A randomized controlled 30 years follow up of three conventional resin composites in Class II restorations

Abstract

Objective

The aim of this 30 year randomized controlled study was to evaluate, by intraindividual comparisons, the durability of three conventional resin composites in Class II restorations.

Methods

Each of 30 participants, 21 female and 9 male (mean age 30 years, range 20–43), received at least three (one set) as similar as possible Class II restorations of moderate size. After cavity preparation, the three cavities were chosen at random to be restored with two chemical-cured (P10, Miradapt) and one light-cured resin composite (P30). A chemical-cured enamel bonding agent was applied after etching of the enamel. The chemical-cured resin composites were placed in bulk and the light-cured in increments. One operator placed 99 restorations (33 sets). The restorations were evaluated with slightly modified USPHS criteria at baseline, 2, 3, 5, 10, 15, 20 and 30 years. Statistical analyses were performed by the Kaplan–Meier, log-rank test and Cox regression analyses.

Results

After 30 years, 5 participants with 15 restorations (15%) could not be evaluated during the whole evaluation. Seven participants were considered as caries risk and eight participants as having active parafunctional habits. Postoperative sensitivity was observed in 24 teeth. In total 28 restorations, 9 P10, 12 P30 and 7 Miradapt restorations failed during the 30 years. The main reasons for failure were secondary caries (39.2%) and material fracture (35.7%). Sixty-four percent of the secondary caries lesions were found in high caries risk participants and 70% of the material fractures occurred in participants with active parafunctional habits. The overall success rate at 30 years was 63%, with an annual failure rate of 1.1%. 68–81% of the restorations showed non-acceptable color match. No statistical significant difference in survival rate was found between the three resin composites ( p = 0.45). The variables tooth type, cavity size, age, and gender of the participants did not significantly affect the probability of failure.

Significance

The three conventional resin composites showed good clinical performance during the 30 year evaluation. The chemical cured resin composites showed better performance than the light-cured composite.

Introduction

Since the development of the first resin based restorative materials with inorganic fillers in the seventies, their popularity have increased during the years due to a growing demand for aesthetics and a need for substitution of amalgam. The first resin composites were difficult to work with and dentists were not familiar with the placement and finishing techniques. Anecdotal reports indicated that resin composite restorations failed much sooner than amalgams . Many dentists were discouraged to place posterior restorations and during the first decades after introduction on the market, resin composites were mostly used in anterior teeth . Due to a growing concern about the biocompatibility of amalgam, the use of posterior resin composite restorations increased during the nineties . The earliest two-paste chemical-cured materials were successively replaced by light-cured resin based materials, which could be placed in increments. Today, advantages like aesthetics and tooth substance saving techniques make resin composites the predominant material in all placed and replaced cavity types in many countries .

The first posterior resin composites exhibited clinical problems like rapid occlusal and proximal wear, rough surface characteristics, marginal staining and bulk discoloration . Inferior bonding and resin composite shrinkage resulted in lack of adaptation to the cavity walls, and as a consequence, an increased susceptibility to caries and high frequency of postoperative symptoms was expected . Secondary caries has been the main reason for replacement of restorative materials as reported in cross sectional and prospective studies .

Dental restorations do have a limited lifespan and replacement of a failed restoration leads to increase of cavity size and destruction of tooth tissues . Increasing the longevity of restorations is therefore an important aim in dentistry. Longevity of restorations is influenced by multiple factors, i.e. the operator, materials selected, and also the patients with their differences in oral environment, including location and size of the restoration, caries risk and bruxing habits, play a major role . To estimate the longevity of dental restorations, long-term clinical evaluations are needed . Compared to amalgam surprisingly little long-term evidence exists for resin composites . Resin composites seem to be the most common replacement material for posterior amalgam restorations and the dental profession therefore needs long time information on these materials to guaranty their effectiveness .

The aim of this randomized controlled prospective study was to evaluate the long term clinical behavior of three conventional resin composites, two chemical-cured and one light-cured, in Class II cavities after 30 years. The hypothesis to be tested was that the light-cured resin composite would perform better than the chemical-cured materials.

