239

BaselineComposite

Restorationsn = 891

BaselineAmalgam

Restorationsn = 2

BaselineComposite

Restorationsn = 1

BaselineAmalgam

Restorationsn = 854

Figure 1. Number of subjects in each group and number of eachtype of restoration included in the analysis for each group. *Nine-teen subjects had no dental examinations after baseline and 16 sub-jects had no restorations to permanent posterior teeth at baseline.Two amalgam restorations accidentally were placed in posteriorteeth in subjects in the composite group.


but were lower than those of amalgam restora-tions in all instances. We found that among thecomposite restorations, single-surface and smallrestorations had the highest survival rate (93.6percent) and restorations with four or more sur-

faces had the lowest survival rate (50.0percent).

Mean annual failure rates rangedfrom 0.16 to 2.83 percent for amalgamrestorations and from 0.94 to 9.43 per-cent for resin-based composite restora-tions (Table 2). The reasons for failureare shown in Table 3. All restorationsfailed because of either secondary(recurrent) caries or restoration frac-ture. Secondary caries was the mainreason for failure in both amalgam andcomposite restorations, accounting for66 percent (32 of 48) and 88 percent(113 of 129) of all failures, respectively.This was true independently of thearch, tooth type, number of surfacesrestored and restoration size.

Proportionally, more composite resto-rations than amalgam restorationsfailed because of secondary caries. Ofthe failures due to secondary caries,77.9 percent were in the compositegroup, while only 22.1 percent were inthe amalgam group. On the other hand,

among the restorations that failed because offracture, exactly one-half were composite restora-tions and the other one-half were amalgam resto-rations. Table 4 presents the RR of failure due tosecondary caries in composite versus amalgam



TABLE 1

Number of posterior restorations placedfor each restorative material, by arch,tooth type and restoration characteristics.

RESTORATIONCHARACTERISTIC

Amalgam AllComposite

RESTORATION TYPE

Arch MaxillaryMandibular

Tooth TypePremolarMolar

Number of RestoredSurfaces1234 or more

SizeSmallMediumLarge

ALL

439417

91765

4293387811

263460133

856

892856

2031,545

87969415223

545891312

1,748

453439

112780

4503567412

282431179

892

TABLE 2

Mean annual failure rates and survival at seven years, by arch, toothtype and restoration characteristics.RESTORATIONCHARACTERISTIC

SURVIVAL ATSEVEN YEARS (%)

Amalgam Composite Amalgam Composite

MEAN ANNUALFAILURE RATES (%)

ArchMaxillaryMandibular


Restored Surfaces1234 or more


ALL

95.293.5

94.594.4

98.890.588.581.8

98.993.389.5

94.4

84.586.6

85.785.5

93.680.666.250.0

93.684.974.3

85.5

0.700.95

0.800.82

0.171.411.742.83

0.160.991.58

0.82

2.372.04

2.182.21

0.953.035.729.43

0.942.314.15

2.21


restorations. Table 5 presents the sameinformation for failures due to fracture.

The RR of developing secondary carieswas significantly higher in composite resto-rations for both arches, for molars, for res-torations involving up to three surfacesand for all restoration sizes (P < .05). Therisks were not significantly different in res-torations with four or more surfacesinvolved (owing to small numbers) or inpremolars. The overall risk of secondarycaries was 3.5 times greater in compositerestorations than in amalgam restorations.Even after adjustment for sex, baselineage, and tooth and restoration characteris-tics, the RR was still 3.4 (95 percent CI,2.1-5.4). Figure 2A (page 781) shows thediverging survival curves due to secondarycaries.

On the other hand, the overall risk offracture for composite restorations wasslightly (0.9 times) lower than that foramalgam restorations. After adjustment,the RR was 1.1 (95 percent CI: 0.5-2.4), butneither risk ratio is significantly differentfrom a no-effect ratio of 1.0. Figure 2B (page781), in which the survival curves for fractures inthe two kinds of restorations are superimposed,shows the lack of effect.

