Chapter  36:  Diagnosis and Management of Dental Caries: Evidence Report/Technology Assessment Number 36

Search Process and Inclusion Criteria

The Evidence-based Practice Center (EPC) review and investigative team conducted two detailed searches of the relevant English language literature from 1966 to October 1999 using MEDLINE, EMBASE, and the Cochrane Controlled Trials Register. The team did not pursue reports in the gray literature (i.e., information not reported in the periodic scientific literature). The team hand-searched current journals up to the end of 1999.

One search focused on the following diagnostic methods -- visual and visual tactile inspection, radiography, fiberoptic transillumination, electrical conductance, laser fluorescence, and combinations of these methods -- using keywords for the disease (dental caries, tooth demineralization), diagnostic concepts (oral diagnosis, oral pathology, dental radiography), and study characteristics and design.

A second search focused on dental caries preventive or management methods, using keywords for methods (fluorides, pit and fissure sealants, health education, dental prophylaxis, oral hygiene, dental plaque, chlorhexidine dental sealants, cariostatic agents) and study characteristics and design in addition to the disease keywords.

The EPC team applied several inclusion and exclusion criteria to the reports identified in our literature search. The team included studies in the diagnostic review that used histological validation of caries status, and either reported results as sensitivity and specificity of the diagnosis or reported data from which these measures could be calculated. The team excluded reports of diagnostic methods not commercially available. For the review of the dental caries management literature, the team included only reports concerning methods applied or prescribed in a professional setting, and only studies performed in vivo and having a comparison group.

The two disease management questions that were addressed by the team used the results of the management review and featured additional inclusion criteria. For the management of non-cavitated carious lesions, the team included only studies where the lesion was the unit of analysis. The team accepted several different descriptions of noncavitated lesions (including the terms "incipient" and "initial)." From the literature describing the management of subjects at elevated risk for dental caries, the team included only studies where the classification of elevated risk had been made for individual subjects and was based on carious lesion experience and/or bacteriological testing. The team accepted the elevated risk classification described in the paper.

The EPC team selected studies for inclusion from among 1,407 diagnostic and 1,478 management reports through independent duplicate reviews of titles, abstracts, and, where necessary, full papers, with discussion leading to consensus where disagreement occurred. Two team reviewers agreed on inclusion status for 97 percent of the reports at this stage. In addition, the reviewers separately identified six studies evaluating preventive methods in patients who had received radiotherapy for head and neck neoplasms (a special high-risk group) and seven studies evaluating preventive methods in patients with orthodontic bands or brackets (another special high-risk group). The team believed that these studies should be included in the review, but not combined with the main group of studies due to substantial differences in lesions and study methods.

The team abstracted data (single abstraction, subsequent independent review) on 39 diagnostic studies and 27 management studies, using different forms for the diagnostic and management studies. Four reviewers were involved in the abstraction process, with reviewer agreement rates of 100 percent for results and 88 percent for other study descriptors. Separate quality rating forms were completed by the EPC team's scientific director for the two types of studies. The quality rating scales assessed several elements of internal validity, including study design, duration, sample size, blinding, baseline assessments of differences among groups, loss to followup, and examiner reliability. Two items also requested the reviewer's subjective assessment of both the internal and external validity of the study.

The team compiled the abstracted data in a series of six evidence tables, one each for in vivo and in vitro radiographic studies, studies of management of noncavitated carious lesions and individuals at elevated risk for carious lesions, and studies of special populations of orthodontic patients and patients who received head and neck radiotherapy. The team then graded the evidence summarized in the tables.

For the diagnostic question, the strength of the evidence was judged in terms of the extent to which it offered a clear, unambiguous assessment of the validity of a particular method for identifying a specific type of lesion on a specific type of surface. The three possible ratings were:

• Good (A): The number of studies is large, the quality of the studies is generally high, and the results of the studies represent narrow ranges of observed sensitivity and specificity.

• Fair (B): There are at least three studies, the quality of the studies is at least average, and the results represent moderate ranges of observed sensitivity and specificity.

• Poor (C): There are fewer than three studies, or the quality of the available studies is generally lower than average, and/or the results represent wide ranges of observed sensitivities and/or specificities.

For purposes of this question, a narrow range is defined as no more than 0.15 on a scale of 0.0 to 1.00, a moderate range is no more than 0.35, and a wide range is more than 0.35. High quality is defined as most study scores at or above 60, and average quality is defined as most study scores at or above 45.

For the management studies, the team used a scheme based on several considerations, including the magnitude of the results reported, the quality rating scores of the studies, the number of studies, and the consistency of the results across studies. The EPC team's scientific and clinical directors independently rated the interventions and developed an adjudicated final rating. The four possible ratings were:

• Good (A): Data are sufficient for evaluating efficacy. The sample size is substantial, the data are consistent, and the findings indicate that the intervention is clearly superior to the placebo/usual care alternative.

• Fair (B): Data are sufficient for evaluating efficacy. The sample size is substantial, but the data show some inconsistencies in outcomes between intervention and placebo/usual care groups such that efficacy is not clearly established.

• Poor (C): Data are sufficient for evaluating efficacy. The sample size is sufficient, but the data show that the intervention is no more efficacious than placebo or usual care.

• Incomplete Evidence (I) Data are insufficient for assessing the efficacy of the intervention, based on limited sample size and/or poor methodology.

Diagnostic Methods

The EPC team evaluated the strength of the evidence describing the performance of diagnostic methods separately for cavitated lesions, lesions involving dentin, enamel lesions, and any lesions. The team also separated the evaluations by the surface and tooth type involved. The team found 39 studies reporting 126 histologically validated assessments of diagnostic methods.

• There are few assessments of the performance of any diagnostic methods for primary or anterior teeth and no assessments of performance on root surfaces. The strength of the evidence describing the performance of any method for these teeth and surfaces is poor.

• Among studies assessing diagnostic performance for proximal and occlusal surfaces in posterior teeth, the team rated the strength of the evidence describing the performance of visual/tactile, FOTI, and laser fluorescence methods as poor due to the small numbers of studies available.

• The team also rated the strength of the evidence for radiographic, visual, and EC methods as poor for all types of lesions on posterior proximal and occlusal surfaces. However, these ratings were due less to inadequate numbers of assessments than to variation among reported results. In one instance, the quality of the available studies was the principal reason for the rating.

• For all but EC assessments, specificity of a diagnostic method was generally higher than sensitivity. Thus, false negative diagnoses are proportionally more apt to occur in the presence of disease than are false positive diagnoses in the absence of disease.

• The evidence did not support the superiority of either visual or visual/tactile methods. The number of available assessments was small and there was substantial variation among reports for each method.

• The evidence suggests, but is not conclusive, that some digital radiographic methods offer small gains in sensitivity compared with conventional film radiography on both proximal and occlusal surfaces.

• The evidence also suggests, but is not conclusive, that EC methods may offer heightened sensitivity on occlusal surfaces, but at the expense of specificity.

• The diagnostic performance literature is limited in terms of numbers of available assessments for most diagnostic techniques overall, and especially for primary teeth, anterior teeth, and root surfaces and for visual/tactile and FOTI methods. The literature is further limited by threats to both internal and external validity represented by incomplete descriptions of selection and diagnostic criteria and examiner reliability, the use of small numbers of examiners, nonrepresentative teeth, samples with high lesion prevalence, and a variety of reference standards of unknown reliability.

Management of Noncavitated Carious Lesions

We found only five studies addressing this topic. The evidence was rated as incomplete.

This literature is limited by:

• Differences in treatment provided to comparison groups and in how noncavitated lesions are defined.

• Problems in the identification and control of patient exposure to community-based and individual preventive dental procedures.

• High loss to followup due in part to limiting analyses only to full participants.

All of these limitations make drawing conclusions difficult.

Management of Caries-Active Individuals

The EPC team evaluated the evidence for nine management methods: fluoride varnishes, fluoride topical solutions, fluoride rinses, chlorhexidine varnishes, chlorhexidine topicals, chlorhexidine rinses, combined chlorhexidine-fluoride applications, sealants, and other approaches. The team based its review on 22 studies that described 29 experimental interventions evaluating these. The team also examined 13 studies of special at-risk populations (orthodontic and head and neck radiotherapy patients).

• The team rated the evidence for the efficacy of fluoride varnishes as fair, and the evidence for all other methods as incomplete.

• The evidence for efficacy was suggestive for chlorhexidine varnishes and gels, for combination treatments including chlorhexidine, and for sucrose-free gum, but in each instance the number of studies was too small or the results were too variable to be conclusive.

• Among subjects undergoing orthodontic treatment with attached bands or brackets, the team found the evidence for efficacy of fluoride interventions to be suggestive but incomplete. Evidence was also incomplete for all other prevention methods for these subjects.

• Among patients receiving head and neck radiotherapy, the literature offers fair evidence of the efficacy of fluoride-based interventions. The evidence was incomplete for any other types of preventive interventions among these patients.

• The team found no reports of substantive harms associated with any interventions.

• The team found the number of available studies for any specific method to be a serious limitation. Among studies addressing a method, the variety of experimental protocols, comparison groups, and other community and individual preventive dentistry exposures further restricted the opportunity to draw conclusions about the efficacy of the method. Finally, generalization from the studies to the broader U.S. population is problematic, as nearly all studies included only children and evaluated changes only in the permanent dentition.