Materials and methods

Experimental design

During March–June 1984, adult patients attending the Dental School in Copenhagen, who were in need of three or six similar sized Class II restorations and requested tooth-colored restorations, were invited to participate in the follow-up. No participant was excluded because of high caries activity, periodontal condition or parafunctional habits. Excluded were participants with known allergic symptoms for dental resins, patients younger than 18 years as well as patients with partial prosthesis or orthodontic apparatus. Reasons for placement were replacements of amalgam fillings for aesthetic reasons, concern of amalgam (60%) and secondary caries or primary caries (40%). All teeth were in occlusion and had at least one synergist. In order to make an intra-individual comparison possible of the three tested resin materials (P10, P30 (3M Dental Products, St Paul, Minn., USA) and Miradapt (Johnson & Johnson Ltd, New Brunswich, NY, USA) ( Table 1 ) each patient received three (one set) or six (two sets) as similar sized and located restorations as possible. After cavity preparation of the teeth in each set, the cavities were randomly distributed to be restored with the different materials according to a predetermined scheme of randomization. The participants were not aware of in which cavity the experimental materials were placed. Vitality of the tooth was tested electronically (Dental Electronic, Denmark, interval 0–32).

Table 1
Resin composites and bonding system used.
Material Composition Type Application steps Manufacturer
P10 Matrix:
Filler: quarts, Sb–Al–Mg–Fe-glass
SiO 2 (pyrogenic) (van Dijken et al., 1989)
Chemical-cured two paste hybrid resin composite Bulk applikation 3M Dental Products, St Paul, Minn., USA
P30 Filler: ZrO 2 , Zn—B-Ti-glass, 1–3 μm
SiO 2 (pyrogenic) (van Dijken et al. 1989)
Light-cured one paste hybrid resin composite 2 mm layering technique 3M Dental Products, St Paul, Minn., USA
Miradapt Filler: bariun glass 12 μm, SiO 2 (pyrogenic) 0.04 μm Chemical-cured two paste hybrid resin composite Bulk application Johnson & Johnson Ltd, New Brunswich, NY, USA
Scotchbond Halophosphorous esters of Bis-GMA, alcoholic sulphanate solution 2-component chemical- cured enamel bonding Etch (35% H 3 PO 4 ) of enamel 30 s. Water rinse 20 s. Air drying. Twice application 99% ethanol. Air blowing. Application bonding agent 2 times for 10 s. Air thinning bonding layer. 3M Dental Products, St Paul, Minn., USA
Concise enamel bond TEGDMA (45–55%), Bis-GMA (40–50%), dibenzoylperoxid <2% Unfilled chemical-cured enamel bonding agent After polishing of restoration etch with 37% phosphoric acid of restoration margins. Water rinse. Air drying. Application of 99% ethanol. Air blowing. Application bonding agent on all cavity margins 3M Dental Products, St Paul, Minn., USA

Patients were informed on the study and special information was given of the use of resin materials in posterior teeth, which was at the start of the study not recommended or indicated by the manufacturers. All patients were promised to receive free gold-inlays, if durability did not comply with comparable amalgam restorations. All restorations were made by one clinician (UP). At the start of the study neither signed consent nor approval by the ethics committee in Denmark was necessary. In 1993 accept from the Danish Data Protection Agency was obtained (J. No. 1993-1110-1449). The study was performed in accordance with the Helsinki Declaration of 1975.

Thirty-three sets of 3 Class II restorations were placed in 30 participants (9 male with 10 sets, 21 females with 23 sets) with a mean age of 29.6 years (20–43). Three of the participants received 2 sets. The restorations were placed: 76 in premolars (maxilla 54, mandible 22), and 23 in molars (maxilla 14, mandible 9) ( Table 2 ). The sample size was based on the basis of previous similar designed clinical studies and recommendations and later ADA Acceptance Guidelines of at least 30 restorations per material to be included in the original study design (American Dental Association (ADA) Acceptance Program Guidelines for Restorative Materials, American Dental Association Council on Scientific Affairs, March 1996). The low drop out frequency during the 30 years resulted in that the number per resin composite group was still close to the sample size recommended by the ADA. Despite that the CONSORT statement, an evidence-based minimum set of recommendations for reporting randomized trials, was published considerably after the start of the study, most of the statement’s checklist were followed.