DISCUSSION

Our study used data collected in a randomized,controlled clinical trial designed to assess thesafety of low-level mercury exposure arising from



TABLE 3

Reasons for failure, by arch, tooth type and restoration characteristics.


NUMBER (%) OF TEETH WITH RESTORATION FAILURE

Secondary Caries Fracture

Amalgam Composite

Secondary Caries Fracture

ArchMaxillaryMandibular


Restored Surfaces1234+


ALL

15 (3.4)17 (4.1)

5 (5.5)27 (3.5)

2 (0.5)22 (6.5)6 (7.7)2 (18.2)

2 (0.8)23 (5)7 (5.3)

32 (3.7)

6 (1.4)10 (2.4)

0 (0)16 (2.1)

3 (0.7)10 (3)3 (3.8)0 (0)

1 (0.4)8 (1.7)7 (5.3)

16 (1.9)

64 (14.1)49 (11.2)

16 (14.3)97 (12.4)

26 (5.8)59 (16.6)23 (31.1)5 (41.7)

16 (5.7)57 (13.2)40 (22.3)

113 (12.7)

6 (1.3)10 (2.3)

0 (0)16 (2.1)

3 (0.7)10 (2.8)2 (2.7)1 (8.3)

2 (0.7)8 (1.9)6 (3.4)

16 (1.8)

TABLE 4

Relative risk of secondary caries incomposite restorations compared withamalgam restorations, by arch, tooth typeand restoration characteristics.


RELATIVERISK

P VALUE95%CONFIDENCEINTERVAL





ALL

4.32.8

2.63.6

12.42.74.32.6

7.42.74.8

3.5

< .0001.0003

.0639< .0001

.0006< .0001

.0015

.2482

.0075< .0001

.0001

< .0001

(2.4-7.5)(1.6-4.8)

(0.9-7.1)(2.4-5.5)

(2.9-52.1)(1.6-4.3)(1.7-10.5)(0.5-13.5)

(1.7-32.3)(1.6-4.3)(2.1-10.7)

(2.3-5.1)


participating in this study graduatedfrom the Lisbon School of Dental Medi-cine, but their dates of graduationspanned 15 years, so it is possible thattheir training in the placement of com-posite restorations may have varied.Nevertheless, the subjects assignmentto the dentists was completely random.We evaluated the distribution of thedentists by restorative material in astudy subsample and found that thisdistribution was even.

The mean annual failure rates wefound in this study for amalgam andcomposite restorations placed in pos-terior permanent teeth over a seven-year period of follow-up were, respec-tively, 0.82 percent and 2.21 percent.These values are in agreement withthose found in the dental literature andmay be considered as small. In clinicalstudies, the annual failure rates ofstress-bearing restorations have rangedfrom 0 to 7 percent for amalgam resto-rations and from 0 to 9 percent for com-posite restorations.1,16-25 Some studies

that compared posterior amalgam restorationsand composite restorations have shown thatamalgam restorations have lower annual failurerates.1,3,18,26 This also was true in our study, inwhich the mean annual failure rates of compositerestorations were almost three times greater thanthose of amalgam restorations.

Survival results obviously are consistent withthe mean annual failure results. Of the amalgamrestorations, 94.4 percent remained intact andwere considered as clinically successful afterseven years, while only 85.5 percent of the com-posite restorations were so. Other survival ratesin studies with approximately the same durationas this study have varied between 80 and 95 per-cent for amalgam restorations, and between 50and 92.9 percent for composite restorations.1,3,19,27-30 In almost all of these studies, survival ofamalgam restorations was greater than that ofcomposite restorations. Only one of the citedstudies found a better survival for composite restorations.30

The reported approximate median survivaltimes ranged from three to eight years for com-posite restorations and from five to 15 years foramalgam restorations.2,4,31-35 Since most publishedclinical studies have reported follow-up times of



dental amalgam restorations. Even though thisstudy had different objectives, the fact that one-half the subjects had posterior teeth restored withamalgam restorations and the other one-half withresin-based compositeand that the groupassignment was randomizedmakes the datavalid for the purposes of our study. Additionally,the restorative materials and techniques werepredetermined and standardized, and all thetreatments were performed in the same univer-sity clinic, as they would be in a controlled clin-ical study designed to compare both types of res-torations. The only less standardized conditionwas the fact that the restorations were placed bya larger number of dentists than is usual in clin-ical studies. On the other hand, the largernumber of dentists participating may make theresults more comparable with what would be seenin clinical practice.

