Longevity of posterior composite restorations: a systematic review and meta-analysis.

The aim of this meta-analysis, based on individual participant data from several studies, was to investigate the influence of patient-, materials-, and tooth-related variables on the survival of posterior resin composite restorations. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we conducted a search resulting in 12 longitudinal studies of direct posterior resin composite restorations with at least 5 years' follow-up. Original datasets were still available, including placement/failure/censoring of restorations, restored surfaces, materials used, reasons for clinical failure, and caries-risk status. A database including all restorations was constructed, and a multivariate Cox regression method was used to analyze variables of interest [patient (age; gender; caries-risk status), jaw (upper; lower), number of restored surfaces, resin composite and adhesive materials, and use of glass-ionomer cement as base/liner (present or absent)]. The hazard ratios with respective 95% confidence intervals were determined, and annual failure rates were calculated for subgroups. Of all restorations, 2,816 (2,585 Class II and 231 Class I) were included in the analysis, of which 569 failed during the observation period. Main reasons for failure were caries and fracture. The regression analyses showed a significantly higher risk of failure for restorations in high-caries-risk individuals and those with a higher number of restored surfaces.

Treatment effect of ozone and fluoride varnish application on occlusal caries in primary molars: a 12-month study.

AIM: The aim of this study is to evaluate the effect of ozone and fluoride varnish on occlusal caries in primary molars in a split-mouth study. MATERIALS AND METHODS: Caries risk was estimated by treating Public Dental Health Service dentists. Children with occlusal caries with Ekstrand index scores ≤3 (VI ≤3) were included. Selection of caries lesions was discontinued for ethical reasons due to non-acceptable clinical results during the follow-up. In the continued evaluation pairs of teeth with non-cavitated caries lesions, Ekstrand score ≤2a (VI ≤2) were selected. Fifty pairs of carious primary molars were included, 18 boys and 15 girls (mean 4.7 years, range 3-8). At baseline, the lesions were assessed by visual inspection (VI) and laser-induced fluorescence (LF), in each pair to treatment with 40 s ozone (HealOzone(TM), 2,100 ppm) or fluoride varnish Duraphat®. The treatments and evaluations were repeated at 3, 6 9 months and evaluations only at 12 months. RESULTS: Medium-high caries risk was observed in VI ≤3 children and low-medium risk in VI ≤2a children. In the 15 pairs VI ≤3 lesions, 8 treated with ozone and 9 with fluoride progressed to failure. In the 35 pairs VI ≤2a lesions, one lesion failed. Median baseline LF values in the VI ≤3 group were 76 and 69, for ozone and fluoride lesions, respectively, and 21 and 19 in the VI ≤2a group. At 12 months, LF values in the VI ≤2a group were 15 and 18. No improvement or difference in LF values was found over time between the caries lesions treated with ozone or fluoride. CONCLUSIONS: Neither ozone nor fluoride varnish treatments stopped the progression of caries in cavitated lesions. In low and medium caries risk children, non-cavitated lesions following both treatments showed slight or no progression. The use of ozone or fluoride varnish treatments in this regime as caries preventive method, added to the daily use of fluoridated toothpaste, to arrest caries progression in primary molars must therefore be questioned.

Antibacterial effect of ozone on cariogenic bacterial species.

OBJECTIVE: The aim was to evaluate the antibacterial effect of ozone on cariogenic bacterial species with and without the presence of saliva and a possible effect on the salivary proteins. METHODS: Suspensions of Actinomyces naeslundii (ACTCC 12104(T)), Lactobacilli casei (N CTC 151) and Streptococcus mutans (NCTC 10449), in salt buffer or in saliva, were exposed to ozone gas delivered by the ozone generator Healozone 2130C. Aliquots of the suspensions were taken after 10, 30 and 60s ozone exposures and cultivated on agar plates. Initial number of bacteria per ml was 8.0 x 10(7) (SD 2.2 x 10(7)) (A. naeslundii), 1.0 x 10(8) (SD 3.1 x 10(6)) (L. casei) and 1.0 x 10(8) (SD 7.0 x 10(5)) (S. mutans), respectively. The proteins were separated by SDS electrophoresis and visualized by silver staining. RESULTS: In salt buffer 92%, 73% and 64% of the initial numbers of A. naeslundii, S. mutans and L. casei, respectively, were killed already after 10s ozone exposure, while approximately 99.9% of the bacteria were dead after a 60s exposure. After 10 and 30s, but not after 60s exposure to ozone, S. mutans and L. casei were less efficiently killed in saliva compared to the salt buffer. Various saliva proteins were degraded by ozone after a 60s exposure. CONCLUSIONS: The cariogenic species S. mutans, L. casei and A. naeslundii were almost eliminated following 60s of ozone treatment. This killing was reduced in the presence of saliva although increasing the ozone application time to 60s overcame these reductants in saliva. Detection of altered salivary proteins indicates that saliva components constitute additional targets for ozone.

