Journal of Dentistry (2006) 34, 292–297
www.intl.elsevierhealth.com/journals/jden
A randomized clinical trial on proximal contacts of posterior composites B.A.C. Loomansa,*, N.J.M. Opdama, F.J.M. Roetersa, E.M. Bronkhorsta, ¨rferb R.C.W. Burgersdijka, C.E. Do a
Department of Preventive and Restorative Dentistry, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands b Clinic for Conservative Dentistry, School of Medicine, University of Heidelberg, Germany Received 21 February 2005; received in revised form 12 July 2005; accepted 18 July 2005
KEYWORDS Randomized clinical trial; Class II composite resin restorations; Tooth pressure meter (TPM); Matrix systems; Separation rings
Abstract Objective: The objective of this study was to investigate clinical changes in proximal contact strength inserting Class II composite resin restorations according to one of three randomly assigned protocols. Materials and methods: Seventy-one Class II restorations (MO/DO) were placed by two calibrated operators. Restorations were randomly assigned to one of three groups: one using a circumferential and two a sectional matrix system with separation rings. Proximal contacts were measured by one independent observer with a Tooth Pressure Meter immediately before treatment, and directly after finishing the restoration. Results: Compared to the situation before treatment groups with a sectional matrix system resulted in a statistical significant stronger mean proximal contact strengths (p!0.05), whereas the use of a circumferential matrix system with hand-instrument resulted in a lower proximal contact strength (p!0.05). Conclusion: Class II posterior composite resin restorations placed with a combination of sectional matrices and separation rings resulted in a stronger proximal contact than when a circumferential matrix system was used. Q 2005 Elsevier Ltd. All rights reserved.
Introduction To obtain, a tight proximal contact with Class II composite resin restorations the clinical procedure has to compensate for the thickness of the matrix
* Corresponding author. Tel.: C31 24 3616410; fax: C31 243540265. E-mail address:
[email protected] (B.A.C. Loomans).
as well as the polymerization shrinkage of the composite resin. With amalgam, a tight proximal contact can be obtained by condensation of the material. The assumption that applying pressure on the composite resin material would have the same effect was one of the main reasons to develop highviscous composites or ceromers. However, in vitro studies show that packing of high-viscous composites does not contribute to a tight proximal contact, due to the visco-elastic behavior of
0300-5712/$ - see front matter Q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jdent.2005.07.008
Clinical trial on proximal contacts of composites composite resin materials.1,2 Another technique which is recommended to achieve tight proximal contacts is the ‘pre-wedging’ or ‘multiple wedging’ technique. A wooden wedge is pressed firmly in the interdental space before cavity preparation and kept in place during preparation.3–5 Another study shows that a special separation ring combined with a sectional matrix resulted in tight proximal contacts in vitro.2 However, no clinical research is available to prove the effectiveness of both these techniques in the amount of obtained interdental separation. Moreover, clinical studies are lacking on the reconstruction of the proximal contact with composite resin restorations. In clinical studies, the tightness of the proximal contact can be measured by passing dental floss interdentally and scoring the proximal contact strength as ‘satisfactory’, ‘acceptable’ or ‘unacceptable’.6 Unfortunately, this method it is not applicable to record minor changes in proximal contact strength. A more accurate device to measure clinically, the proximal contact strength has been used in a clinical trial and is able to record minor changes.7 The objective of this randomized clinical trial was to compare a circumferential matrix system with a hand-instrument and two sectional matrix systems combined with separation rings in the reconstruction of the proximal contact of a Class II composite resin restoration.