The identification of the dentist who placedeach restoration, though noted on the clinicalcharts, was never recorded in the computer data-base. This is the reason why we did not introducedentist identification as a covariate in theanalysis. It has been reported that the dentistslevel of experience may influence the clinical per-formance of dental restorations.4 All the dentists

TABLE 5

Relative risk of fracture in composite restorations compared with amalgam restorations, by arch, tooth type and restoration characteristics.


RELATIVERISK

P VALUE95%CONFIDENCEINTERVAL





ALL

0.90.9

NA*1.0

0.90.90.7NA

1.81.00.6

0.9

.9139

.8483

NA.8858

.9182

.8667

.6460NA

.6260

.9659

.4139

.8350

(0.3-2.9)(0.4-2.2)

NA(0.5-1.9)

(0.2-4.6)(0.4-2.2)(0.1-3.9)

NA

(0.2-2.0)(0.4-2.7)(0.2-1.9)

(0.5-1.9)

* NA: Not applicable.


three years or less, thosestudies yielded higher survivalrates for composite restora-tions than ours did. Our studyhad a follow-up period thatwas twice as long, which exac-erbated the observed differ-ences between amalgam andcomposite restoration survival.

Even though the failurerates we found for compositerestorations seem to be amongthe lowest reported, large res-torations and restorations withthree or more surfaces in-volved had mean annualfailure rates ranging from 4.15to 9.43 percent. Norman andcolleagues6 also reported thatlarger restorations performedmore poorly, regardless ofmaterial. Even though themean annual failure rates forboth amalgam and compositerestorations increased as the sizeand the number of restored surfaces increased,this increase clearly was more accentuated forcomposite restorations.

Secondary caries and restoration fracture werethe only two reasons we found for failure. Wedefined secondary caries as the presence of car-ious lesions located on the margins of existingrestorations, and restoration fracture as the lossof material whose location or extension compro-mises function or the prevention of new disease.Other reasons for failure commonly reportedinclude marginal defects, endodontic treatment,excessive wear and color changes. Most of thesereasons are associated with earlier generations ofresin-based composites, and they are becomingmore rare and less significant in currentmaterials.36

Fracture frequently is reported as a reason forfailure in load-bearing restorations. The adhesivetechnique used in composite restorations pro-duces a reinforcement of the restoration-toothsystem, reducing the risk of fracture.37 The factthat we found no significant difference in risk offracture between composite and amalgammaterials is consistent with the intended effect ofthis adhesive system.

Secondary caries was the main reason forfailure in both amalgam and composite restora-

tions. It accounted for 66.7 percent and 87.6 per-cent of the failures that occurred in amalgam andcomposite restorations, respectively. Substantialevidence has been reported to confirm that theclinical diagnosis of secondary caries is the prin-cipal reason for failure of both posterior compositeand amalgam restorations.2,18,30-32,34,35,38-48 Studies inwhich fracture was found to be more frequentthan secondary caries are rare.16,33,49

In our study, amalgam restorations performedconsistently better than composite restorations,independently of the type of tooth, number ofrestored surfaces or size of the restoration. Lein-felder50 affirmed that the evolution of caries adja-cent to composite restorations was faster thanthat of caries adjacent to amalgam restorationsbecause some resin-based composite componentshave the ability to promote bacterial growth.Additional explanation for these findings maycome from the age group of the subjects enrolledin this study. Poor oral hygiene habits, which arecharacteristic of teenagers, may have beenresponsible for the elevated rates of secondarycaries in the more susceptible composite restora-tions. One study that compared longevity ofamalgam and composite restorations in teenagersand adults found that the longevity of the restora-tions placed in teenagers was, on average, five to



