| Sign In to gain access to subscriptions and/or personal tools. |
© 2003 SAGE Publications Informatics Systems to Assess and Apply Clinical Research on Dental Restorative MaterialsPresented at "Dental Informatics & Dental Research: Making the Connection", a conference held in, Bethesda, MD, USA, June 12–13, 2003, sponsored by the University of Pittsburgh Center for Dental Informatics and supported in part by award 1R13DE014611-01 from the National Institute of Dental and Craniofacial Research/National Library of Medicine.
Department of Dental Biomaterials, College of Dentistry, University of Florida, Gainesville, FL 32610-0446; kanusavice{at}dental.ufl.edu
Dental biomaterials are used clinically for one or more of the following purposes: to restore function, to enhance esthetics, and to prevent or arrest demineralization of tooth structure. Studies of the clinical performance of restorations and prostheses made from these materials have generally focused on quality assessment and survival statistics. Data from these studies should provide probabilities of specific treatment outcomes that are useful for practicing dentists. However, the utility of these data is limited by the lack of national and international standards for assessing these clinical outcomes. Standardized approaches toward clinical informatics and treatment-decision analysis are urgently needed to minimize the variability of clinical outcomes reported in publications associated with direct and indirect restorative materials used for dental restorations and prostheses.
Key Words: Comparative study • dental materials • dental restoration • medical informatics • human • tooth diseases/diagnosis • survival analysis
Dental biomaterials are used clinically to restore function, to enhance esthetics, and to prevent or arrest demineralization of tooth structure. Biomaterials used for these purposes fall into one of two classes, direct restorative materials and indirect restorative materials. The types of materials in each class are listed in Fig. 1  . The results from clinical studies of restoration or prosthesis survival times and overall quality are variable because of changes in material composition and microstructure, the skill level of the dental lab technician and the dentist, the technique sensitivity of the materials, and confounding variables associated with human subjects. Of particular importance is the need for translational research that links the outcomes from randomized and controlled clinical trials to those associated with private practice treatment that is provided under less ideal conditions.
Survival analysis is a relatively recent approach that can be used to estimate the probability of success of dental fillings and prostheses. There are several ways in which survival analysis data can be expressed. The simplest in vitro failure analyses of simple geometries are performed under a variety of test conditions. Ultimately, clinically shaped prostheses are tested under simulated intra-oral conditions, which include a specific environment and loading condition for either one cycle or many cycles. Because the in vitro data are not usually normally distributed, Weibull analyses are performed and fracture stresses are determined as a function of failure probability. Shown in the Table is a list of failure stresses and Weibull moduli of monolithic bar specimens for 8 dental ceramic products, which include 6 core ceramics and 2 veneering ceramics. The data are represented for three levels of failure probabilities—1%, 5%, and 63.2%—adapted from the study by Tinschert et al.(2000). As shown in the Table, the greater the induced stress, the higher the probability of fracture. The Weibull moduli are also important, since they are indirect measures of the flaw distributions in these structures. Smaller moduli indicate a broader distribution of flaws and a greater scatter in the distribution of strength values as a function of failure probability.
What is unknown are the effects of cyclic fatigue and the environment, which were not tested in the study by Tinschert et al.(2000). However, additional tests of the same materials can be performed to address some of these variables. NASA has developed a computer program (CARES/Life) (Sokolowski et al., 1996) that incorporates strength data and stress distribution information determined from finite element stress analysis of ceramic prostheses to predict survival probabilities under a wide variety of simulated clinical conditions. When clinical data are not available for specific types and shapes of prostheses or tooth fillings, these analyses represent a preliminary approach for predicting the relative potential for fracture of these materials. Traditionally, failure data have been presented as mean values of stress or failure load, which are often determined from a static test performed over one stress cycle (loading-failure event-unloading). If survival had been assessed at functional stress levels defined by mean strength values, 50% of the prostheses would have failed. Clearly, this would represent an unacceptable level of clinical performance. However, cyclic loading is responsible for virtually all clinical fractures of ceramic prostheses. It would be more beneficial to analyze the stress levels at which 1% or 5% of the prostheses would have failed. However, since clinical fractures typically occur over many stress cycles, the 1% or 5% failure stresses should be determined from cyclic loading tests.