Materials and methods This randomized clinical trial was performed between October 2002 and December 2003. The study design was approved by the Central Committee on Research Involving Human Subjects (CMO-nr: 2001/056). Patients who required a Class II composite resin restoration (MO or DO) were asked to participate in the study in two Dutch general dental practices. Inclusion criteria were: good general health, minimum age of 18, and fully erupted occluding teeth. Exclusion criteria were diastema between posterior teeth, presence of fixed partial dentures, severe periodontal diseases and tooth mobility more than mobility-score 1. A poweranalysis was performed with Alpha at 5%, Beta at 90% in a three-group design. In this calculation the standard deviation within the groups (matrix systems) was estimated to be 3, and the standard deviation between the three groups was assumed to be 1.4. This resulted in an estimated group size of 30 restorations. Informed consent was obtained and all 63 patients (19 males/44 females, age 18–66 years, mean age of 33.9) participated in the study. Seventy-one Class II composite resin restorations
293 (MO/DO) were placed by two calibrated operators (Operator 1: 42 restorations in 39 patients, Operator 2: 29 restorations in 24 patients). In six patients multiple restorations were placed (in four patients two restorations and in two patients three restorations). Each restoration was placed in a separate session, and therefore, nesting of multiple restorations within patients was ignored in the analysis. All restorations were placed according to the following protocol: local anesthesia (Ultracaı¨ne DS, Aventis) was administered if necessary. A wooden wedge (Hawe Neos, Bioggio, Switzerland) was placed interproximally at the surface to be restored. The wedge was pushed tightly in the interdental space from the palatal side. Wedges were kept in place tightly to obtain separation of teeth and to prevent damage of the papilla during the whole preparation procedure, also called ‘prewedging’.3 To open the cavity or to remove an existing restoration a high-speed hand-piece (KaVo Intramatic LUX 2, 25 LN, Biberach, Germany) with a diamond bur (FG 108-009, Horico) was used. Carious tissue was excavated with low speed hand-piece (KaVo Intramatic LUX 2, 29 LN) with round steel burs (RA 001-012/014, Meiss) and hand-excavators. The outline was finished in a bevel with either finegrit diamond bur (FG 249-F012, Horico) or a Sonicsys preparation device (KaVo Dental, Biberach, Germany) to prevent damage of adjacent teeth.8 Restorations in adjacent teeth were adjusted and polished with fine-grit diamond burs and Sof-LexTM discs (3M Espe, St Paul, MN, USA). After finishing the preparation teeth were randomly assigned to one of three experimental groups with a special balancing computer program (Trial Balance Program, NRC Nestle, Vevey, Switzerland). This program distributes the samples to one of the experimental groups, based on existing clinical variations other than the experimental variables. Balancing factors were: bucco-lingual width of the preparation (%3 mm or O3 mm), depth of the box (%3 mm or O3 mm.), age of patient (18–29, 30–49 or 50–69), location of tooth (4, 5 or 6), treated surface (MO-DO) and jaw (upper-lower). For moisture control cotton rolls were combined with a tongue shield and suction. Group 1 A circumferential pre-contoured matrix system (Tofflemire, Produits Dentaire SA, Vevey, Switzerland and matrix Hawe Neos 1001c, 0.05 mm thick, Bioggio, Switzerland) was placed around the tooth and fixed with an interdental wooden wedge (Hawe Neos, Bioggio, Switzerland). The contact area was burnished with a handinstrument so that no visual space was left
294 Table 1
B.A.C. Loomans et al. Specification of experimental groups. N (rest.)
Pre-wedging
Matrix systems
Separation ring
Additional force with handinstrument (PFI 49)
Group 1
28
Yes
–
Yes
Group 2
20
Yes
23
Yes
Bi-Tine type I ring Contact Matrix ring
–
Group 3
TofflemireCHawe Neos 1001C matrix Palodent, Standard matrix (dead-soft) Danville, Lite Flex (flexible)
between matrix and adjacent tooth. During polymerization of the first horizontal layer of the composite resin a handinstrument (PFI 49, Weybridge, UK) was placed on the contact area to press the matrix against the adjacent tooth surface. Group 2 A sectional contoured dead-soft Standard Matrix (0.04 mm thick, Palodent, Dentsply Caulk, Milford, DE) was combined with a separation ring (Bi-Tine ring type-I, Palodent, Dentsply Caulk, Milford, DE). The sectional matrix was fixated with a wooden wedge (Hawe Neos, Bioggio, Switzerland) and subsequently the separation ring was placed. The matrix was burnished with a hand-instrument (PFI 49) against the proximal surface of the adjacent tooth. Group 3 For this group, the same procedure was followed as in group 2, but instead of the Palodent matrix system, the sectional Contact Matrix System combining a contoured flexible Thin Flex Matrix (0.04 mm, Danville Materials, San Ramon, California) and a separation ring (Contact Rings, inward/outward, Danville Materials, San Ramon, California) were applied. Table 1 summarizes the specifications of the experimental groups used in this study. The cavities were restored using a three-step adhesive system (Clearfil SA Primer, Clearfil Photo Bond (nZ44)) or a two-step self-etching system (Clearfil SE Bond (nZ27)) and a highly filled hybrid composite (Clearfil AP-X (A3), Kuraray Co., Osaka, Japan). The composite resin was injected from a preloaded tip in layers of approximately !2 mm, which were cured for 20 s from the occlusal surface using a polymerization unit (Optilux 401, Demetron, Kerr, USA, light intensity 800 mW/cm2). After removal of the matrix, restorations were postcured for 20 s from both buccal and lingual side. Restorations were finished using fine grit diamond burs (FG 249-F012 and FG 277-F023, Horico) and