Mutans streptococci and lactobacilli in plaque on a leucite-reinforced dental ceramic and on a calcium aluminate cement.

In this in vivo study, the proportions of mutans streptococci and lactobacilli in plaque were examined (1) on proximal surfaces of bonded, leucite-reinforced ceramic crowns and (2) on class V restorations of calcium aluminate cement (CAC). The examined proportions were intraindividually compared with those of resin composite and enamel. Mutans streptococci and lactobacilli in samples from plaque that was accumulated for 10 days on the following surfaces were determined by cultivation on blood agar plates and species-selective plates: (1) proximal leucite-reinforced ceramic crown, class II composite and enamel (n=11); and (2) class V restoration of CAC and composite, and enamel (n=17). Mutans streptococci and lactobacilli in the samples were distributed in three groups: 0, >0-1, and >1% of total bacteria. The surfaces with detected mutans streptococci were similarly distributed between the materials and enamel. The highest proportion of mutans streptococci and lactobacilli were observed on ceramic followed by composite and enamel. A higher proportion of lactobacilli, but not of mutans streptococci, was detected on enamel compared to CAC and composite. However, no significant differences were found between the surfaces. Conclusively, the materials investigated did not show different relative proportions of mutans streptococci and lactobacilli in plaque, compared to enamel.

Effect of a hydroxyl ion-releasing composite resin on plaque acidogenicity.

The aim of this in vivo study was to evaluate the neutralizing capacity, registered as change of plaque acidogenicity, on aged proximal restorations of an ion-releasing composite resin (IRCR), which releases hydroxyl, calcium, and fluoride ions at low pH. Twenty patients, with a mean age of 63 years (range 43-85), participated. All had one aged proximal IRCR restoration (mean age 15 months) and one nonrestored enamel surface to make an intraindividual comparison possible. The neutralizing effect of the IRCR was evaluated by measuring plaque pH, using the microtouch method, after a mouthrinse with 10% sucrose. The plaque pH measurements were repeated 1.5 years later on the IRCR (mean age 34 months), the enamel surfaces and a universal hybrid composite resin (CR). At both 15 and 34 months, the plaque on the IRCR surfaces showed the least acidogenic potential for the whole 60-min time interval. The largest differences between the IRCR, CR and enamel were found during the first 15 min. At 15 months, the total areas under the plaque pH curve (AUC(5.7) and AUC(6.2)) differed significantly between the IRCR and enamel surfaces for the time periods 0-5 min and 5-15 min. At 34 months, significant differences were found between IRCR and CR at the 0- to 5-min time period. It can be concluded that IRCR restorations countered the plaque pH fall and maintained it at levels where less enamel and dentin demineralization can occur.

In vivo interfacial adaptation of class II resin composite restorations with and without a flowable resin composite liner.

The aim of this study was to evaluate in vivo the interfacial adaptation of class II resin composite restorations with and without a flowable liner. In 24 premolars scheduled to be extracted after 1 month, 48 box-shaped, enamel-bordered class II cavities were prepared and restored with a flowable liner (FRC, Tetric Flow/Tetric Ceram/Syntac Single-Component) or without (TRC), cured with three different curing modes: soft start and 500- or 700-mW/cm2 continuous irradiation. Interfacial adaptation was evaluated by quantitative scanning electron microscopic analysis using replica method. Gap-free adaptation in the cervical enamel (CE) was observed for FRC and TRC in 96.2 and 90.2%, for the dentin (D) in 63.6 and 64.9%, and for occlusal enamel (OE) in 99.7 and 99.5%, respectively. The difference between the two restorations was not statistically significant (ns). Significant better adaptation was observed for OE than CE and D (p<0.01), and for CE than D (p<0.01). Gap-free adaptation with the soft-start and 500- and 700-mW/cm2 continuous-curing modes was observed for CE: 88.7%, 92.7%, 97.9% (ns); OE: 99.8%, 98.7%, 100% (ns); and D: 64.0%, 63.9%, and 64.6% (ns), respectively. It can be concluded that neither the use of flowable resin composite liner nor the curing mode used influenced the interfacial adaptation.