Prevalence of Carious Lesions

The Third National Health and Nutrition Examination Survey-Phase I (NHANES III), conducted from 1988 to 1991, provides the most recent estimates of the prevalence of carious lesions in the United States.^9,10 This survey, which produced nationally representative estimates for the civilian noninstitutionalized U.S. population, found that the mean decayed and filled surfaces (dfs) of primary teeth score in children age 2 to 9 was 3.1. The score varied among racial-ethnic categories: non-Hispanic whites (2.5), non-Hispanic blacks (2.7), and Mexican-Americans (4.8). Overall, 83 percent of children age 2 to 4 years had experienced no lesions in the primary dentition, with this percentage dropping to 50 percent in children 5 to 9 years of age.

Table 1. Mean DS, DMFS, and % DS/DMFS per person by race-ethnicity (more...)

Table 1. Mean DS, DMFS, and % DS/DMFS per person by race-ethnicity in U.S. children and adolescents age 5 to 17, 1988-91

Race/Ethnicity	DS (SE)	DMFS (SE)	% DS/DMFS (SE) 1
Total	0.4 (0.1)	2.5 (0.2)	19.7 (1.5)
Non-Hispanic Whites	0.3 (0.0)	2.4 (0.3)	14.6 (2.2)
Non-Hispanic Blacks	0.8 (0.1)	2.5 (0.2)	37.9 (3.1)
Mexican-Americans	0.7 (0.1)	2.7 (0.1)	36.4 (2.8)

1

With at least one decayed or filled surface.

DS - decayed surfaces on permanent teeth.

DMFS - decayed, missing, and filled surfaces on permanent teeth.

%DS/DMFS - percent of DMF surfaces composed of decayed surfaces.

In children age 5 to 11, 74 percent had no carious lesions in the permanent dentition, whereas for children age 12 to 17, the proportion falls to 33 percent. Overall, carious lesions in children are not evenly distributed; 75 percent of overall caries experience in permanent teeth occurred in approximately 25 percent of the population.³ Thus, although the majority of children have moderate decay or less, carious lesions are a recurring problem for a substantial minority. Carious lesions are also not equally distributed across tooth surfaces in this population, as occlusal surfaces experience lesions five times more frequently than the mesial and distal (smooth) surfaces.

Table 2. Mean DS, DFS, and %DS/DFS per person by race-ethnicity in (more...)

Table 2. Mean DS, DFS, and %DS/DFS per person by race-ethnicity in U.S. dentate adults, ages 18 Years and older, 1988-91

Race/Ethnicity	DS (SE)	DFS (SE)	%DS/DFS (SE) 1
Total	1.8 (0.0)	22.2 (0.9)	14.2 (0.8)
Non-Hispanic Whites	1.5 (0.1)	24.3 (1.5)	10.6 (0.8)
Non-Hispanic Blacks	3.4 (0.3)	11.9 (3.4)	35.4 (2.8)
Mexican-Americans	2.8 (0.3)	14.1 (2.8)	31.0 (2.6)

1

With at least one decayed or filled surface.

DS - decayed surfaces on permanent teeth.

DFS - decayed, and filled surfaces on permanent teeth.

%DS/DFS - percent of DFS surfaces composed of decayed surfaces.

Carious root lesions were found in 23 percent of the dentate population overall and in more than 47 percent of individuals 65 or older. The average number of treated and untreated root surfaces ranged from a low in non-Hispanic whites of 1.1 to 1.4 in Mexican-Americans and 1.6 in non-Hispanic blacks.

Burden of Illness from Dental Caries

Because treatment and/or prevention of carious lesions is one of several reasons for visiting a dentist, and because accurate information describing reasons for dental visits is not available, the extent to which carious lesions necessitate dental visits is not known. However, from the preceding section on the prevalence of carious lesions and filled tooth surfaces, it would seem that racial and ethnic minorities receive proportionally less treatment for carious lesions than do white non-Hispanics. This observation is supported from existing data on dental visits. The Surgeon General's report on oral health¹¹ has assembled data from a number of Federal agencies that paint a clear picture of differences in the receipt of oral health care by race/ethnicity, income, and insurance status. In white, non-Hispanic adults, 64 percent reported a dental visit in 1993 compared with 47 percent of black, non-Hispanic adults and 46 percent of Hispanic adults. For individuals with incomes at or above poverty level, 64 percent reported a dental visit in the previous year compared with 36 percent for those with incomes below poverty level. In 1989, 70 percent of individuals 2 years and older with private dental insurance reported a dental visit within the preceding year compared with 51 percent for those without private dental insurance. Finally, those individuals who rate their oral health as very good or excellent are more likely to have visited a dentist in the preceding year (61 percent) than were those assessing their health as fair or poor (45 percent).

The economic cost of dental caries is also difficult to assess precisely. In 1998, Americans spent more than $53 billion on dental services.¹¹ From an analysis of insurance claims, approximately 40 percent of charges are related to restorative dental services, which are usually, but not always, performed to repair teeth damaged by carious lesions.¹² Thus, even without adding in the cost of more complex services necessary to restore function lost as a result of the sequalae of the dental caries process, expenditures related to caries were more than $20 billion.

Beyond the direct economic costs of dental treatment, there are the less directly calculable costs associated with the loss of working time, missed school, and reduced levels of social functioning. The National Health Interview Survey (NHIS) indicates that 2.9 million acute dental conditions occurred in the U.S. population during 1994. These dental conditions accounted for an estimated 3.9 million days of missed work in persons 18 years of age and over, 1.2 million days of missed school in youth 5 to 17 years of age, and 12.2 million days of restricted activity across all ages (e.g., nonperformance of usual family role activities).¹³ The NHIS methods may underestimate the actual amount of missed time from school and work and restricted activity days for dental conditions.^14,15

Studies of how dental caries affects quality of life are much less empirically compelling, but experts agree on what the potential effects of dental caries are likely to be in the short and long term.¹⁶ In the short term, physical discomfort and pain are the most likely consequences of untreated lesions. Physical impacts can be felt directly as through the pain of toothaches, infections, and temporomandibular joint disorder resulting in part from a loss of posterior teeth and the failure to replace them when necessary. The possible eventual inability to eat -- both bite and chew -- because of tooth loss can lead to unnecessary dietary restrictions and nutritional deficiencies as well as complicate the dietary management of other chronic health conditions.

The psychological pain of self-consciousness and social isolation may also accompany the embarrassment of the unsightly deterioration of anterior teeth caused by dental caries. The same psychological distress can result from the embarrassment of missing anterior teeth, the communication dysfunction associated with not being able to be easily understood by others, and the isolation or withdrawal from social intercourse because of missing teeth. In the long term, left untreated, carious lesions may lead to the loss of such teeth, the replacement of which may be needed for functional, social, cosmetic, and physical and mental health reasons.

Caries Diagnosis

The local result of the dental caries infection is a process of demineralization of tooth tissue (enamel, dentin, cementum). Acid produced by bacteria as a product of carbohydrate fermentation causes the demineralization. The diagnosis of dental caries at a particular site on a tooth is based on either direct or indirect detection of the demineralized tooth structure.

The principal methods dentists use to diagnose carious lesions -- visual and visual/tactile examinations and radiographic assessment -- have been employed with little change for decades. Refinement in techniques, rather than development of new technology, has characterized these methods over the years. Illumination has improved and magnification is more easily employed for visual examinations, whereas radiation doses have decreased for radiographic assessment as both equipment and film have been improved.

Visual inspection is based on a search for signs of demineralization, which include changes in color and in surface consistency and contour. Tactile inspection is usually accomplished with a fine-tipped dental explorer or probe that is passed over smooth surfaces of teeth as well as pits and fissures. On smooth surfaces, the surface texture is assessed for roughness as well as breaks in contour. In pits and fissures, the probe is usually pressed with differing levels of force into depressed areas to assess whether any penetration is possible and whether there is any resistance to withdrawal of the probe. Radiographic assessment is based on identification of demineralization of tooth tissue through differential exposure of film. Demineralized tooth tissue is less resistant to the passage of ionizing radiation and thus appears darker on film images.

More recently, new technologies have begun to appear that further refine radiographic diagnosis of carious lesions and offer alternatives to this technology. Digital radiographic techniques eliminate film by capturing radiographic images on phosphor storage plates or charge-coupled devices. The images can then be manipulated to enhance diagnostic features. Fiberoptic transillumination, passing a narrow beam of light through tooth tissue, has become an adjunctive diagnostic method now used for both anterior and posterior teeth, principally on proximal surfaces. Demineralized tooth tissue appears dark when transilluminated because of its decreased transmission of light. This method represents refinement of the traditional technique of transilluminating the anterior proximal surfaces using a mouth mirror and the operatory light. Measuring the resistance of tooth tissue to an electrical current passed through it is another approach to caries diagnosis, especially of occlusal fissure caries. First demonstrated in the 1950s, the technique has been progressively refined, with devices available commercially since the 1980s. The technique depends on the fact that when enamel becomes demineralized, it loses much of its resistance to electrical charges, hence its conductance increases.