–
Sof-LexTM discs and the occlusion was checked with articulating paper (Bausch Articulating Paper, 40 mm). Proximal contact strengths were measured using the Tooth Pressure Meter (TPM) that was constructed at the University of Technology Delft in The Netherlands according to the principles as described by Do ¨ rfer et al.7 Proximal contact strengths were recorded by one independent observer before treatment (T0) and directly after finishing the restoration (T1) at the contact of the treatment site (CT), at the contact mesial of the treatment site (CM) and at the contact distal of the treatment site (CD) as can be seen in Fig. 1. As a control the proximal contact strengths were recorded at the same locations in the contra-lateral quadrant. A 0.05 mm thick metal strip is connected to this device, which is inserted interdentally from the occlusal surface. Proximal contact strength (N)
Figure 1 Location of the measuring sites (CM, CT and CD) in case when restoring a DO preparation in tooth 14.
Clinical trial on proximal contacts of composites is quantified as the maximum frictional force when the strip is slowly removed in occlusal direction. Custom-written software in Excel (MS Office 2000, Windows) is used for data acquisition and for construction of diagrams, relating force (N) to seconds (s). At each measuring site three measurements were taken, of which outcomes should be within a maximum range of 0.5 N. Measurement outcomes exceeding this range or presence of visible irregularities in the diagram, e.g. due to deformations of the strip, excluded data for analysis and leads to repetition of the measurement. The final result of a single measuring site is the mean value of these three outcomes. Due to the fact that the patient is treated in a horizontal position, possible changes of proximal contact strength occur between both measurements before and after treatment.8,9 Therefore, the contact strength at the experimental side after treatment (T1) was corrected with the mean difference (T1KT0) in proximal contact strength at the control side. Data were statistically analyzed using SPSS 12. To find differences between groups ANOVA and Tukey tests were used at p!0.05 and a linear regression analysis was used to determine the influence of the following variables: matrix systems, jaw (right-left and upper-lower), box width and depth, tooth surfaces (MO-DO), age, gender (male–female), adhesive system and the interactions between operator and matrix system.
Results A large inter-individual variation in proximal contact strengths before treatment was found to
Table 2 Groups
295 exist (0.10 N!T0!12.43 N). The mean proximal contact strength before treatment at the contact of the treatment site (CT) was 3.19 N (95% CI [2.49– 3.89 N]), at the contact site mesial of the treatment site (CM) the mean proximal contact strength was 2.42 N (95% CI [1.97–2.87 N]) and at the contact site distal of the treatment site (CD) it was 2.65 N (95% CI [2.17–3.13 N]). No statistical significant differences of proximal contact strengths before treatment were found between the three experimental groups at treatment side (between groups 1 and 2: pZ0.351; between groups 1 and 3: pZ0.714; between groups 2 and 3: pZ0.814). Differences in proximal contact strengths before and after treatment (DT10ZT1KT0) for all three restorative techniques at three measuring sites are shown in Table 2. All data were normally distributed so that parametric statistical analyses were suitable. Before statistical analyses were performed data were corrected with the mean difference (T1KT0) in proximal contact strength at the control side. At the treatment site (CT) a decrease of proximal contact strength was found for the circumferential matrix (group 1) and an increase of proximal contact strength was found for both sectional matrix systems combined with separation ring (groups 2 and 3). The differences between the circumferential system (group 1) and the two sectional matrix systems (groups 2 and 3) were statistical significant (p!0.001 resp. pZ0.03). No statistical significant difference was found between groups 2 and 3 (pZ0.247). At the contacts mesial and distal of the treatment site (CM and CD) both sectional matrix systems resulted in an increase of proximal contact strength but no statistical significant differences were found between groups 2 and 3
Differences in proximal contact strength divided over site and matrix system. N
Contact of treatment site (CT) 1. Tofflemire 28 2. Palodent 20 3. Contact matrix 23 Contact mesial of treatment site (CM) 1. Tofflemire 28 2. Palodent 20 3. Contact matrix 23 Contact distal of treatment site (CD) 1. Tofflemire 28 2. Palodent 20 3. Contact matrix 23
DT10 (N)
SEM (N)
95% CI (N)
K2.11a 2.56b 0.96b
0.61 0.59 0.77
[K3.33. K0.88] [1.39. 3.73] [K0.57. 2.50]
K0.45c 0.61d 1.41d
0.23 0.22 0.31
[K0.92. 0.02] [0.18. 1.05] [0.79. 2.04]
K0.18e 0.61f 0.33ef
0.16 0.16 0.16
[K0.50. 0.15] [0.28. 0.93] [0.01. 0.66]
Statistical significant differences between experimental groups at each site have been marked with a different characters (a–c, etc.) (p!0.05).