Table 2
Distribution of the experimental restorations.
Surfaces Mandible Maxilla Total
Premolars Molars Premolars Molars
2 Surfaces 18 6 28 11 63
3 Surfaces 4 3 26 3 36
Total 22 9 54 14 99

Clinical procedures

Preparations were performed with high speed air-rotor and low speed micro-motor with rounded cylindrical and round burs. No bevels were prepared and the outline of the cavity was determined by the caries lesion or the size of the restoration to be replaced. Caries excavation was performed by round burs at low speed or manually with hand instruments. After cleaning of the cavity by water rinsing, lining of exposed dentin was performed with a chemical-cured calcium hydroxide cement (Dycal, DeTrey Dentsply, Konstanz, Germany). Isolation of the operative field was performed with cotton rolls and suction device. Transparent celluloid matrix bands (Contact-Molar-band R , Hawe Neos, Switzerland) in Nyström matrix retainers and wooden wedges (Hawe Neos, Switzerland) were placed. Etching of the enamel margins was done with a 37% phosphoric acid gel for 30 s, followed by rinsing with water for 20 s and drying with air . Ethanol (99%) was applied twice with a cotton pellet for 10 s and removed with air resulting in an opaque dry enamel surface (dry bonding technique). The chemical-cured enamel bonding Scotchbond™ Dental Adhesive System (3M) was mixed and applied in two thin layers during 5–10 s and then thinned with air. The light-cured resin composite (P30) was applied in maximally 2 mm thick oblique layers which were light-cured from the occlusal surface for 40 s per layer (Elipar curing light, 3M) and for 40 s gingivally after removal of wedges and matrix retainer (from both facial and palatinal side). For the two chemical-cured materials (P10 and Miradapt) equal amounts of catalyst and base material were mixed for 30 s. Universal colors were used for all materials. The cavities with the chemical-cured resin composites were filled in bulk with excess using a syringe (Hawe Neos). The occlusal surfaces were covered by a celluloid matrix band, which was pressed against the surface during curing for 4 min. After at least 10 min, finishing and polishing was done under water cooling with finishing diamond burs, silicone polisher, polishing discs and strips. Then control of marginal adaptation, anatomic form, occlusion and articulation contacts and proximal contacts was performed by probing, articulation foil and dental floss. In order to seal cavity margins directly after polishing, etching of all margins of the restorations was performed with 37% phosphoric acid during 10 s, followed by water rinse, intensive air drying and application of 99% ethanol. A hydrophobic chemical cured bonding agent (Concise Enamel Bond, 3M) was then applied with a cotton pellet in a thin layer. Excess material was removed by a cotton pellet, dental floss and dental probe. The resin was allowed to cure for at least two minutes in an isolated operative field. Excess of cured bonding agent was then removed by a scaler and occlusion was checked again.

Evaluation

At baseline (after placement of the restorations) and after 2, 3, 5, 10, 20 and 30 years the restorations were assessed clinically and radiographically. During the study period dental care of the patients was performed by the patient’s private dentists. Slightly modified USPHS criteria were used for the clinical evaluation containing the following parameters: anatomic form, marginal adaptation, marginal discoloration, surface roughness, color match and secondary caries Table 3 ) . Clinical registrations were performed blindly by the operator and at regular intervals by two calibrated evaluators. Post-operative sensitivity was recorded by telephone interview 1–2 weeks after the placement of the restorations or when the patient addressed the clinic with pain. At all recalls the participants were questioned again about persisting or new sensitivity symptoms. Bite-wing radiographs were taken at all the study recalls by the evaluating team and in between these by their own dentists. The experimental resin composites exhibited low radiopacity on the recall radiographs compared to newer resin composites. Change of the experimental resin composites by later marketed resin composites by other dentists could therefore be easily detected. When a replacement or repair had taken place, the private dentist was contacted after a signed consent from the patient, and reason for repair or replacement (symptoms, fracture of restoration, fracture of tooth, secondary caries, caries at new surface, or other reasons) date was collected from the file. Clinical photos were taken of most teeth for documentation.