The extent of variation in the diagnosis of dental caries is substantial among dental practitioners using the traditional techniques. Typically agreement among several dentists is poor to moderate, with kappa values ranging from 0.30 to 0.60 in several studies.⁵ The range of positive diagnoses (proportion of teeth diagnosed as carious) is typically wide for any given sample, often spanning 30 to 40 percentage points.⁵ The problem of calibrating dental practitioners to an objective standard for caries diagnosis results to a large extent from the absence of objective criteria for the diagnosis;^17,18 consequently, dentists tend to develop widely different subjective patterns or "scripts" that they then use for identification of carious lesions.¹⁹ This variation in the diagnosis of carious lesions is a principal contributor to the still greater variation in the decision to restore teeth through irreversible surgical intervention^20,21 and the concomitant variation in associated costs of those decisions.²²

Variation in Methods to Control Noncavitated Lesions

Little is known about dentists' strategies to reduce or eliminate progression of noncavitated carious lesions and hence the necessity for surgical intervention. The previously cited literature on variation in dentists' decisions to initiate treatment includes some studies of dentists' treatment thresholds. These studies suggest there is variation in the extent of progression of a carious lesion that individual dentists are willing to tolerate before they intervene surgically. Unfortunately, these types of studies must be done using patient vignettes, and there is some suggestion that what dentists say they do with respect to intervention is often different than what they actually do in practice.^23-26 These studies show that a sizable proportion of dentists routinely intervene when radiographic evidence of dental caries manifests itself in the enamel prior to cavitation. No recent studies are available to document circumstances surrounding application of nonsurgical means of control, although the continuing controversy about "sealing over caries"²⁷ suggests that dentists vary in their willingness to use sealants as a method for the control of unidentified occlusal lesions.

Variation in Methods to Control Caries in Caries-Active Individuals

Knowledge of dentists' practices in addressing caries control in caries-active individuals is exceedingly limited. Only recently have the concepts of "caries risk" and "medical management" emerged in the clinical dental literature.^4,7,28,29 These discussions suggest that practitioners' preventive approaches may not be routinely based on a careful assessment of the magnitude of the caries challenge. Information from insurance claims suggests that topical fluoride applications tend to vary by practitioner, but not by patients within a practice, who all receive the same preventive care even though they have different rates for restoration receipt.³⁰ Also, a survey of practitioners shows that commonly used clinical protocols are not congruent with current recommendations for low-risk individuals.³¹ Clearly, there is a potential for both over use and under use of prevention and control methods in a caries-active population; but studies that examine dentists' preventive treatment behaviors are rare, and none differentiate treatment by an individual's caries activity.

Final Key Questions

The key questions address issues of caries diagnosis and management that arise in the professional treatment of dental caries, i.e., those procedures that are provided by dentists and allied dental personnel in dental practices and clinics. Thus, the procedures are limited to those commercially available at the time of this review. Further, the caries management questions focus on issues that accompany the "modern" view of dental caries as an oral infection that, at specific sites, initially leads to demineralization and ultimately destruction of tooth tissue. The key questions, stated in final form, are as follows:

• Question 1.

• What are the validities of the available diagnostic methods for detecting carious lesions in primary and permanent teeth?

• Question 2.

• What are the efficacies of the nonsurgical methods available for stopping or reversing the progression of a noncavitated coronal carious lesion in a primary or a permanent tooth?

• Question 3.

• What are the efficacies of the methods available for reducing the incidence of new coronal carious lesions in primary and permanent teeth in individuals who are deemed to be "caries active" or at "high caries risk"?

The first question addresses only the diagnosis of primary caries, i.e., the first carious lesion on a tooth surface. Both coronal and root surfaces are included in the review, and for coronal surfaces, both primary and permanent teeth are included. Following discussion with the TEAG, assessment of test validity was operationalized as the sensitivity and specificity of a diagnostic test. The methods to be assessed included all those diagnostic methods that are commercially available, including visual and visual-tactile inspection, radiography, FOTI, EC, laser fluorescence, and combinations of those methods.

The second question focuses on individual early carious lesions, where demineralization has occurred but cavitation has not yet occurred. In the past, this type of lesion was either removed surgically and replaced with a restoration or monitored or "watched." Dentists generally assumed that many noncavitated lesions would progress to cavitation, and based treatment decisions on this assumption. More recently, the possibility of remineralizing or at least arresting the demineralization of these noncavitated lesions has been considered as an alternative to surgical removal and restoration. Another nonsurgical technique, placing dental sealants, is also available for noncavitated lesions on fissured surfaces. The question includes consideration of a still wider range of potentially useful methods, including professional fluoride applications and prescribed supplements, other remineralization agents, professional oral hygiene and plaque control programs, and combinations of these methods.

The third question focuses on patients rather than individual carious lesions. It reflects the need for information about how to manage patients who have active carious lesions or who are at risk of developing such lesions. Recommendations for the "medical management" of such patients have appeared; yet the methods to be included in such an approach are not well defined. This question includes consideration of professional fluoride applications and prescribed supplements, sealants, antimicrobial therapy, salivary enhancements, nutritional/diet counseling, professional oral hygiene/plaque control programs, and combinations of these methods.

Causal Pathways

Figure 1. Causal pathways for the diagnosis, nonsurgical management, (more...)

   Figure 1. Causal pathways for the diagnosis, nonsurgical management, and prevention of carious lesions

1 Determined separately for primary teeth and for coronal and root surfaces of permanent teeth.

2 Methods to be evaluated are radiographs, visual and visual/tactile inspection, FOTI, EC, fluorescence, and combinations of these methods.

3 Risks associated with the diagnostic procedure may include exposure to ionizing radiation, inadvertent cavitation, and site inoculation as well as outcomes of false positive and false negative diagnoses.

4 Validity evaluated as sensitivity and specificity.

5 Methods to be evaluated include professional (in-office) fluoride applications and/or prescribed supplements, other remineralization agents, sealants, professional oral hygiene/plaque control programs, and combinations of these methods.

6 Risks associated with the nonsurgical treatment intervention may include drug reactions, loss of tooth pulp, and dentist concern over "failure to treat."

7 Methods to be evaluated include professional (in-office) fluoride applications and/or prescribed supplements, sealants, antimicrobials, salivary enhancers, nutritional/diet counseling, professional oral hygiene/plaque control programs, and combinations of these methods.

8 Caries-active designation operationalized as existing caries lesion(s) and/or high caries-risk level based on criteria other than current lesions.

9 Risks associated with infectious disease management strategy include drug reactions, loss of tooth pulp, and dentist concern over "failure to treat."

Figure 1

The degree of penetration of the lesion is thought to be the principal criterion that most dentists use in making treatment decisions, with penetration to the dentin seemingly the threshold for restoration reported most often. In view of caries progression, whether a lesion is cavitated or not may represent a more logical criterion for differentiating between opportunities to arrest or reverse caries progression nonsurgically and the necessity for removal of the lesion and replacement of the lost tissue. The use of dentin penetration as the surgical intervention criterion may result in the treatment of noncavitated, potentially reversible lesions. The causal pathway reflects the lack of a cavitation criterion for nonsurgical intervention.

For those patients found to have one or more carious lesions, in addition to surgical or nonsurgical treatment directed specifically at the lesion(s), there is an opportunity to provide treatment for the purpose of reducing the likelihood for the development of further lesions. As noted, although dentists have long provided professional preventive procedures, linking the provision of these procedures to a patient's caries activity status, when it has been done, usually has been done informally, with little knowledge of the effectiveness of such preventive procedures in patients with high rates of disease. Extending this type of targeted intensified prevention to patients identified as being at risk for the development of carious lesions is less common. Caries risk assessment is a relatively recent development in dentistry; and even though a number of risk assessment instruments have been described, the approach has not been validated when applied to individual patients.

Search Terms

Table 3. Strategy and results of MEDLINE caries diagnosis search (more...)

Table 3. Strategy and results of MEDLINE caries diagnosis search

Wide search of early caries literature
1  exp dental caries/pa,di.ra	2,846
2  limit to human, English, 1966-75	219
Defining studies of caries
3  exp tooth demineralization/pa,di,ra	2,928
4  exp dental caries/	21,830
5  3 or 4	21,904
Limiting 5 to diagnostic methods
6  exp diagnosis/, oral diagnosis/	2,420
7  exp radiography/, dental radiography/, digital dental radiology/	816
8  exp pathology/, oral pathology/	4
9  1 or 6 or 7 or 8	2,539
10  limit to human, English	1,776
Limiting 10 to various study types
11  controlled clinical trial	21
12  meta analysis	4
13  randomized controlled trial	50
14  epidemiologic study characteristics	244
15  epidemiologic research design	333
16  comparative study	457
Combining results of 1966-75 "wide" search and searches for specific study types
17  2 or 11 or 12 or 13 or 14 or 15 or 16	1,266
Adding all root caries studies
18  exp root caries/pa,di,ra	62
Total	1,328

exp: explode (i.e., include all subheadings, or those listed after the slash)

pa: pathology

di: diagnosis

ra: radiography

Table 4. Strategy and results of MEDLINE caries management search (more...)

Table 4. Strategy and results of MEDLINE caries management search

Identifying management methods
1  exp fluorides, topical/tu	2,061
2  exp tooth remineralization/	445
3  exp pit and fissure sealants/tu	667
4  exp health education, dental/	4,287
5  exp dental prophylaxis/	3,699
6  exp oral hygiene/	8,624
8  exp dental plaque/pc,dh,dt,th	3,423
9  exp chlorhexidine/tu	1,126
10  exp xylitol/tu	162
11  exp tooth demineralization/pc,dt,th	10,162
12  exp cariostatic agents/tu	3,994
13  fluoride supplements	60
14  1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13	26,902
Identifying caries management methods
15  exp dental caries/pc,dh,dt,th	10,064
16  14 and 15	10,058
17  limit to human, English	5,057
Limiting 17 to various study types
18   controlled clinical trial	122
19   randomized clinical trial	177
20   epidemiologic study characteristics	762
21  epidemiologic research design	266
22  comparative study	758
Total of 18 or 19 or 20 or 21 or 22	1,435

exp: explode (i.e., include all subheadings, or those listed after the slash)

tu: therapeutic use

pc: prevention and control

dh: diet therapy

dt: drug therapy

th: therapy

In the absence of effective exclusion criteria available in MEDLINE, it still might have been possible to design relatively "tight" search strategies if certain critical keywords were available to narrow the search focus. Unfortunately, this was not the case for either search. The diagnosis search returned a large number of potentially eligible studies (1,328) because preliminary searches had demonstrated that a key term, "sensitivity and specificity," could not be assumed to identify accurately all eligible studies.