Based on a clinical study, failures may occur at loads well below those predicted from in vitro data. For example, the fracture of the three-unit fixed partial denture (FPD) in Fig. 2
Although the results from this study are published in a dental journal, this information may not have been readily assessable by practicing dentists if these clinical data were published in an engineering journal. Thus, unless there is a national or international database that can be easily accessed by practicing dentists, failures of the type described will likely continue to occur for a period of time. Do dentists base clinical decisions mostly on personal practice experience or on scientific evidence combined with personal experience? If the former situation applies, it is likely that decision-making errors will occur at higher levels than should occur if scientific evidence was also applied. However, no databases exist to verify this hypothesis.
For optimal clinical decisions to be made regarding restorative treatment, the following steps and information are required:
However, there is considerable variability in the clinical assessment of restoration quality, because there is no standard for judging success or failure. In addition, clinicians are known to disagree with themselves over time and with other technicians at any given time (Bader and Shugars, 1993, 1997; Bader et al., 1995; Shugars and Bader, 1996). Perhaps the best place to begin standardized training of dentists’ clinical decision-making is in dental school. One of the methods that may be useful for such training is the use of relatively simple decision trees such as shown in Fig. 3
An example of this process is presented in Fig. 4 . Here, a noncavitated occlusal caries lesion extending into dentin is presented, and the dentist considers two very conservative options: (1) applying chlorhexidine and fluoride periodically to arrest and possibly remineralize the demineralized area, and (2) applying a sealant to the occlusal fissures. Assuming that there would be a 50% probability of cavitation and the need for subsequent restoration after a period of nine years for the chlorhexidine/fluoride treatments and a probability of only 25% for the sealant option, the probable values for the two treatment options are 75 and 88, respectively. However, what is the source of data for these options? Unfortunately, the data do not exist. In this case, one must make rough estimates of these probabilities based on studies of treatments in similar clinical situations. For example, treatments involving chlorhexidine and fluoride have shown positive outcomes in patients whose saliva flow rates had been severely reduced by irradiation therapy. Sealing of caries lesions has also been reported to be quite successful in one 10-year study (Mertz-Fairhurst et al., 1998). None of the caries lesions that were sealed had progressed over the 10-year period of this study. Confirmation of the efficacy of sealing in the caries lesions was validated by digitizing and analyzing (with use of the CADIA algorithm) the images of the radiographs obtained at each recall appointment (Briley et al., 1997).
However, the condition presented in Fig. 4  is not precisely the same as those associated with the study by Mertz-Fairhurst et al.(1998). Thus, there is a great need for randomized, controlled studies in each of these areas to provide reliable predictions of outcomes under specific conditions of caries risk and lesion characteristics. Treatment decisions should vary depending on the caries risk of the individual patients in a clinical practice. A high-risk patient should not receive costly dental prostheses that are likely to fail because of secondary caries. However, there is a great need for early caries lesion detection devices and tests. Furthermore, a database on practice-based trials is urgently needed to assess treatment outcomes for populations at various risk levels for caries. For this database to be useful, the protocol descriptions and risk classification criteria must be standardized. At least two objective examiners must be used as well as a human filter to screen the clinical data and to develop models to ensure that treatment protocols and outcomes are more consistent. A national center should be established for storage and sharing of clinical databases so that filtered information can be made available to clinicians on a timely basis. However, clinicians must also be informed frequently of these clinical research findings and their implications, so that changes in treatment philosophies can be made in a timelier manner. Recent surveys suggest that clinicians either are unaware of scientific evidence or are unwilling to change their decision-making options based on this evidence. For example, explorers are still used for caries lesion detection, sealants are underutilized, marginal gap size is used as a predictor of caries processes, noncavitated tooth surfaces are being restored, and the caries disease process is still treated by restoring teeth. These traditions continue in spite of a growing body of evidence that supports minimally invasive treatment choices.