Table 3
Criteria for direct clinical evaluation (van Dijken, Acta Odontol Scand 1986).
Category Score (acceptable/unacceptable) Criteria
Anatomical form 0 The restoration is continuous with tooth anatomy
1 Slightly under- or over-contoured restoration; marginal ridges slightly undercontoured; contact slightly open (may be self-correcting); occlusal height reduced locally
2 Restoration is undercontoured, dentin or base exposed; contact is faulty, not self-correcting; occlusal height reduced; occlusion affected
3 Restoration is missing, partially fractured or shows traumatic occlusion; restoration causes pain in tooth or adjacent tissue
Marginal adaptation 0 Restoration is continuous with existing anatomic form, explorer does not catch
1 Explorer catches, no crevice is visible into which explorer will penetrate
2 Crevice at margin, enamel exposed
3 Obvious crevice at margin, dentin or base exposed
4 Restoration mobile, fractured or missing
Color match 0 Very good color match
1 Good color match
2 Slight mismatch in color, shade or translucency
3 Obvious mismatch, outside the normal range
4 Gross mismatch
Marginal discoloration 0 No discoloration evident
1 Slight staining, can be polished away
2 Obvious staining can not be polished away
3 Gross staining
Surface roughness 0 Smooth surface
1 Slightly rough or pitted
2 Rough, cannot be refinished
3 Surface deeply pitted, irregular grooves
Caries 0 No evidence of caries contiguous with the margin of the restoration
1 Caries is evident contiguous with the margin of the restoration

The following baseline data were registered: name, gender, age of the participant, date of placement, indication for placement (patient request for aesthetic or non-metallic restoration, caries, fracture of tooth or restoration, reinforcement of tooth), tooth number, surfaces involved, and presence of enamel or dentin at the gingival margins of proximal cavities.

The caries risk for each participant and their parafunctional habits activity during the follow up were estimated at the end of the evaluation period by the treating clinician by means of clinical and socio-demographic information routinely available at the clinical examinations, e.g. incipient caries lesions, former caries history, frequency and symptoms related to bruxing activity .

Statistical analysis

The endpoint of each restoration, replacement or repair, was defined as a failed restoration. Replacements or repairs due to caries in a non-filled surface of a tooth with an acceptable RC were not considered as reasons for failures. The evaluated scores of the restorations are described by cumulative relative frequencies. The data were analyzed with the MedCalc statistical software package for biomedical science, version12.1.4.0 (MedCalc Software, www.medcalc.org ) and Stata/SE version 12.0 (StataCorp LP, www.stata.com ). Survival curves of the restorations and differences between subgroups (resin composite type, tooth type, cavity, age and gender of participants) were created with Kaplan–Meier statistics. The survival of the restorations grouped on the basis of the covariables was evaluated with log rank test. Univariate and multivariate Cox regression analyses were used with shared frailty to test the influence of the co-variables to test the influence of the co-variables resin composite, tooth type, cavity size, gender, and age on survival time. Hazard ratios with respective 95% confidence intervals were determined by Cox regression analyses. The null hypothesis was rejected at the 5% level. Associations of the covariates with the main reasons for failure were described by descriptive statistics.

Materials and methods

Experimental design

During March–June 1984, adult patients attending the Dental School in Copenhagen, who were in need of three or six similar sized Class II restorations and requested tooth-colored restorations, were invited to participate in the follow-up. No participant was excluded because of high caries activity, periodontal condition or parafunctional habits. Excluded were participants with known allergic symptoms for dental resins, patients younger than 18 years as well as patients with partial prosthesis or orthodontic apparatus. Reasons for placement were replacements of amalgam fillings for aesthetic reasons, concern of amalgam (60%) and secondary caries or primary caries (40%). All teeth were in occlusion and had at least one synergist. In order to make an intra-individual comparison possible of the three tested resin materials (P10, P30 (3M Dental Products, St Paul, Minn., USA) and Miradapt (Johnson & Johnson Ltd, New Brunswich, NY, USA) ( Table 1 ) each patient received three (one set) or six (two sets) as similar sized and located restorations as possible. After cavity preparation of the teeth in each set, the cavities were randomly distributed to be restored with the different materials according to a predetermined scheme of randomization. The participants were not aware of in which cavity the experimental materials were placed. Vitality of the tooth was tested electronically (Dental Electronic, Denmark, interval 0–32).