In the management search, two critical features of eligible studies could not be isolated through use of indexing terms. Neither noncavitated lesions nor caries-active or "at-risk" patients are identifiable through the keyword structure. Thus, the management search had to be designed to identify all possible evaluations of the eligible preventive methods, with subsequent inspection of the abstract or full paper required for a final determination of eligibility for either of the systematic reviews based on patient sample or type of lesion included. The result was the identification of 1,435 citations.

Additional Searching

Table 5. Strategy and results for EMBASE caries diagnosis search (more...)

Table 5. Strategy and results for EMBASE caries diagnosis search

1. dental adjacent to caries	1,554
2. diagnosis	248,652
3. dental radiography	121
4. 2 and 3	248,677
5. 1 and 4	87
6. New citations added (not duplicates with MEDLINE)	79

Source: Excerpta Medica, copyright Elsevier Science B.V.

Table 6. Strategy and results for EMBASE caries management search (more...)

Table 6. Strategy and results for EMBASE caries management search

1. dental adjacent to caries	1,554
2. topical fluorides	6
3. remineralizaion	79
4. dental sealants	13
5. sealants	96
6. chlorhexidine	0
7. cariostatic agents	0
8. 2 or 3 or 4 or 5 or 6 or 7	181
9. 1 and 8	48
10. New citations added (not duplicates with MEDLINE)	43

Source: Excerpt Medica, copyright Elsevier Science B.V.

A valuable supplemental search strategy was perusal of the reference sections of papers identified in the searches. Again, the reason for the seeming inefficiency of the MEDLINE searches is in large measure the imprecise indexing characteristics of dental studies in the 1970s and 1980s. Not only are descriptors of study design characteristics inexact or missing, but descriptors related to the condition or process of interest are also often tangential in nature. This forces the search to be less exclusive and at the same time increases the likelihood that some studies will be missed, even with a fairly broad search strategy such as the one employed.

The team had elected at the outset not to complete a detailed search of the gray literature. This is information not appearing in the periodic scientific literature, such as dissertations, theses, industry reports, unpublished studies, abstracts, and other nontraditional sources. The team made this decision because of both limited resources and the prevailing experience opinion among RTI-UNC EPC staff that in the absence of known sources for such literature, searches were unlikely to yield useful information. The team did query NIDCR to identify any in-progress studies that might have recently reported relevant data. The team did not identify other potentially fruitful sources for gray literature for dental topics. Thus, no other sources were searched.

Finally, because the addition of dental articles to the MEDLINE database tends to lag behind publication date by at least 6 months, the team hand-searched six of the more fruitful journals for relevant studies for the period January 1998-December 1999 (Caries Research, Community Dentistry Oral Epidemiology, European Journal of Oral Sciences, Journal of Dental Research, Journal of Dentistry, and Journal of Public Health Dentistry).

Inclusion and Exclusion Criteria for the Literature Searches

Table 7. Inclusion (I) and exclusion (E) criteria for (more...)

Table 7. Inclusion (I) and exclusion (E) criteria for caries diagnosis studies

(1)  Diagnostic question

I  accuracy of the determination of the presence/absence of natural caries at defined levels

E  accuracy of measure of caries depth, volume, extent, etc. (exclude if only question addressed)

E  artificial caries achieved through demineralization or drilling

(2)  Diagnostic method

I  visual or visual and tactile inspection

I  film radiography with D- or E-speed film

I  digital radiography (charged coupled devices, storage phosphor screens)

I  fiberoptic transillumination (FOTI)

I  electrical conductance techniques (Vanguard, Caries Meter L, ECM, etc.)

I  fluorescence/optical techniques if commercially available

E  methods using equipment not available commercially

(3)  Validation technique

I  sectioning with visual inspection, microscopy, stereomicroscopy, or macroradiography

I  visual/tactile inspection of intact surface for cavitation only

E  visual or visual/tactile inspection for caries level other than cavitation

E  radiography or other nonhistologic technique for caries level other than cavitation

(4)  Sensitivity/specificity determination

I  reported or calculable from results presented

E  not determinable

Table 8. Inclusion (I) and exclusion (E) criteria for noncavitated (more...)

Table 8. Inclusion (I) and exclusion (E) criteria for noncavitated lesions studies

(1)  Study design

I  in vivo studies

I  studies with concurrent comparison group (nil, placebo, or active)

I  studies with the lesion as the unit of analysis

E  in vitro, in situ studies

E  studies without concurrent comparison

E  studies without baseline determination of individual lesion status

(2)  Intervention

I  interventions requiring professional application and/or prescription

I  interventions likely to be undertaken only upon the recommendation of a dentist

E  fluoride dentifrice studies regardless of concentration, if only intervention/control component

(3)  Sample size

no exclusion criterion

(4)  Outcome

I  outcome expressed or calculable as percent of lesions identified at baseline that progressed

Table 9. Inclusion (I) and exclusion (E) criteria for (more...)

Table 9. Inclusion (I) and exclusion (E) criteria for studies of caries prevention in caries-active individuals

(1)  Study design

I  in vivo studies

I  studies with concurrent comparison group (nil, placebo, or active)

E  in vitro, in situ studies

E  studies without concurrent comparisons

(2)  "Caries active/high caries risk" (CA) designation

I  studies where CA status is designated at the level of the individual

I  studies where DS, DFS, DMFS, ds, dfs, or defs score is used for CA designation

I  studies where microbiological testing is the basis for CA designation

E  studies where CA is designated at a group, community, or population level

E  studies where CA is based solely on sociodemographic characteristics

(3)  Intervention

I  interventions requiring professional application and/or prescription

I  interventions likely to be undertaken only upon the recommendation of a dentist

E  fluoride dentifrice studies regardless of concentration, if only intervention/control component

(4)  Sample size

no exclusion criterion

(5)  Outcome

I  outcome expressed as change in DS, DFS, DFT, DMFS, DMFT, ds, dfs, dft, defs, or deft

D=decayed permanent tooth structure, F=filled permanent tooth structure, M=missing permanent tooth structure, S=permanent tooth surface, T=permanent tooth, d=decayed primary tooth structure, f=filled primary tooth structure, e=missing/indicated for extraction primary tooth structure, s=primary tooth surface, t=primary tooth.

A key criterion for the diagnostic question was the requirement for histologic validation of caries status for each surface studied. By requiring this level of validation, the team eliminated a large number of studies that compared two or more diagnostic methods, with one method designated as the reference standard. No currently available clinical diagnostic method is perfectly valid, i.e., has 100 percent sensitivity and specificity compared with histologic evaluation. Thus, the team excluded such studies because members were unwilling to include studies that would automatically introduce error into the assessment process. The team made an exception to the histologic reference standard where cavitation was the extent of lesions to be detected. Here a reference standard of direct visual clinical examination was deemed acceptable. Both in vivo and in vitro studies were accepted, although the number of in vivo studies was understandably limited because of the requirement for histologic validation.

The inclusion criteria for diagnostic studies also required that outcomes be expressed in terms of sensitivity and specificity. This criterion resulted in the exclusion of several studies where results were expressed only as receiver operating characteristic (ROC) curves. Such outcomes are typically obtained when observers indicate their level of certainty about a diagnosis on a five-point scale. The argument for using such an analytic approach is that asking the observer to state a level of certainty helps disassociate an observer's degree of leniency from the implications of any given decision criterion, thus permitting diagnostic performance to be reflected independent of an observer's perceived "cost" of an incorrect diagnosis. Many studies that employ ROC analyses also report sensitivity and specificity outcomes for the combined levels of "reasonably certain" and "certain" that a lesion is present. If these outcomes were reported, the study was included in the evidence table. However, when it was necessary to estimate values for sensitivity and specificity outcomes directly from ROC curves because no data were reported in text or table, a study was excluded.

Finally, the team had originally set arbitrary limitations for caries prevalence and sample size, but with no objective support for those specifications. Subsequently, the team found that 15 percent of studies would be ineligible because they either did not report caries prevalence in the sample or had a prevalence over the maximum of 80 percent. The team also found that the sample size exclusion criterion (less than 30) would exclude 10 percent of identified studies. In light of the limited number of studies available when other inclusion and exclusion criteria were applied, the team decided to drop both of these exclusion criteria and include the studies.

Two sets of management inclusion and exclusion criteria were applied to the single set of studies identified in the management literature. These two sets shared some common criteria. Only in vivo studies were included; in vitro studies and in situ studies, where exogenous tooth tissue was placed in the oral environment, were excluded. Studies without concurrent comparison groups (either nil, active, or placebo comparisons) were excluded. The team kept only studies with interventions requiring professional provision (e.g., application, prescription, etc.) or with interventions unlikely to be undertaken without the recommendation of the dentist (e.g., daily OTC mouthrinses, gums, etc.).