Another important question is, "What is the survival probability of restorative treatments if a restoration option is chosen in lieu of the two conservative options proposed in Fig. 4 Systematic reviews of clinical studies are intended to eliminate poorly designed or performed clinical trials and to allow for comparisons of data from properly designed clinical studies. However, systematic reviews of several hundred publications may yield a very small number of acceptable clinical studies. Consider, for example, the study by Hayashi and Yeung (2003). To compare the effectiveness of ceramic inlays in posterior teeth with other posterior restorations, these investigators searched the Cochrane Oral Health Group Trials Register, the Cochrane Central Register of Controlled Trials (The Cochrane Library, Issue 1, 2002), MEDLINE, and EMBASE from 1990 to 2001. For inclusion, each study must have been performed as a randomized controlled trial in which the longevity of ceramic inlays was compared with those of other types of posterior restorations. Only one study met the criteria for inclusion in the review. In this study, evaluation of 60 ceramic inlays and 20 gold inlays over a five-year period revealed 11.7% failures for the ceramic inlays and 10% failures for the gold inlays. The statistical power of this study was insufficient to detect statistically significant differences in longevity or post-operative sensitivity. These authors concluded that there is a great need for optimal design and data reporting for future trials of dental ceramics.
Although the 10-year clinical survival probabilities for amalgam and composite restorations are quite variable (Fig. 5
The greatest uncertainty overall in estimating the survival times for dental restorations is associated with the lack of precision associated with detection of caries lesions and the lack of a standard for classifying the caries risk of individual patients. If restorative materials are placed inappropriately in teeth with no caries lesions (false positives), the efficacy of these materials in preventing secondary caries will be overestimated. Conversely, if a treatment is selected for the prevention of primary caries, and several lesions were undetected at baseline, the efficacy of the preventive treatment will be underestimated. In either case, treatment decisions involving restorative materials will be flawed. Thus, there is a critical need for improved diagnostic methods with much greater sensitivity and specificity. In the absence of such technology, restorative treatments must be based on maximum conservation of tooth structure and optimal control of disease-promoting factors. Although evidence-based reviews of clinical performance provide the most standardized method of comparing different restorative materials and techniques, other long-term studies (lasting longer than five years) offer additional information of value to clinicians. However, a secondary set of inclusion and exclusion criteria must be formulated to answer specific questions that cannot be answered by the limited number of evidence-based reviews. Many long-term studies on restoration survival should be considered further, based on less restrictive criteria (Smales, 1991; Creugers et al., 1992; Qvist and Strom, 1993; Tolman and Laney, 1993; Mahmood and Smales, 1994; Andreasen et al., 1995; Decock et al., 1996; Einwag and Dunninger, 1996; Buser et al., 1997; Millar et al., 1997; Probster and Henrich, 1997; Roulet, 1997; Kreulen et al., 1998; Plasmans et al., 1998; Djemal et al., 1999; Dumfahrt, 1999; Lekholm et al., 1999; Priest, 1999; Raskin et al., 1999, 2000; Frankenberger et al., 2000; Nicolaisen et al., 2000; Reiss and Walther, 2000; Gaengler et al., 2001; Kindberg et al., 2001; Malament and Socransky, 2001; Norton, 2001; Reiss, 2001; Bogacki et al., 2002; El-Mowafy and Brochu, 2002; Otto and De Nisco, 2002; Sethi et al., 2002; Sjögren and Halling, 2002; Holm et al., 2003; Malament et al., 2003; Olsson et al., 2003; Wagner et al., 2003; Zalkind et al., 2003).
Many in vivo studies of restorative dental materials have been performed, but few meet the rigorous criteria for inclusion in systematic reviews of the literature. Even when most criteria have been met, the statistical power may be inadequate for dentists to judge whether differences in outcome measures are statistically significant. Thus, dentists are faced with the challenge of deciding whether they should rely on poorly designed or poorly performed clinical trials to assist in their decision-making processes. Lacking any clinical data, in vitro or in situ data may be needed to screen out poorly performing materials. One solution may solve this major problem: Establishing a national center as a clearinghouse for experimental designs and for storage of clinical research databases should ensure the accessibility of more consistent studies that are appropriate for inclusion in subsequent systematic reviews. In addition, much greater emphasis should be placed on courses in dental schools that provide instruction on (1) clinical study design, (2) interpretation of the results of clinical trials, (3) comparative analyses of clinical data from many studies, and (4) the appropriate use of informatics systems for data acquisition, data analysis, and development of appropriate treatment models to enhance transfer of this information to clinical practices.
Publication supported by Software of Excellence (Auckland, NZ)
Advances in Dental Research, Vol. 17, No. 1,
43-48 (2003)
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
Abstract
Introduction