0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0

0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7

SECONDARY CARIES

YEARS SINCE RESTORATION

PR

OP

OR

TIO

N S

UR

VIV

ING

PR

OP

OR

TIO

N S

UR

VIV

ING

FRACTURES

YEARS SINCE RESTORATION

Amalgam Restorations

Composite Restorations

Amalgam Restorations

Composite Restorations

A B

Figure 2. Survival curves. A. Secondary caries. B. Fractures.


six years shorter than that of restorations inadults.4

CONCLUSIONS

After seven years of follow-up, the mean annualfailure rates of posterior composite restorationswas significantly higher than those of amalgamrestorations, and the corresponding survival ratewas significantly lower. This was true irrespec-tive of the arch, type of tooth, number of restoredsurfaces or restoration size. The main reason forfailure was secondary caries, followed by fracture.This was true for both amalgam and compositerestorations. The overall risk of failure due to sec-ondary caries was 3.5 times higher in compositerestorations than in amalgam restorations. Therisk of failure due to fracture was equal inamalgam and composite restorations.

On the basis of the results of this study andwithin its limitations, posterior amalgam restora-tions performed better than composite restora-tions. The difference in performance was accentu-ated in restorations with more than three surfacesrestored and in large restorations. When one takesinto consideration that certain factors such as poortraining in adhesive procedures, a lack ofadequate equipment and insufficient conditionsrequired to execute highly technique-sensitivecomposite restorations (which still is a reality inmany countries and regions of the world),amalgam more often seems to be preferable toresin-based composites for use in direct restora-tion of large posterior teeth, particularly when therestorations are large. This may be true even inenvironments in which human and technical con-ditions exist to produce high-quality dental resto-rations, as was the case in our study.

The research described in this article was funded by the NationalInstitute of Dental and Craniofacial Research, Cooperative AgreementU01 DE11894.

The authors wish to acknowledge the many contributions by otherson the team of the Casa Pia Study of the Health Effects of DentalAmalgams in Children, as well as the cooperation of the administra-tion, faculty and especially the students of the Casa Pia school system,Lisbon, Portugal.

1. Mjr IA, Jokstad A. Five-year study of Class II restorations in per-manent teeth using amalgam, glass polyalkenoate (ionomer) cermentand resin-based composite materials. J Dent 1993;21(6):338-43.

2. Mjr IA, Moorhead JE. Selection of restorative materials, reasonsfor replacement, and longevity of restorations in Florida. J Am CollDent 1998;65(3):27-33.

3. Mair LH. Ten-year clinical assessment of three posterior resincomposites and two amalgams. Quintessence Int 1998;29(8):483-90.

4. Mjr IA, Dahl JE, Moorhead JE. Age of restorations at replace-ment in permanent teeth in general dental practice. Acta OdontolScand 2000;58(3):97-101.

5. Johnson GH, Bales DJ, Gordon GE, Powell LV. Clinical perform-

ance of posterior composite resin restorations. Quintessence Int1992;23(10):705-11.

6. Norman RD, Wright JS, Rydberg RJ, Felkner LL. A 5-year studycomparing a posterior composite resin and an amalgam. J ProsthetDent 1990;64(5):523-9.

7. Drake CW. A comparison of restoration longevity in maxillary andmandibular teeth. JADA 1988;116(6):651-4.

8. Kolker JL, Damiano PC, Jones MP, et al. The timing of subsequenttreatment for teeth restored with large amalgams and crowns: factorsrelated to the need for subsequent treatment. J Dent Res 2004;83(11):854-8.

9. Jokstad A, Mjr IA. Replacement reasons and service time of class-II amalgam restorations in relation to cavity design. Acta OdontolScand 1991;49(2):109-26.