Evaluation of a skin reference electrode used for intraoral pH measurements in combination with a microtouch electrode.

The aim of this in vivo study was to evaluate a skin reference electrode used with the microtouch method for measurements of intraoral pH. Registrations of plaque pH were made in ten healthy subjects before and up to 30 min after a mouth rinse with 10% sucrose. A microtouch electrode (Beetrode) was used in combination with two different half-cell reference electrodes: a skin silver-silver chloride plate or a glass capillary reference electrode. The skin electrode was placed on the subject's forearm in combination with a conductive gel. The glass capillary electrode, together with one of the test subject's fingers, was placed in a beaker with a 3 M KCl solution. High agreement between the two reference systems resulted in almost identical plaque-pH curves (r=0.97). The mean differences between the two systems at the different time points varied between -0.15 and +0.01 pH units, with an overall mean pH difference of 0.03 pH units. It can be concluded that the use of the skin reference electrode may very well replace the traditional reference electrode used in combination with a microtouch electrode for intraoral use.

Interfacial adaptation of a calcium aluminate cement used in class II cavities, in vivo.

The aim of this in vivo study was to evaluate the interfacial marginal adaptation of a calcium aluminate cement, Doxadent (DD), and to compare it intra-individually with a resin composite, Tetric Ceram/Syntac Single-Component (TC/SS), in Class II cavities. Sixteen Class II box-shaped, enamel-bordered cavities were prepared in eight premolars scheduled to be extracted after 1 month's service for orthodontic reasons. The interfacial marginal adaptation (internal surfaces) of the restorations was evaluated by a quantitative scanning electron microscope analysis using a replica method. DD showed a statistically significant, lower degree of gap-free adaptation to enamel compared with TC/SS: 84% vs. 93%. To dentin, DD showed a significantly better adaptation than TC/SS: 72% vs. 49%. A high frequency of enamel fractures perpendicular to the margins was observed for the DD restorations, which may be explained by an expansion of the calcium-aluminate cement. It can be concluded that DD showed a better adaptation to dentin while TC/SS showed a better adaptation to enamel. The dimensional changes of DD have to be investigated before clinical use can be recommended.

Surface roughness of a novel "ceramic restorative cement" after treatment with different polishing techniques in vitro.

The aim of this study was to determine the surface roughness of a novel calcium aluminate cement (CAC) intended for posterior restorations after treatment with different polishing devices in vitro. Forty-eight CAC specimens were polished with diamond burs at 15550 rpm or 27000 rpm, Sof-Lex discs, Jiffy points, Shofu silicone points, and Aaba universal polisher. Amalgam specimens were polished with Shofu silicone points and used as reference. Roughness was measured using a profilometer. The smoothest CAC surface was observed after use of the fine Sof-Lex disc (roughness average [Ra] 0.26 microm). Diamond burs at higher speed, points, and polisher gave rather similar results (Ra 0.58-0.72 microm). An increase in surface roughness could be seen from using diamond burs at lower speed (Ra 2.3 microm). Extra fine Shofu points and Sof-Lex discs as a final step reincreased surface roughness. Polished amalgam showed the smoothest surface in the study (Ra 0.17). It can be concluded that the smoothest CAC surfaces were obtained with the fine Sof-Lex discs. Different polishing points and diamond burs at higher speed, which are suitable polishing devices for posterior restorations, also gave relatively smooth surfaces.

Restorations with extensive dentin/enamel-bonded ceramic coverage. A 5-year follow-up.

The durability of restorations with extensive dentin/enamel-bonded posterior partial and complete ceramic coverages were investigated. The effect of luting with a dual-cured and a self-cured luting agent was also studied. In 110 patients, 182 ceramic coverages (IPS Empress) were placed. In 58 restorations, Syntac was used in combination with the dual-cured resin composite Variolink. In the other restorations luted with the chemically cured resin composite Bisfil 2B, 25 were bonded with Gluma, 57 with Allbond 2, and 42 with Syntac. Of the 182 ceramics, 13 (7.1%) were assessed as non-acceptable after a mean observation period of 4.9 yr (range 4.3 7.5 yr). The reasons for failure were fracture (5), lost restorations (4), secondary caries (3) and endodontic treatment (1). No significant differences in failure rate were seen between the two luting agents or between the three dentin-bonding agents. Ceramic coverages placed on non-vital teeth failed in 9.7% of cases (3/31) and on vital teeth in 6.6% (10/151). The success rate of the dentin-enamel-bonded ceramic coverages reduces the need for a traditional full-coverage therapy and/or post or pin(s) and core placement. The technique investigated showed many clinical advantages such as less destruction of healthy tissue, and avoidance of endodontic treatment and/or deep cervical placement of restoration margins.