Table 1
Resin composites and bonding system used.
Material Composition Type Application steps Manufacturer
P10 Matrix:
Filler: quarts, Sb–Al–Mg–Fe-glass
SiO 2 (pyrogenic) (van Dijken et al., 1989)
Chemical-cured two paste hybrid resin composite Bulk applikation 3M Dental Products, St Paul, Minn., USA
P30 Filler: ZrO 2 , Zn—B-Ti-glass, 1–3 μm
SiO 2 (pyrogenic) (van Dijken et al. 1989)
Light-cured one paste hybrid resin composite 2 mm layering technique 3M Dental Products, St Paul, Minn., USA
Miradapt Filler: bariun glass 12 μm, SiO 2 (pyrogenic) 0.04 μm Chemical-cured two paste hybrid resin composite Bulk application Johnson & Johnson Ltd, New Brunswich, NY, USA
Scotchbond Halophosphorous esters of Bis-GMA, alcoholic sulphanate solution 2-component chemical- cured enamel bonding Etch (35% H 3 PO 4 ) of enamel 30 s. Water rinse 20 s. Air drying. Twice application 99% ethanol. Air blowing. Application bonding agent 2 times for 10 s. Air thinning bonding layer. 3M Dental Products, St Paul, Minn., USA
Concise enamel bond TEGDMA (45–55%), Bis-GMA (40–50%), dibenzoylperoxid <2% Unfilled chemical-cured enamel bonding agent After polishing of restoration etch with 37% phosphoric acid of restoration margins. Water rinse. Air drying. Application of 99% ethanol. Air blowing. Application bonding agent on all cavity margins 3M Dental Products, St Paul, Minn., USA

Patients were informed on the study and special information was given of the use of resin materials in posterior teeth, which was at the start of the study not recommended or indicated by the manufacturers. All patients were promised to receive free gold-inlays, if durability did not comply with comparable amalgam restorations. All restorations were made by one clinician (UP). At the start of the study neither signed consent nor approval by the ethics committee in Denmark was necessary. In 1993 accept from the Danish Data Protection Agency was obtained (J. No. 1993-1110-1449). The study was performed in accordance with the Helsinki Declaration of 1975.

Thirty-three sets of 3 Class II restorations were placed in 30 participants (9 male with 10 sets, 21 females with 23 sets) with a mean age of 29.6 years (20–43). Three of the participants received 2 sets. The restorations were placed: 76 in premolars (maxilla 54, mandible 22), and 23 in molars (maxilla 14, mandible 9) ( Table 2 ). The sample size was based on the basis of previous similar designed clinical studies and recommendations and later ADA Acceptance Guidelines of at least 30 restorations per material to be included in the original study design (American Dental Association (ADA) Acceptance Program Guidelines for Restorative Materials, American Dental Association Council on Scientific Affairs, March 1996). The low drop out frequency during the 30 years resulted in that the number per resin composite group was still close to the sample size recommended by the ADA. Despite that the CONSORT statement, an evidence-based minimum set of recommendations for reporting randomized trials, was published considerably after the start of the study, most of the statement’s checklist were followed.

Table 2
Distribution of the experimental restorations.
Surfaces Mandible Maxilla Total
Premolars Molars Premolars Molars
2 Surfaces 18 6 28 11 63
3 Surfaces 4 3 26 3 36
Total 22 9 54 14 99

Clinical procedures

Preparations were performed with high speed air-rotor and low speed micro-motor with rounded cylindrical and round burs. No bevels were prepared and the outline of the cavity was determined by the caries lesion or the size of the restoration to be replaced. Caries excavation was performed by round burs at low speed or manually with hand instruments. After cleaning of the cavity by water rinsing, lining of exposed dentin was performed with a chemical-cured calcium hydroxide cement (Dycal, DeTrey Dentsply, Konstanz, Germany). Isolation of the operative field was performed with cotton rolls and suction device. Transparent celluloid matrix bands (Contact-Molar-band R , Hawe Neos, Switzerland) in Nyström matrix retainers and wooden wedges (Hawe Neos, Switzerland) were placed. Etching of the enamel margins was done with a 37% phosphoric acid gel for 30 s, followed by rinsing with water for 20 s and drying with air . Ethanol (99%) was applied twice with a cotton pellet for 10 s and removed with air resulting in an opaque dry enamel surface (dry bonding technique). The chemical-cured enamel bonding Scotchbond™ Dental Adhesive System (3M) was mixed and applied in two thin layers during 5–10 s and then thinned with air. The light-cured resin composite (P30) was applied in maximally 2 mm thick oblique layers which were light-cured from the occlusal surface for 40 s per layer (Elipar curing light, 3M) and for 40 s gingivally after removal of wedges and matrix retainer (from both facial and palatinal side). For the two chemical-cured materials (P10 and Miradapt) equal amounts of catalyst and base material were mixed for 30 s. Universal colors were used for all materials. The cavities with the chemical-cured resin composites were filled in bulk with excess using a syringe (Hawe Neos). The occlusal surfaces were covered by a celluloid matrix band, which was pressed against the surface during curing for 4 min. After at least 10 min, finishing and polishing was done under water cooling with finishing diamond burs, silicone polisher, polishing discs and strips. Then control of marginal adaptation, anatomic form, occlusion and articulation contacts and proximal contacts was performed by probing, articulation foil and dental floss. In order to seal cavity margins directly after polishing, etching of all margins of the restorations was performed with 37% phosphoric acid during 10 s, followed by water rinse, intensive air drying and application of 99% ethanol. A hydrophobic chemical cured bonding agent (Concise Enamel Bond, 3M) was then applied with a cotton pellet in a thin layer. Excess material was removed by a cotton pellet, dental floss and dental probe. The resin was allowed to cure for at least two minutes in an isolated operative field. Excess of cured bonding agent was then removed by a scaler and occlusion was checked again.