The key inclusion criterion for noncavitated lesion studies involved the type of lesion examined. The question addressed lesions for which there was some likelihood that remineralization treatment would be successful. To identify studies that included only this type of lesion, the team originally established "noncavitated" as a definition for acceptable lesions in conjunction with the TEAG. However, the term "noncavitated" was not in widespread use for much of the search period. Thus, the inclusion criterion were broadened to include the terms "incipient" and "initial" lesions. This decision permitted the inclusion of studies in which outcomes of interest had been reported in analyses stratified by caries status at baseline.

One other criterion for studies of noncavitated lesions deserves mention. The unit of analysis was the individual carious lesion. This is not the usual analytical unit in caries studies, which are typically analyzed and reported in terms of total decayed, missing, and filled (DMF) surfaces for a single subject. However, when the reversal of individual lesions is at issue, an aggregated analytical unit that combines cavitated and noncavitated lesions and cannot distinguish multiple lesions on a single surface is unusable.

For the review of studies involving caries-active individuals, the definition of subjects was the key inclusion criterion. As noted earlier, no consensus exists on characteristics of either caries-active or at-risk individuals. For this reason, and because the team realized that most of the extant studies would represent subgroup analyses of controlled trials rather than studies recruiting at-risk or caries-active subjects exclusively, the inclusion criteria were intentionally broad. The team accepted designation of at-risk and caries-active individuals through any combination of caries experience and/or mutans streptococci concentrations. The team did not specify cut points or limit the relative size of the risk group with that of the total sample. The team did insist on individual identification of subjects, thus excluding studies where schools or communities were selected on the basis of mean caries experience, socioeconomic status, or other group-level predictors of caries activity or caries risk.

Criteria for outcomes of studies eligible for inclusion allowed a range of traditional measures of caries experience in primary and permanent teeth. Again, studies were required to have concurrent comparison groups; but because some findings were expected to be subgroup analyses, minimum sample sizes were not established.

Data Extraction Forms and Reviewers

The scientific director and clinical director collaborated on the development of three data extraction forms for the diagnostic, noncavitated lesion, and caries-active individual questions. Draft versions of the forms, together with lists of evidence table columns linked to the extraction forms, were circulated to the TEAG for review and comment. Final versions of the forms incorporated TEAG comments and necessary modifications that had been identified through pretesting. Copies of the forms appear in Appendix D. Based on experience with previous abstraction forms developed for other RTI-UNC EPC projects, the team included essential directions and criteria on the forms and also endeavored to collect little information beyond that planned for inclusion in the evidence tables. All three forms included discrete sections addressing study design, subjects, examiners, caries criteria, intervention information, and results.

Data extraction for the two management questions was performed by the scientific director as the sole reviewer. In addition to extracting data, the reviewer also calculated the number needed-to-treat (NNT) statistic from the data reported in the study, where possible. The clinical director subsequently reviewed the evidence tables, confirming entries directly with the published papers. All disagreements were discussed and tables changed when indicated by the discussion.

Data extraction for each paper in the diagnosis review was completed by one of three reviewers (a pediatric dentistry resident, a dental epidemiology resident [a dentist], and a dental hygienist with a masters in education who specializes in dental radiology). The reviewers were not blinded to journal, author(s), or institution. The scientific director reviewed the completed extraction forms and subsequently verified evidence table entries against the published papers. The three reviewers participated in a training session that used six of the included studies, and all also completed an extraction form for the same study toward the end of review period.

Quality Control, Adjudication, and Reliability

Quality control mechanisms for determining eligibility for abstraction have already been described. All articles were reviewed by title by the scientific and clinical directors and disagreements settled by consensus. Agreement on retention status for this process was 96 percent for diagnosis articles and 97 percent for management articles. At the next stage (abstract/full paper review), all articles were again reviewed by both directors, with disagreement again resolved by consensus. An agreement rate was not calculated for this stage. Finally, the pool of potentially eligible articles was again subjected to a final review by both directors, with disagreement arising on one diagnostic article (2 percent) and one management article (3 percent).

Agreement among the three reviewers and the scientific director on descriptive data for the study abstracted by all reviewers was 100 percent for results (sensitivity and specificity for four different diagnostic methods) and 88 percent for study description items. Agreement statistics for confirmation of evidence table entries by direct comparison with articles were not recorded.

Quality Rating of Individual Articles

The team developed separate sets of quality rating items for the diagnostic and management articles. The sets of items are unique, but most individual items are either modified from or taken directly from existing rating scales used by the RTI-UNC EPC. In developing the quality rating item sets, the team was guided by the suggestions advanced by Lohr and Carey,³² both investigators in the RTI-UNC EPC. For the prevention questions, CONSORT criteria³³ figured prominently in the design; and for the diagnostic question, several items were included in response to issues examined in Lijmer, Mol, Heisterkamp, et al.³⁴

The rating scales assess several elements of internal validity, including study design, duration, sample size, blinding, baseline assessments of differences among groups, loss to followup, and examiner reliability. Two items also request the reviewer's subjective assessment of both internal and external validity of the study. The forms were pretested on small groups of studies, which resulted in some changes in wording.

Figure 2. Quality rating form -- caries diagnosis studies (more...)

Figure 2. Quality rating form -- caries diagnosis studies

Number of sites assessed:
3	150 or more
2	75-149
1	40-74
0	fewer than 40
Area assessed for any site:
1	Entire surface (occlusal, proximal, etc.)
0	Specific site on surface
Setting:
2	In vivo
0	In vitro
Tooth selection:
3	Both posterior and anterior teeth
2	Only posterior or only anterior teeth
1	Selected posterior or selected anterior teeth
0	Single tooth type (e.g., max. or mand. 3rd molar)
Validation method:
2	Light microscopy (stereo/mono) w/wo dye
1	Other visual or radiographic assessment of sectioned tooth
0	Assessment of unsectioned tooth
Validation criteria:
1	Criteria explicitly stated
0	Criteria not explicitly stated
Validation reliability:
1	Interevaluator or intraevaluator reliability reported
0	No validation reliability reported
Caries prevalence (calculate score for each lesion type evaluated):
2	Less than 20%
1	20-49%
0	50% or more
Number of test evaluators:
2	4 or more
1	2-3
0	1
Test reliability reported:
2	Interevaluator and intraevaluator reliability reported
1	Interevaluator reliability reported or intraevaluator reliability reported
0	No reliability reported
Criteria for caries call:
1	Specified prior to evaluation
0	Developed post hoc

Figure 3. Quality rating form -- caries management studies (more...)

Figure 3. Quality rating form -- caries management studies

Study type:
2	RCT
1	Other prospective design with control/comparison group
0	Uncontrolled or cross-sectional design
Duration:
3	5 years or more
2	2-4.9 years
1	1-1.9 years
0	Less than 1 year
Blinding:
2	Examiners and subjects
1	Examiners only
0	Subjects only or none
Baseline assessment of equality of treatment groups:
1	Reported and adjusted if necessary
0	Not reported or unadjusted when differences reported
Previous and concurrent caries prevention exposures (other than intervention) described:
1	Yes
0	No
Sample size:
3	50 or more in smallest analysis group
2	20-49 in smallest analysis group
1	10-19 in smallest analysis group
0	<10 in smallest analysis group
Loss to follow-up per year:
1	Fewer than 15%
0	15% or greater or unreported
Analysis:
1	Intention to treat
0	Includes only those remaining at final exam
Reliability:
2	Intrarater and interrater reliability reported and interrater reliability above 0.6 kappa, or 90%
1	Intrarater or interrater reliability only reported and above 0.6 kappa, or 90%
0	No rater reliability reported or reported level(s) below minimum
Active group (for Caries-Active Individual studies only):
1	Active group represents less than 50% of total sample/population from which it was selected
0	Active group proportion >50% or unknown
Probing (for Noncavitated lesion studies only):
1	Criteria for lesion not dependent on probing
0	Probing used in lesion criteria
Subjective assessment of external validity:
1	Reasonably wide generalization possible
0	Applicability limited to very specific populations
Subjective assessment of internal validity
1	Reasonably "tight" methods and design
0	One or more concerns re measurement, selection, etc.

Grading the Evidence

For the diagnostic question, the strength of the evidence was judged in terms of the extent to which it offered a clear, unambiguous assessment of the validity of a particular method for identifying a specific type of lesion on a specific type of surface. The three possible ratings were:

• Good (A). The number of studies is large, the quality of the studies is generally high, and the results of the studies represent narrow ranges of observed sensitivity and specificity.

• Fair (B). There are at least three studies, the quality of the studies is at least average, and the results represent moderate ranges of observed sensitivity and specificity.

• Poor (C). There are fewer than three studies, or the quality of the available studies is generally lower than average, and/or the results represent wide ranges of observed sensitivities and/or specificities.

For purposes of this question, a narrow range is defined as no more than 0.15 on a scale of 0.00 to 1.00, a moderate range is no more than 0.35, and a wide range is more than 0.35. High quality is defined as most study scores at or above 60; average quality is defined as most study scores at or above 45.

As in previous RTI-UNC EPC systematic reviews, we used a four-level grading scheme for judging the overall efficacy of each of the interventions reviewed in the two caries management reviews. The scheme was based on four aspects of the situation as depicted in the evidence tables, the number of studies, the magnitude of the effects reported, the quality rating scores of the studies, and the consistency of the evidence across studies. The scientific and clinical directors independently rated the evidence and developed an adjudicated final rating. The four possible ratings were:

• Good (A). Data are sufficient for evaluating efficacy. The sample size is substantial, the data are consistent, and the findings indicate that the intervention is clearly superior to the placebo/usual care alternative.