Analysis of factors affecting failure of glass cermet tunnel restorations in a multi-center study.

The aim of this study was to analyze factors influencing the failures of tunnel restorations performed with a glass cermet cement (Ketac Silver). Caries activity, lesion size, tunnel cavity opening size, partial or total tunnel, composite lamination or operating time showed no significant correlation to failure rate. Twelve dentists in eight clinics clinically experienced and familiar with the tunnel technique placed 374 restorations. The occlusal sections of fifty percent of the restorations were laminated with hybrid resin composite. The results of the yearly clinical and radiographic evaluations over the course of 3 years were correlated to factors that could influence the failure rate using logistic regression analysis. At the 3-year recall a cumulative number of 305 restorations were available. The cumulative replacement rate was 20%. The main reasons for replacement were marginal ridge fracture (14%) and dentin caries (3%). Another 7% of the restorations which had not been replaced were classified as failures because of untreated dentin caries. The only significant variable observed was the individual failure rate of the participating dentists varying between 9 and 50% (p=0.013).

Durability of new restorative materials in Class III cavities.

PURPOSE: The aim of this study was to evaluate the durability of new hybrid tooth-colored restorative materials in Class III cavities. MATERIALS AND METHODS: 154 large-sized Class III restorations were placed in 50 patients. The patients received one of each of the three following materials: a resin composite (RC, Pekafill), a polyacid-modified resin composite (PMRC, Dyract) (compomer), and a resin-modified glass-ionomer cement (RMGIC, Fuji II LC). The restorations were evaluated yearly with slightly modified USPHS criteria. RESULTS: Of 141 restorations evaluated at 6 years, 16 were estimated as unacceptable. Seven of these were replaced: 2 fractures, 3 recurrent caries, and 2 due to unacceptable color match. In 9 other restorations with unacceptable color match, the patients did not request replacements. No significant differences were seen between the materials concerning the occurrence of recurrent caries. The RC showed significantly better color match. Significantly higher surface roughness was seen for the aged RMGIC restorations. Fracture of the incisal enamel corner was observed contiguous to ten restorations in the 6-year evaluation. None of the restorative techniques resulted in postoperative sensitivity or loss of vitality. CONCLUSION: The resin composite showed the best durability.

Interfacial adaptation of a Class II polyacid-modified resin composite/resin composite laminate restoration in vivo.

The aim of this in vivo study was to evaluate the interfacial adaptation of Class II resin composite open sandwich restorations with a polyacid-modified resin composite as a stress-absorbing layer (PMRC/RC). Twenty Class II box-shaped, enamel-bordered cavities were prepared in 10 premolars scheduled to be extracted for orthodontic reasons. An open PMRC/RC sandwich restoration was placed in 1 of the cavities of each tooth. The first layer, PMRC, in the proximal box extended to the periphery in the cervical part of the cavity. The following RC layers were placed with a horizontally incremental technique. The PMRC was excluded from the control cavity. The teeth were extracted after 1 month and the interfacial adaptation of the restorations was studied with quantitative scanning electron microscope analysis using a replicate technique. Gap-free interfacial adaptation was observed for the PMRC/RC and RC restorations in cervical enamel in 97% and 73%, respectively (P = 0.006). The gap-free scores for dentin were 87% and 64%, respectively (P = 0.022). Excellent interfacial adaptation was observed in both groups for the occlusal enamel 99% and 100%, respectively. The adaptation to occlusal enamel for the direct resin composite restorations was significantly better than to dentin or cervical enamel. A higher frequency of enamel fractures was observed parallel to the cervical margins compared to the occlusal. No dentin fractures were observed in the experimental groups. The PMRC/RC sandwich technique showed a statistically significant improved interfacial adaptation to dentin and cervical enamel in Class II enamel-bordered cavities. The clinical significance of the differences has to be evaluated.

Clinical evaluation of three adhesive systems in class V non-carious lesions.