10. Lucarotti PS, Holder RL, Burke FJ. Outcome of direct restora-tions placed within the general dental services in England and Wales(Part 1): variation by type of restoration and re-intervention. J Dent2005;33(10):805-15.

11. Wahl MJ, Schmitt MM, Overton DA, Gordon MK. Prevalence ofcusp fractures in teeth restored with amalgam and with resin-basedcomposite. JADA 2004;135(8):1127-32.

12. Mjr IA. Amalgam and composite resin restorations: longevityand reasons for replacement. Paper presented at: International Sympo-sium on Criteria for Placement and Replacement of Dental Restora-tions; Oct. 19-21, 1989, Lake Buena Vista, Fla.

13. Tobi H, Kreulen CM, Vondeling H, van Amerongen WE. Cost-effectiveness of composite resins and amalgam in the replacement ofamalgam Class II restorations. Community Dent Oral Epidemiol1999;27(2):137-43.

14. DeRouen TA, Leroux BG, Martin MD, et al. Issues in design andanalysis of a randomized clinical trial to assess the safety of dentalamalgam restorations in children. Control Clin Trials 2002;23(3):301-20.

15. DeRouen TA, Martin MD, Leroux BG, et al. Neurobehavioraleffects of dental amalgam in children: a randomized clinical trial.JAMA 2006;295(15):1784-92.

16. Opdam NJ, Loomans BA, Roeters FJ, Bronkhorst EM. Five-yearclinical performance of posterior resin composite restorations placed bydental students. J Dent 2004;32(5):379-83.

17. Collins CJ, Bryant RW, Hodge KL. A clinical evaluation of pos-terior composite resin restorations: 8-year findings. J Dent 1998;26(4):311-7.

18. Burke FJ, Cheung SW, Mjr IA, Wilson NH. Restorationlongevity and analysis of reasons for the placement and replacement ofrestorations provided by vocational dental practitioners and theirtrainers in the United Kingdom. Quintessence Int 1999;30(4):234-42.

19. Raskin A, Michotte-Theall B, Vreven J, Wilson NH. Clinicalevaluation of a posterior composite 10-year report. J Dent 1999;27(1):13-9.

20. Ernst CP, Martin M, Stuff S, Willershausen B. Clinical perform-ance of a packable resin composite for posterior teeth after 3 years.Clin Oral Investig 2001;5(3):148-55.

21. Geurtsen W, Schoeler U. A 4-year retrospective clinical study ofClass I and Class II composite restorations. J Dent 1997;25(3-4):229-32.

22. Oberlander H, Hiller KA, Thonemann B, Schmalz G. Clinicalevaluation of packable composite resins in Class-II restorations. ClinOral Investig 2001;5(2):102-7.

23. Rasmusson CG, Kohler B, Odman P. A 3-year clinical evaluationof two composite resins in class-II cavities. Acta Odontol Scand 1998;56(2):70-5.

24. Turkun LS, Aktener BO, Ates M. Clinical evaluation of differentposterior resin composite materials: a 7-year report. Quintessence Int2003;34(6):418-26.

25. Barnes DM, Blank LW, Thompson VP, Holston AM, Gingell JC. A 5- and 8-year clinical evaluation of a posterior composite resin. Quin-tessence Int 1991;22(2):143-51.

26. Smales RJ, Gerke DC, White IL. Clinical evaluation of occlusalglass ionomer, resin, and amalgam restorations. J Dent 1990;18(5):243-9.

27. Roulet JF. Longevity of glass ceramic inlays and amalgam:results up to 6 years. Clin Oral Investig 1997;1(1):40-6.

28. Plasmans PJ, Creugers NH, Mulder J. Long-term survival ofextensive amalgam restorations. J Dent Res 1998;77(3):453-60.

29. Mertz-Fairhurst EJ, Curtis JW Jr, Ergle JW, Rueggeberg FA,Adair SM. Ultraconservative and cariostatic sealed restorations:results at year 10. JADA 1998;129(1):55-66.