• Fair (B). Data are sufficient for evaluating efficacy. The sample size is adequate, but the data show some inconsistencies in outcomes between intervention and placebo/usual care groups such that efficacy is not clearly established.

• Poor (C). Data are sufficient for evaluating efficacy. The sample size is sufficient, but the data show that the intervention is no more efficacious than placebo or usual care.

• Insufficient Evidence (I). Data are insufficient for assessing the efficacy of the intervention, based on limited number of studies and/or poor methodology.

Cavitated Lesions on Proximal Surfaces

Table 12. Sensitivity and specificity values for the (more...)

Table 12. Sensitivity and specificity values for the detection of proximal cavitated lesions

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Radiographic -- in vivo
Rugg-Gunn, 197237								X													O
Downer, 1975109																X				O
Mejare, 198538								X													O
Pitts, 199263																		X			O
Hintze, 199836														X						O
Radiographic -- in vitro
Espelid, 1986110															X					O
Visual/Tactile -- in vivo
Mejare, 198538							X														O
Hintze, 199836								X													O
Visual/Tactile -- in vitro
Downer, 197535																				X	O
Visual -- in vitro
Downer, 197535																			O	X
FOTI -- in vivo
Hintze, 199836		X																			O

X = sensitivity, O = specificity

Among the assessments of radiographic techniques, two reported sensitivity levels substantially lower (0.34, 0.35) than the remainder, (0.63 to 0.87).^37,38 There were no obvious characteristics of these two studies that could explain the difference. Although one study³⁷ required lesion penetration only to two-thirds of the thickness of the enamel, the other employed a criterion essentially similar to the remaining studies, with lesion penetration into the dentin. Both assessments used D-speed film as did three of the four other assessments, and both were conducted in samples with low caries prevalence (5 and 9 percent), similar to three of the four other radiographic studies.

The variation in sensitivities displayed among studies of radiographic and visual/tactile methods and the availability of only single studies of visual and FOTI methods for the diagnosis of cavitated lesions on proximal surfaces of permanent posterior teeth results in a poor rating for the strength of the evidence for assessing the validity of these diagnostic methods. All of the assessments display high specificity, but the range of sensitivities is wide within methods. However, one pattern evident in these results is the tendency for sensitivity to be higher in in vitro than in in vivo assessments.

Lesions Penetrating into Dentin on Proximal Surfaces

Table 13. Sensitivity and specificity values for the (more...)

Table 13. Sensitivity and specificity values for the detection of proximal carious lesions penetrating into dentin

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Radiographic -- in vitro
Mileman, 199039												X								O
Verdonschot, 199141											X									O
Russell, 199342				X																O
Russell, 199342							X												O
Russell, 199342							X													O
Ricketts, 1997d40				X																	O
Visual/Tactile -- in vitro
Verdonschot, 199141											X				O

X = sensitivity

O = specificity

The radiographic assessments all showed high specificity values, ranging from 0.92 to 0.99, whereas sensitivity values were lower and more variable, ranging from 0.16 to 0.54. One of the two studies with the lowest sensitivity employed a direct digital radiographic technique.⁴² The single visual/tactile assessment⁴¹ showed lower specificity (0.71) and sensitivity equivalent to the better radiographic assessments (0.56). The diagnostic criteria for this visual/tactile study stressed the cavitation of the lesion in addition to penetration to dentin, which may have caused more frequent misidentification of enamel lesions as having penetrated dentin.

The strength of the evidence for assessing the validity of methods for detecting proximal lesions penetrating into dentin is rated as poor for both methods. Although the number of studies was too small for conclusions, the available studies suggested that the radiographic method features high specificity, but a range of sensitivity levels, from low values of 0.16 to 0.30 to moderate levels around 0.55. Thus, in any given assessment, little more than one-half of all lesions extending into dentin will be identified radiographically. The single visual/tactile study precludes any conclusions about this method for diagnosing proximal lesions penetrating into dentin.

Any Carious Lesions on Proximal Surfaces

Table 14. Sensitivity and specificity values for the (more...)

Table 14. Sensitivity and specificity values for the detection of any proximal carious lesions

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Radiographic -- in vitro
Heaven, 199243																					OX
Russell, 199342			X																O
Russell, 199342			X																O
Russell, 199342				X															O
Ricketts, 1997d40						X														O
Firestone, 199844																O	X
Firestone, 199844													X					O
Firestone, 199844																X	O

X = sensitivity

O = specificity

The specificities were generally high, although there was a greater range than was seen for diagnosis of cavitated lesions and lesions penetrating the dentin (74 to 100 percent). The assessment with perfect specificity also reported perfect sensitivity. This assessment, which employed image analysis software, evaluated only 16 surfaces, 75 percent of which were carious. Three other assessments reported moderately high sensitivity levels, all stemming from a single study that evaluated automatic image analysis software, D-speed film, and a digital image of the D-speed film read with the results of the automatic image analysis displayed.⁴⁴ Lesion prevalence in this study was also high at 66 percent. The remaining four assessments returned low sensitivity levels, 0.15 to 0.27. One of these assessments also employed digital imaging techniques.⁴²

Three of the studies reflected the same patterns seen in the performance of radiographs in detecting lesions extending into dentin. Two studies reported assessments with relatively low sensitivity and high specificity,^40,42 whereas one study reported assessments with higher sensitivity and reduced specificity.⁴⁴ Again, the variation among studies limited our ability to reach firm conclusions about the performance of radiographs in detecting proximal carious lesions of any extent. The strength of the evidence for assessing the validity of the method is rated as poor. The specificity of this method appears to be slightly more variable than when it is used to diagnose lesions into dentin or cavitated lesions. Sensitivity levels varied widely and were highest when computer-based image analysis was applied to digital images.³⁰

Cavitated Lesions on Occlusal Surfaces

Table 15. Sensitivity and specificity values for the (more...)

Table 15. Sensitivity and specificity values for the detection of occlusal cavitated lesions

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Visual/Tactile -- in vivo
Downer, 1975109																	O	X
Visual/Tactile -- in vitro
Downer, 197535																		O	X
Visual -- in vitro
Downer, 197535																	O		X
Ketley, 199345							X														O

X = sensitivity

O = specificity

Lesions Penetrating into Dentin on Occlusal Surfaces

Table 16. Sensitivity and specificity values for the (more...)

Table 16. Sensitivity and specificity values for the detection of occlusal carious lesions penetrating into dentin

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	8O	85	90	95	100
Radiographic -- in vitro
Wenzel, 199049														X						O
Wenzel, 199049															X						O
Wenzel, 199150														X				O
Wenzel, 199150															X			O
Wenzel, 199150													X					O
Wenzel, 199150															X			O
Wenzel, 199150														X			O
Wenzel, 199251											X						O
Wenzel, 199251															X			O
Wenzel, 199251												X				O
Nytun, 199248											O			X
Ketley, 199345														X					O
Russell, 199342					X															O
Russell, 199342					X															O
Russell, 199342					X															O
Lussi, 199355										X							O
Verdonschot, 199356													X				O
Lussi, 199547													X			O
Ricketts, 1994111													X			O
Huysmans, 1997112													X						O
Ekstrand, 199757												X									O
Ricketts, 1997d40				X																O
Ashley, 199852						X													O
Ashley, 199852					X														O
Huysmans, 1997112													X							O
Visual/Tactile -- in vitro
Penning, 199254						X															O
Lussi, 199355				X																O
Visual -- in vivo
Ricketts, 199546		X																		O
Visual -- in vitro
Nytun, 199248									O						X
Wenzel, 199251												X					O
Lussi, 199355			X																	O
Lussi, 199355					X														O
Verdonschot, 199357											X								O
Deery, 199563			X																	O
Ekstrand, 199757																			O	X
Ashley, 199852						X														O
Huysmans, 1997112						X															O
EC -- in vivo
Lussi, 199547																O				X
Ricketts, 199546												O								X
Verdonschot, 199356														X			O
Ricketts, 199546														O						X
Ricketts, 199546													O							X
Ricketts, 1997a58																			OX
Ricketts, 1997b59																		O	X
Ricketts, 1997c60																OX
Ekstrand, 199757																		O	X
Lussi, 199961																			X
Huysmans, 199853													X							O
Huysmans, 199853																X			O
Huysmans, 199853																	XO
Ashley, 199852																	XO
FOTI -- in vitro
Ashley, 199852				X																O
Laser Fluorescence -- in vitro
Lussi, 199961																X		O
Lussi, 199961																	O	X

X = sensitivity

O = specificity

The radiographic studies showed moderate variation in specificities, which ranged generally from 0.75 to 1.00. Sensitivities appeared in two distinct ranges of values with one value centered near 0.20 and another ranging from 0.45 to 0.70. Generally, the assessments with the lowest values for sensitivity had specificity values near the high end of the range of values. A single study reported a specificity of 0.50, with a sensitivity within the upper range at 0.66. This assessment was based on teeth selected because they showed signs of "fissure caries," and the prevalence of carious lesions penetrating into the dentin was 77 percent.⁴⁸ Ten of these assessments involved digital techniques.^42,49-53 Two assessments reported sensitivities in the lower range of values (0.21, 0.24)^42,52 and eight were in the higher range (0.54 to 0.72). Five studies reporting assessments of digital methods also included assessments based on direct examination of D- or E-speed films. In three instances, the performance of the film image was not markedly different than any of the digital comparisons reported in the same study.^42,49,52 In one study, one of the digital methods, edge enhancement of the digitized film image, yielded 0.23 and 0.04 improvement in sensitivity and specificity, respectively.⁵¹ This study featured a single examiner, with 1-week intervals separating diagnostic sessions where different methods were employed.