OBJECTIVES: The purpose of this study was to evaluate the clinical retention of three new adhesive systems in non-carious cervical lesions during a 3-year period. METHODS: The adhesive systems, a three-step (EBS/Pertac Hybrid), a one-bottle resin bonding agent (One-Step/Pertac Hybrid) and a resin-modified glass ionomer cement (Fuji II LC) were placed in 148 non-carious cervical lesions, 87 with sclerotic dentin and 61 non-sclerotic. Of the sclerotic lesions treated with the two resin bonding systems, 37 were slightly roughened with a diamond bur before conditioning. The restorations were evaluated every 6 months during a 3-year period. RESULTS: All except six restorations were evaluated during the 3 years. The cumulative loss rates for EBS, One-Step and Fuji II LC were, at 1 year: 2, 24 and 2% and at 3 years: 10, 49, 7%, respectively. The one-bottle adhesive showed significantly more failures. The five lost EBS restorations were found in non-sclerotic lesions, while the three lost Fuji II LC restorations had been placed in sclerotic lesions. For the One-Step material the loss frequency for non-sclerotic versus sclerotic lesions was 31.8 and 65.2%, respectively. Slight roughening of the sclerotic dentin surfaces with a diamond bur did not increase retention of the restorations. SIGNIFICANCE: The three-step resin adhesive and the RMGIC showed clinically acceptable retention rates, while a high failure rate was registered for the one-bottle adhesive.

An 8-year evaluation of sintered ceramic and glass ceramic inlays processed by the Cerec CAD/CAM system.

The purpose of this study was to evaluate Cerec CAD/CAM inlays processed of two industrially made machinable ceramics during an 8-yr follow-up period. Each of 16 patients received two similar ceramic inlays. Half the number of the inlays were made of a feldspathic (Vita Mark II) and the other of a glass ceramic (Dicor MGC) block. The inlays were luted with a dual resin composite and evaluated clinically using modified USPHS criteria at baseline, 8 months, 2, 3, 5, 6 and 8 yr, and indirectly using models. At baseline, 84% of the inlays were estimated as optimal and 16% as acceptable. Postoperative sensitivity was reported by one patient for 8 months. Of the 32 inlays evaluated during the 8 yr, 3 failed due to fracture of the material. No secondary caries was found adjacent to the inlays. No significant differences in the clinical performance were found between inlays made of the two ceramics. It can be concluded that the CAD/CAM inlays processed of the two ceramics functioned well during the 8-yr follow-up period.

Direct resin composite inlays/onlays: an 11 year follow-up.

OBJECTIVES: The aim of this study was to present an 11-year assessment of direct resin composite inlays/onlays. METHODS: One-hundred Class II direct resin composite inlays and 34 direct resin composite restorations were placed in 40 patients. The restorations were evaluated clinically, according to modified USPHS criteria, annually over a 11-year period. RESULTS: Of the 96 inlays/onlays and 33 direct restorations evaluated at 11 years, 17. 7% in the inlay/onlay group and 27.3% in the direct restorations group were assessed as unacceptable. The differences in longevity were not statistically significant. The main reasons for failure for the inlays/onlays and direct restorations were fracture (8.3 and 12. 1%, respectively), occlusal wear in contact areas (4.2 and 6.1%, respectively) and secondary caries (4.2 and 9.1%, respectively). Eight of the non-acceptable inlays/onlays and five of the direct restorations were replaced, while the other ones were repaired with resin composite. Unacceptable wear was observed in occlusal contact areas of six restorations, in patients who were severe bruxers. For the other restorations occlusal wear was not found to be a clinical problem and no difference was observed between the inlays/onlays and direct composite restorations. The marginal adaptation of the inlays/onlays was still good at the end of the study. Ditching was only observed in a few inlays. A higher failure rate was observed in molar teeth than in premolar teeth. CONCLUSIONS: Good durability was observed for the direct resin composite inlay/onlay technique. Excellent marginal adaptation and low frequency of secondary caries in patients with high caries risk were shown. No apparent improvement of mechanical properties was obtained by the secondary heat treatment of the inlays. Also, the difference in failure rate between the resin composite direct technique and the inlay technique was not large, indicating that the more time-consuming and expensive inlay technique may not be justified. The direct inlay/onlay technique is recommended to be used in Class II cavities of high caries risk patients with cervical marginal placed in dentin.

Clinical performance of pressed ceramic inlays luted with resin-modified glass ionomer and autopolymerizing resin composite cements.