30. Opdam NJ, Bronkhorst EM, Roeters JM, Loomans BA. A retro-spective clinical study on longevity of posterior composite and amalgam


782 JADA, Vol. 138 http://jada.ada.org June 2007Copyright 2007 American Dental Association. All rights reserved.

restorations. Dent Mater 2007;23(1):2-8.31. Tyas MJ. Placement and replacement of restorations by selected

practitioners. Aust Dent J 2005;50(2):81-9.32. Forss H, Widstrom E. Reasons for restorative therapy and the

longevity of restorations in adults. Acta Odontol Scand 2004;62(2):82-6.

33. Van Nieuwenhuysen JP, DHoore W, Carvalho J, Qvist V. Long-term evaluation of extensive restorations in permanent teeth. J Dent2003;31(6):395-405.

34. Palotie U, Vehkalahti M. Reasons for replacement and the age offailed restorations in posterior teeth of young Finnish adults. ActaOdontol Scand 2002;60(6):325-9.

35. Mjr IA. The reasons for replacement and the age of failed resto-rations in general dental practice. Acta Odontol Scand 1997;55(1):58-63.

36. Burgess JO, Walker R, Davidson JM. Posterior resin-based com-posite: review of the literature. Pediatr Dent 2002;24(5):465-79.

37. Hansen EK, Asmussen E. In vivo fractures of endodonticallytreated posterior teeth restored with enamel-bonded resin. Endo DentTraumatol 1990;6(5):218-25.

38. Manhart J, Chen H, Hamm G, Hickel R. Buonocore MemorialLecture: review of the clinical survival of direct and indirect restora-tions in posterior teeth of the permanent dentition. Oper Dent2004;29(5):481-508.

39. Brunthaler A, Konig F, Lucas T, Sperr W, Schedle A. Longevityof direct resin composite restorations in posterior teeth. Clin OralInvestig 2003;7(2):63-70.

40. Mjr IA, Shen C, Eliasson ST, Richter S. Placement and replace-

ment of restorations in general dental practice in Iceland. Oper Dent2002;27(2):117-23.

41. Hickel R, Manhart J. Longevity of restorations in posterior teethand reasons for failure. J Adhes Dent 2001;3(1):45-64.

42. Deligeorgi V, Mjr IA, Wilson NH. An overview of reasons for theplacement and replacement of restorations. Prim Dent Care 2001;8(1):5-11.

43. Mjr IA, Moorhead JE, Dahl JE. Reasons for replacement of resto-rations in permanent teeth in general dental practice. Int Dent J2000;50(6):361-6.

44. Kohler B, Rasmusson CG, Odman P. A five-year clinical evalu-ation of Class II composite resin restorations. J Dent 2000;28(2):111-6.

45. Hickel R, Manhart J, Garcia-Godoy F. Clinical results and newdevelopments of direct posterior restorations. Am J Dent 2000;13(spe-cial number):41D-54D.

46. Deligeorgi V, Wilson NH, Fouzas D, Kouklaki E, Burke FJ, MjrIA. Reasons for placement and replacement of restorations in studentclinics in Manchester and Athens. Eur J Dent Educ 2000;4(4):153-9.

47. Wendt LK, Koch G, Birkhed D. Replacements of restorations inthe primary and young permanent dentition. Swed Dent J 1998;22(4):149-55.

48. Mjr IA, Toffenetti F. Placement and replacement of resin-basedcomposite restorations in Italy. Oper Dent 1992;17(3):82-5.

49. Van Dijken JW, Sunnegardh-Gronberg K. A four-year clinicalevaluation of a highly filled hybrid resin composite in posterior cavities.J Adhes Dent 2005;7(4):343-9.

50. Leinfelder KF. Do restorations made of amalgam outlast thosemade of resin-based composite? JADA 2000;131(8):1186-7.


JADA, Vol. 138 http://jada.ada.org June 2007 783Copyright 2007 American Dental Association. All rights reserved.

Composite Failure

Documents

Transcript of Composite Failure