Two in vitro assessments of the visual/tactile method returned sensitivity values of 0.24, with specificities within the range established in the radiographic studies.^54,55 One of these assessments, which reported perfect specificity, reflected standardized (mechanical) probing of 1,140 sites on three teeth.⁵⁴ The criterion for a positive diagnosis was the force necessary to withdraw the probe from the surface of the tooth, or "tugback." The visual component of the examination was limited to directing the probe to all pits and fissures on the surface of the tooth.

Both in vivo and in vitro visual assessments were included in the review. The single in vivo assessment reported a sensitivity of 0.03 and a specificity of 0.97 for what was presumably a single examiner's diagnoses of specific sites on third molars.⁴⁶ The in vitro assessments generally reflected the same levels of performance as were seen in the radiographic assessments. Two groups of sensitivities were reported, more commonly between 0.10 and 0.25, with two studies between 0.45 and 0.55.^51,56 Two assessments reported higher sensitivity values. One assessment was based on the sample of teeth selected for signs of "fissure caries"⁴⁸ and the other used a histologic criterion that required lesions to penetrate more than one-third of the width of the dentin before dentin caries was declared.⁵⁷

The two in vivo EC assessments both showed moderate-to-high specificities (0.56, 0.77) and extremely high sensitivities (0.93, 0.97).^46,47 The in vitro EC assessments showed much more variation in both sensitivities and specificities. In some studies, the sensitivities were equal to or greater than the linked specificities. These studies were characterized by relatively modest sample sizes, all between 76 and 107 sites, and often a single examiner. Five of the 12 assessments were reported by the same principal author,^46,58-60 and four of these assessments were characterized by post hoc determination of the optimal values for caries criteria.

The single FOTI assessment displayed a sensitivity value of 0.14, at the lower end of the range of values in both radiographic and visual examinations, and a high specificity value of 0.95.⁵² Finally, two assessments of a laser fluorescence method utilizing newly available equipment also reported relatively high sensitivity levels compared with radiographic and visual methods (0.76, 0.84), with specificities in the lower half of the range of values for these other methods (0.87, 0.79). The assessments represented one report,⁶¹ with the system tested with the teeth dry and moist.

The general pattern of results suggested that radiographic and visual methods were roughly comparable in their ability to accurately indicate the presence of carious lesions penetrating into the dentin on occlusal surfaces. The two visual/tactile assessments were not markedly different than the larger number of in vitro visual assessments, and the single in vitro FOTI assessment returned a similar performance. EC and laser fluorescence methods showed better sensitivity than these other methods, with only a small penalty reflected in reduced specificities. However, the strength of the evidence for assessing the validity of each of the methods in any specific application is rated as poor because of the variation in sensitivity among studies and the small number of studies reported. The evidence presented by studies describing the EC method is the closest to meeting criteria for a fair rating and would do so if either of two studies were to be discounted. The pattern of sensitivity and specificity values for EC assessments presented a generally narrower range of variation than any other method. Further as noted, specificity values were typically lower than for other methods, and sensitivity values were typically higher, with the result that for 8 of 14 assessments, sensitivity was higher than specificity.

Any Carious Lesions on Occlusal Surfaces

Table 17. Sensitivity and specificity values for the (more...)

Table 17. Sensitivity and specificity values for the detection of any occlusal carious lesions

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Radiographic -- in vitro
Russell, 199342			X																	O
Russell, 199342			X																	O
Russell, 199342				X																O
Ricketts, 1997d40			X																	O
Wenzel, 199049																	X		O
Wenzel, 199049																X	O
Lazarchik, 1995113													X				O
Visual/Tactile -- in vitro
Penning, 199254				X																	O
Lussi, 1991114													X					O
Visual -- in vivo
Ricketts, 199546						X													O
Visual -- in vitro
Wenzel, 199049														O					X
Lussi, 1991114														X				O
Deery, 199562											O		X
EC -- in vivo
Ricketts, 199546																OX
EC -- in vitro
Rock, 1988115															X			O
Ricketts, 199546															X			O
Ricketts, 199546																OX
Ricketts, 1996116																			X		O
Ricketts, 1997a58															X			O
Ricketts, 1997b59													X							O
Ricketts, 1997c60													X					O

X = sensitivity

O = specificity

The radiographic assessments displayed the two ranges of sensitivity seen in summaries of radiographic assessments on other surfaces and for other types of lesions. The lower range of sensitivities was generally associated with higher specificities. This pattern of two ranges of sensitivity levels in the aggregated results of radiographic studies was most likely not associated with a single design feature of these assessments. Principal determinants of sensitivity included the specific criteria used for identifying lesions, the criteria for selection of sample teeth, the selection and training of the examiner(s), and the extent to which the study design protected against upward performance bias through familiarity with the sample. Assessments of digital radiographic techniques reported sensitivities at the higher level for digitized film⁴⁹ and the lower level for a direct digital method.⁴²

The visual/tactile assessments reflected the radiographic performance levels, with one assessment each in the higher and lower sensitivity ranges. The in vivo visual assessment fell in the same lower sensitivity range, whereas the in vitro visual assessments reflected the higher range. In two of these three assessments,^49,62 the sensitivity values were greater than the specificity values. In these instances, the extremely high prevalence of lesions in the samples, 89 and 97 percent, respectively, may have influenced the examiners' diagnostic decisions.

The EC assessments reflected a pattern similar to such assessments of lesions penetrating dentin, with the exception that the sensitivities tended to be slightly lower as a group. Six of the seven studies were reported by the same principal author; all featured one examiner; sample sizes were between 30 and 100 sites with the same sample used in two separate studies;^58,59 and relatively high lesion prevalence, 64 to 80 percent, was reported.

Overall, the performance of diagnostic methods for the detection of any carious lesions on occlusal surfaces was similar to the performance of these methods in detecting lesions penetrating dentin. Again, however, the strength of the evidence is rated as poor for all radiographic visual and visual-tactile methods, as well as for in vivo EC studies because of the variation in sensitivity and/or the number of available studies. The strength of the evidence is rated as poor for in vitro EC studies because of their low quality scores.

Enamel-Only Lesions on Occlusal Surfaces

Table 18. Sensitivity and specificity values for the (more...)

Table 18. Sensitivity and specificity values for the detection of occlusal carious lesions confined to enamel

	0	5	10	15	20	25	30	35	40	45	50	55	60	65	70	75	80	85	90	95	100
Radiographic -- in vitro
Wenzel, 199049										X					O
Wenzel, 199049							X								O
Ashley, 199852						X											O
Ashley, 199852					X												O
Visual -- in vitro
Wenzel, 199049														O	X
Ashley, 199852													X			O
EC -- in vitro
Ashley, 199852														X		O
FOTI -- in vitro
Ashley, 199852					X														O

X = sensitivity

O = specificity

Other Assessments

Thirteen other assessments of diagnostic methods were reported. Three assessments reported the results of methods on primary teeth. Radiographs were extremely sensitive (0.99) and highly specific (0.89) in detecting cavitated lesions on primary tooth proximal surfaces in vivo,⁶³ and slightly less so (sensitivity=0.93, specificity=0.89) in detecting lesions into dentin on occlusal surfaces in vitro.⁴⁵ Visual methods for detecting cavitated surfaces on occlusal services in vitro were less sensitive (0.45) but perfectly specific (1.00).⁴⁵ The strength of the evidence for assessing the validity of any diagnostic methods in primary teeth is rated as poor.

Seven assessments examined performance of radiographic methods in vitro on combined anterior and posterior proximal surfaces. Six of these assessments were reported in the same study⁶⁴ and featured high specificity for dentin lesions (0.94 to 0.95) that drop somewhat for enamel lesions and any lesions (0.76 to 0.80), and moderate sensitivities for dentin and any lesions (0.49 to 0.58) that drop somewhat for enamel-only lesions (0.35 to 0.46).

The other assessment of any lesions reported high specificity (0.95) and low sensitivity (0.33). The strength of the evidence is rated as poor.

Three assessments reported results for combined methods, all visual and radiographic examinations of posterior occlusal surfaces to detect carious lesions penetrating into dentin.^48,55,65 All of the assessments involved at least 10 examiners, and when reported, intra- and interexaminer reliability was low to moderate (kappa=0.30, 0.46). Performance of the combined methods varied, with moderately high sensitivities (0.49 to 0.86) and specificities in the lower half of the range reported for visual-only or radiographic-only methods (0.64 to 0.87). These results suggest that the combination of visual and radiographic methods returns moderate to high sensitivities, i.e., in the upper range of the dual range of sensitivity values described earlier, but that specificity may be reduced from that of either method alone. The strength of the evidence is rated as poor.