296 Table 3
B.A.C. Loomans et al. Linear regression analysis to determine the effect of different variables.
Model
Unstandardized coefficients (b)
Standardized coefficients (b)
Sig. (pZvalue)
95% CI
(Constant) Danville Palodent Tooth (4–6) Tooth surfaces (1ZMO, 2ZDO) Jaw (1Zright, 2Zleft) Jaw (1Zupper, 2Zlower) Box width (1Z%3 mm, 2ZO3mm) Box depth (1Z%3 mm, 2ZO3 mm) Adhesive system (1ZPhotoBondC SA Primer, 2ZSE Bond) Age Gender (1Zmale, 2Zfemale) Interaction between Operator 1 and Danville Interaction between Operator 1 and Palodent Interaction between Operator 1 and Tofflemire
1.52 3.30 6.84 0.15 0.55 K0.83 0.16 0.96 K1.44 K1.69
0.42 0.82 0.03 0.07 K0.11 0.02 0.12 K0.18 K0.22
0.841 0.033 0.001 0.891 0.748 0.413 0.876 0.457 0.247 0.073
[K13.70. 16.75] [0.27. 6.34] [3.01. 10.66] [K2.08. 2.38] [K2.88. 3.98] [K2.86. 1.20] [K1.86. 2.18] [K1.62. 3.54] [K3.91. 1.03] [K3.54. 0.17]
K0.05 K0.26 1.80
K0.13 K0.03 0.18
0.359 0.814 0.319
[K0.16. 0.06] [K2.42. 1.91] [K1.79. 5.39]
K1.49
K0.16
0.471
[K5.63. 2.64]
2.40
0.26
0.130
[K0.74. 5.55]
(CM: pZ0.103 resp. CD: pZ0.496). At the contact mesial of the treatment site (CM) use of the circumferential system with hand-instrument (group 1) resulted in statistical significant weaker proximal contact strength compared to groups 2 and 3 (p!0.001 resp. pZ0.015). At the contact distal of the treatment site (CD) only a statistical significant difference was found between group 1 and 2 (pZ0.003), whereas no statistical significant difference was found between groups 1 and 3 (pZ 0.05). The greatest changes of contact strengths after treatment were observed at the treatment site. Moreover, stronger (or weaker) proximal contact strengths at the treatment site were associated with stronger (or weaker) contacts mesially and distally of the treatment site. From the linear regression analysis it appears that variables such as location in the jaw, tooth surfaces, box dimensions, age, gender and adhesive system did not have a statistical significant effect on the measurements. Neither there were statistical significant interactions between operator and used matrix systems (Table 3).
Discussion In this randomized clinical trial, the reconstruction of proximal contacts of Class II composite resin restorations placed with a traditional technique (circumferential matrix system, strong wedging and applying pressure by hand-instrument) were
compared with two experimental techniques using a sectional matrix and separation ring. The experimental techniques lead to significantly stronger proximal contacts compared to the applied traditional method. Both operators were experienced dentists and well-trained in handling the circumferential matrix system. Regarding the sectional system with separation ring only one operator was well-trained in this technique, while the other operator was relatively inexperienced. Nevertheless, the results showed that no statistically significant differences were found for both operators with various systems. Generally, it is assumed that an optimal contour and proximal contact strength are important characteristics to prevent food impaction, tooth migration, periodontal complications and caries.10–13 As contoured matrix bands contribute to a better contour of the proximal restoration it was decided to use pre-contoured sectional as well as precontoured circumferential matrix bands, but it is unknown whether the use of straight conventional matrix bands would have changed the outcome of the present study.14 When a restoration was present in the adjacent tooth-surface, it was adjusted and to remove marginal overhang or irregularities on the surface of the restoration. This might have some effect on the measurement of the proximal contact strength, but is inevitable as adjustment of the adjacent restoration is mandatory for ‘good clinical practice’.