STATEMENT OF PROBLEM: Several in vitro studies have been published showing the incapability of the chemical cure of dual-cured resin composite luting agents to compensate for absence of visible light activation. PURPOSE: This study evaluated and compared pairs of Empress ceramic inlays luted with 2 chemical-cured luting agents: a resin-modified glass ionomer cement and a resin composite within individual patients. MATERIAL AND METHODS: Seventy-nine ceramic inlays were placed in class II cavities in 29 patients. In each patient half of the inlays were luted with a resin-modified glass ionomer cement and the other half with a chemical-cured resin composite cement. The inlays were evaluated clinically, according to modified USPHS criteria, at baseline, after 6 months, and 1 and 2 years. RESULTS: No failed inlays were observed during the 2-year follow-up. A slight ditching of the cement margins was observed in both luting groups. No significant difference was seen between the 2 luting techniques. CONCLUSION: IPS Empress inlays luted with both chemical-cured luting agents functioned satisfactorily in the short-term follow-up. Longer observation periods are necessary to evaluate the long-term clinical behavior of both luting techniques.

Longevity of new hybrid restorative materials in class III cavities.

The aim of this study was to evaluate the durability of two new tooth-colored restorative materials, a polyacid-modified resin composite (compomer) and a resin-modified glass ionomer cement. In an intra-individual comparison with a resin composite, the materials were studied during a 5-yr period. In 50 patients, 154 large class III restorations were placed. Most patients received one of each of the three materials. The restorations were evaluated by modified United States Public Health Service (USPHS) criteria. Of the 144 restorations evaluated, 15 were found unacceptable. No significant differences were seen between the materials concerning recurrent caries incidence. The resin composite showed a significantly better color match. No difference was found between resin composite fillings in enamel- or total-etched cavities. Significantly higher surface roughness was found in the aged resin-modified glass ionomer cement restorations. Fracture of the incisal enamel corner was observed adjacent to 10 restorations. None of the three total-etch techniques resulted in postoperative sensitivity or loss of vitality. All the restorative techniques showed a low failure rate during the 5-yr follow-up.

Longevity of extensive class II open-sandwich restorations with a resin-modified glass-ionomer cement.

Several new techniques have been introduced for use in the esthetic restoration of posterior cavities to substitute for the presumed toxicity of amalgam. Composite-laminated glass-ionomer cement restorations, the sandwich technique, have been recommended for caries-risk patients. Clinical evaluation of the use of conventional glass-ionomer cements in the open-sandwich restoration has shown a high failure rate. The aim of this study was to evaluate the durability and cariostatic effect of a modified open-sandwich restoration utilizing a resin-modified glass-ionomer cement (RMGIC) in large cavities. The materials consisted of 274 mostly extensive Class II Vitremer/Z100 restorations performed by four dentists in 168 adults. Six experimental groups were investigated. In four groups a thick and in two groups a thin layer of cement was placed. Cavity conditioning before application of the RMGIC self-etching primer was done in 3 groups with polyacrylic acid and in one group with maleic acid; in two groups, only water rinsing was performed. The restorations were evaluated at baseline and after 6, 12, 24, and 36 months according to modified USPHS criteria (van Dijken, 1986). After 3 years, 239 restorations were evaluated. Twelve (5%) were estimated as non-acceptable. Two were replaced, and seven were repaired with resin composite. Tooth fractures were observed in 2.5%. Slight erosion of the RMGIC part was seen in 4%, and in one case operative treatment was indicated. Post-operative sensitivity was reported for 9 teeth. Forty-three percent of the patients were considered as caries-risk patients. Only one restoration showed secondary caries. The three-year results indicated that the modified open-sandwich restoration is an appropriate alternative to amalgam including extensive restorations.

Five year evaluation of class III composite resin restorations in cavities pre-treated with an oxalic- or a phosphoric acid conditioner.

An oxalic acid solution has been proposed as a conditioning agent for resin composite restorations in two commercial adhesive systems. The durability of 163 class III restorations, including 12 class IV restorations, in cavities pre-treated with an oxalic acid total etch technique or an enamel etch with phosphoric acid was studied. Each of 52 patients received at least one of each of three experimental restorations. The restorations were evaluated yearly with slightly modified United States Public Health Service (USPHS) criteria. After 5 years 95% of the restorations were evaluated as acceptable. Reasons for failure were the fracture of four fillings, including three class IV, secondary caries contiguous to two fillings and a non-acceptable colour match for one restoration. For eight class III restorations a fracture of the incisal tooth structure was registered. No differences were seen between the three experimental restorations.