Evaluation

At baseline (after placement of the restorations) and after 2, 3, 5, 10, 20 and 30 years the restorations were assessed clinically and radiographically. During the study period dental care of the patients was performed by the patient’s private dentists. Slightly modified USPHS criteria were used for the clinical evaluation containing the following parameters: anatomic form, marginal adaptation, marginal discoloration, surface roughness, color match and secondary caries Table 3 ) . Clinical registrations were performed blindly by the operator and at regular intervals by two calibrated evaluators. Post-operative sensitivity was recorded by telephone interview 1–2 weeks after the placement of the restorations or when the patient addressed the clinic with pain. At all recalls the participants were questioned again about persisting or new sensitivity symptoms. Bite-wing radiographs were taken at all the study recalls by the evaluating team and in between these by their own dentists. The experimental resin composites exhibited low radiopacity on the recall radiographs compared to newer resin composites. Change of the experimental resin composites by later marketed resin composites by other dentists could therefore be easily detected. When a replacement or repair had taken place, the private dentist was contacted after a signed consent from the patient, and reason for repair or replacement (symptoms, fracture of restoration, fracture of tooth, secondary caries, caries at new surface, or other reasons) date was collected from the file. Clinical photos were taken of most teeth for documentation.

Table 3
Criteria for direct clinical evaluation (van Dijken, Acta Odontol Scand 1986).
Category Score (acceptable/unacceptable) Criteria
Anatomical form 0 The restoration is continuous with tooth anatomy
1 Slightly under- or over-contoured restoration; marginal ridges slightly undercontoured; contact slightly open (may be self-correcting); occlusal height reduced locally
2 Restoration is undercontoured, dentin or base exposed; contact is faulty, not self-correcting; occlusal height reduced; occlusion affected
3 Restoration is missing, partially fractured or shows traumatic occlusion; restoration causes pain in tooth or adjacent tissue
Marginal adaptation 0 Restoration is continuous with existing anatomic form, explorer does not catch
1 Explorer catches, no crevice is visible into which explorer will penetrate
2 Crevice at margin, enamel exposed
3 Obvious crevice at margin, dentin or base exposed
4 Restoration mobile, fractured or missing
Color match 0 Very good color match
1 Good color match
2 Slight mismatch in color, shade or translucency
3 Obvious mismatch, outside the normal range
4 Gross mismatch
Marginal discoloration 0 No discoloration evident
1 Slight staining, can be polished away
2 Obvious staining can not be polished away
3 Gross staining
Surface roughness 0 Smooth surface
1 Slightly rough or pitted
2 Rough, cannot be refinished
3 Surface deeply pitted, irregular grooves
Caries 0 No evidence of caries contiguous with the margin of the restoration
1 Caries is evident contiguous with the margin of the restoration
Only gold members can continue reading. Log In or Register to continue

Stay updated, free dental videos. Join our Telegram channel

Nov 23, 2017 | Posted by in Dental Materials | Comments Off on A randomized controlled 30 years follow up of three conventional resin composites in Class II restorations

VIDEdental - Online dental courses

Get VIDEdental app for watching clinical videos