Clinical trial on proximal contacts of composites In a study of Southard et al.,9 it has been shown that when a patient was positioned horizontally, the proximal contact strength decreased over time. Therefore, in the present study the test readings were corrected based on the difference between baseline and final recording on the untreated control side of each patient in order to exclude possible influences. If no differences were recorded at the control side, no correction of the original test data was executed. The better results of the groups 2 and 3 may be contributed either to the separation rings or the sectional matrix. As in another study of Loomans et al.,15 a statistical significant stronger separation effect of rings was found compared to the use of hand-instruments, it is likely that they play a crucial role in obtaining a tight proximal contact. According to this study, a large variation in proximal contact strength existed between individuals. From the three experimental groups, proximal contacts in the group with a circumferential matrix could not be restored into the original contact strength, whereas in the other groups, using the separation ring and sectional matrix an increase of contact strengths were recorded. From a clinical point of view the use of sectional matrix systems combined with separation rings may result in a more reliable proximal contact when placing posterior composite resin restorations. On the other hand a ‘correction by nature’ can be expected as stronger proximal contacts diminish in strength and weaker contacts may become stronger again over a period of time. However, it is unknown whether this adaptation in contact strength occurs, and therefore, a 6-months follow-up study on the same subjects is now in progress to establish this adaptability.
Conclusion Class II posterior composite resin restorations placed with a combination of sectional matrices and separation rings resulted in a stronger proximal contact than when a circumferential matrix system was used.
297
Acknowledgements The authors would like to thank general practioner A.J.M. Loomans and Dental Laboratory Claeys for their contribution to this research.
References 1. Klein F, Keller AK, Staehle HJ, Do ¨rfer CE. Proximal contact formation with different restorative materials and techniques. Am J Dent 2002;15:232–5. 2. Peumans M, Van Meerbeek B, Asscherickx K, Simon S, Abe Y, Lambrechts P, et al. Do condensable composites help to achieve better proximal contacts? Dent Mater 2001;17: 533–41. 3. Albers HF. Textbook: tooth colored restoratives A text for selection, placement and finishing. 7th ed 1985. 4. Eli I, Weiss E, Kozlovsky A, Levi N. Wedges in restorative dentistry: principles and applications. J Oral Rehabil 1991; 18:257–64. 5. Wang J-C, Charbeneau GT, Gregory WA, Dennison JB. Quantitative evaluation of approximal contacts in class 2 composite resin restorations: a clinical study. Oper Dent 1989;14:193–202. 6. Ryge G. Clinical criteria. Int Dent J 1980;30:347–58. 7. Do ¨rfer CE, von Bethlenfalvy ER, Staehle HJ, Pioch T. Factors influencing proximal dental contact strengths. Eur J Oral Sci 2000;108:368–77. 8. Opdam NJM, Roeters FJM, Berghem van E, Eijsvogels E, Bronkhorst EM. Microleakage and damage to adjacent teeth when finishing class II adhesive preparations using either a sonic device or bur. Am J Dent 2002;15:317–20. 9. Southard TE, Southard KA, Tolley EA. Variation of approximal tooth contact tightness with postural change. J Dent Res 1990;69:1776–9. 10. Abrahams H, Kopczyk RA. Gingival sequela from retained piece of dental floss. J Am Dent Assoc 1983;106:57–8. 11. Ash MM. Wheeler’s dental anatomy physiology and occlusion. 7th ed. Philadelphia: W.B. Saunders; 1993. ¨rztliche 12. Do ¨rfer CE. Der Approximalraum. Deutsche Zahna Zeitschrift 1997;52:151–67. 13. Hancock EB, Mayo CV, Schwab RR, Wirthlin MR. Influence of interdental contacts on periodontal status. J Periodontol 1980;51:445–449.1. 14. Shennib HAF, Wilson NHF. An investigation of the adequacy of interproximal matrices commonly used with posterior composite restorations. J Dent 1986;14:84–6. 15. Loomans BAC, Opdam NJM, Roeters FJM, Bronkhorst EM, Burgersdijk RCW. Comparison of proximal contacts of class II composite resin restorations in vitro. Oper Dent; 2005, Submitted for publication.