Limitations to the Evidence Base

The literature describing the histologically determined validity of methods for diagnosing carious lesions had a variety of limitations, many of which represented potentially serious threats to internal validity, and most of which represented barriers to generalization of the reported results to dental practice. The most obvious limitation has already been noted -- the virtual absence of any assessments of diagnostic methods applied to primary teeth and to root surfaces of permanent teeth. The breadth of reported studies also seriously restricted any conclusions about differences in the validities of visual and visual/tactile examinations and possible advantages of combining examination methods. A minority of method/surface/lesion type combinations were represented by more than three studies; and for most of these combinations, the variation among reported performances was extensive. Although the team might have meta-analyzed these results for several combinations of method, surface, and lesion type, two characteristics of the assessments discouraged the use of meta-analysis. First, many were not independent assessments, reflecting common examiners and sample teeth across studies. Second, the studies did not all assess the same "outcomes," since criteria for diagnosis were different. Thus, the available literature could not support specific conclusions about the performance of various diagnostic methods and permitted only the occasional generalization about relative differences in performance.

The quality scores for these studies tended to cluster in the low to mid range of the scale of possible scores. The mean for in vivo studies was 61, and for in vitro studies it was 45. The range was 5 to 70. Most studies had sample sizes of 75 or more sites or surfaces. The choice of sites rather than surfaces may be problematic for external validity, however, because most occlusal surfaces will present multiple sites for assessment. The results of site assessment did not summarize the status of the entire surface, as is routinely done in clinical practice. As noted, most of the studies were performed in vitro, a practical necessity if histologic validation is to be easily accomplished. There is some indication that in vitro assessments tended to return higher sensitivity values than in vivo assessments, although the numbers of studies were too small and the number of study variables too large to assume that the small observed differences were the results of the setting per se. Most in vivo studies were limited to premolars and third molars, which tend to be extracted more frequently in good clinical condition. In vitro studies also often relied on these teeth for the same reason -- that they are more frequently available with intact crowns. The problem is that the teeth that most frequently experience occlusal and proximal surface decay -- the first and second molars -- differ from the premolars and third molars in ways that may affect the performance of diagnostic methods. For example, occlusal surfaces of third molars tend to have less well-coalesced fissures; and proximal enamel thicknesses, both buccolingually and mesiodistally, are less in both premolars and third molars.

The inclusion criteria for the review required a histologic validation, and a variety of validation methods were represented in the included studies, with little assurance that different methods are equivalent.^66,67 Slightly fewer than one-half of the studies relied on light microscopy, with an identical number using other methods for evaluating the extent of caries on sectioned teeth. The remainder used visual criteria to confirm cavitated lesions. Further, a majority of studies supplied no explicit criteria for the validation, and a large majority did not report reliability information for the validation despite known variability in this procedure.⁶⁶ The result was that the reported diagnostic performances may be biased, although not in any predictable direction, by the validation procedures employed, making comparisons problematic.

The percent of sites to be evaluated that actually included a carious lesion was less than 20 percent in only 5 percent of in vitro assessments compared with 53 percent of in vivo assessments. Further, most of the in vivo sites with lesion prevalences above 20 percent were selected sites on third molars. Lesion prevalence proportions above 20 percent across all surfaces (a D value of 40 in a fully dentate individual) are rare in clinical practice. The effect of elevated frequencies of occurrence in assessment samples raised issues about examiner bias, since unusual presentations may alter examiner alertness and behavior, albeit in an unknown manner. The criteria used for selecting teeth for the samples also raised issues about both the comparability of studies and the generalization of results to clinical practice. The criteria described for selection did not always seem intended to reflect typical presentations in clinical practice. When this limitation was coupled with the previously noted limitations resulting from restricted tooth types and the use of sites rather than surfaces, generalization to clinical practice was again problematic.

Only one-half of the studies reported the combined performance of four or more evaluators. The studies relying on a smaller number of evaluators may present difficulties in generalization because of the positive influence of particularly skilled investigator/examiners. Single examiner studies, of which there were 17 (46 percent of studies), were especially vulnerable to this phenomenon. Further, 17 studies reported no reliability information for the examiner(s). When interexaminer reliability was reported, the values often were low enough to underscore the threat to external validity represented by the use of single examiners.

Finally, 14 of the assessments described post hoc determination of the optimal criteria for lesion designation. Although the development of new diagnostic techniques often requires such procedures, it is usually expected that the criteria will then be tested in a second "validation" sample. Such a procedure was not reported in any of the 14 assessments.

Fluoride Rinses, Solutions, and Varnishes

The results of evaluations of professionally applied fluoride therapies were equivocal. One study examined both acidulated phosphate fluoride (APF) and stannous fluoride (SnF) solutions applied one time to mesial surfaces of permanent first molars via access created during operative treatment of adjacent teeth.⁶⁹ Both solutions resulted in statistically significant reductions in progression, 38 percent and 18 percent, respectively, compared with placebo treatment. More than 50 percent of lesions progressed in 24 months in these experimental groups. The technique used depends on access to the mesial surface of the first molar through a cavity preparation in the second primary molar and hence has limited applicability in most clinical encounters. Because the lesions were detected and progression was assessed radiographically, it is likely that the sample was skewed toward lesions with more advanced demineralization, including clinically detectable cavitation. A second study examined lesions on facial surfaces of maxillary teeth.⁷² Rinsing with a 0.5 percent sodium fluoride (NaF) solution every 2 weeks during the school year resulted in 50 percent greater lesion progression and 22 percent less reversal in experimental group lesions compared with a placebo rinse. Approximately one-fourth of lesions in the experimental group progressed over 36 months. No statistical testing was reported.

In a third study, weekly application of a 2.0 percent NaF solution for 3 weeks, repeated after 3 months, resulted in an nonsignificant 8 percent decrease in occlusal lesion progression (progression rate of 35 percent over 20 to 32 months), compared with twice yearly rinsing with the same solution.⁷¹ As a field trial, this study employed relatively crude assessment techniques. School Dental Service personnel recorded lesions as a part of regular examinations, and progression was noted when these same personnel determined that the lesion required restorative treatment according to criteria in routine use in the clinics. The personnel were not blind to subject assignment to treatment group. A fourth study reported an analysis that examined the effects on noncavitated lesions of 5 percent fluoride varnish applied every 3 months to proximal surfaces of posterior teeth, compared with twice monthly rinsing with 0.2 percent NaF.⁷⁰ Progression was similar in the experimental and comparison groups (61 percent, 60 percent, respectively, over 36 months), and regression was identical (7 percent). No statistical testing was reported. Lesions were evaluated radiographically, again raising the likelihood that the sample of lesions studied was more advanced than those included in studies with visual criteria for lesion inclusion and progression.

Given the small number of studies and the equivocal nature of the results, it is clear that there is incomplete evidence for the efficacy of professionally applied fluoride treatments for stopping and reversing noncavitated carious lesions. Because each of the studies examines a different fluoride therapy on a different set of tooth surfaces, because assessment techniques are different in the four studies, and because two of the studies compared study effects in placebos whereas two used alternative fluoride therapies for comparison, it is not possible to make any comparisons among the studies. Thus, the efficacy of neither any specific fluoride therapy nor fluoride therapies in general for arresting and/or reversing noncavitated carious lesions enjoys adequate support from the literature.

Other Preventive Methods

Two other preventive methods were evaluated in a single study. Treatment with ammonical silver nitrate was compared with a placebo in a single direct application to mesial surfaces of first molars.⁶⁹ The 24-month progression rate of 69 percent was 16 percent less than the comparison, which was statistically significant. The application of dental sealants to occlusal and buccolingual pits and fissures was compared with no intervention in a subanalysis of a nonrandomized clinical trial.⁶⁸ Progression of 12 percent in the experimental group was 77 percent lower than in the comparison group, a statistically significant difference. Given that these two studies are the only reports of nonfluoride interventions meeting our inclusion criteria, the evidence is incomplete for the efficacy of methods other than fluoride therapies for stopping or reversing the progression of noncavitated carious lesions.

Limitations to the Evidence Base

The principal limitation to the evidence base is the extremely small number of studies assessing management of noncavitated lesions. Five comparative studies would represent only limited evidence even if most aspects of the study designs were similar or identical. When the subjects, tooth surfaces, lesion criteria, progression criteria, experimental intervention, and comparison group intervention all vary across studies, as was the case with the reviewed studies, it is impossible to reach any conclusions concerning the efficacy of any single method. When the results of any five studies display as much variation in efficacy determinations as the five included in this analysis, generalizations about the merits of the entire approach of stopping and reversing noncavitated lesions are also not possible. Although there is a large literature addressing the efficacy of fluorides and dental sealants in preventing carious lesions, the studies found in this literature generally lack two requirements for inclusion in the current review. First, they typically did not analyze outcomes for noncavitated lesions separately, and second, they almost invariably presented results aggregated to the tooth surface, tooth, or individual, but did not report lesion-specific results.

The most problematic aspect among the studies included in the review was the lack of standardized criteria for initially identifying noncavitated lesions and for assessing their progression. It is not at all clear that the lesions included in these five studies were equivalent in terms of their depth of penetration or their activity status and, for the lesions identified initially using radiographs, whether they were in fact noncavitated.

Although differences in the individual arrangements of the studies were not limitations of the individual studies per se, the extensive variation in these arrangements rendered comparisons difficult. For example, results based on placebo comparisons, a feature of the earlier fluoride studies as well as the sealant study, were difficult to compare with results based on active comparison groups. The "background" individual and community preventive dentistry exposures received by subjects during the course of the investigation, when reported, also varied across studies, again rendering comparisons difficult.

The quality scores for the individual studies were generally in the middle of the range of possible scores, with a mean of 53 and a range from 40 to 65. Most of the studies did not report intra- and, when appropriate, interexaminer reliabilities. All studies employed analyses that included only those subjects with final exams. The loss to followup was either over 15 percent per year or unreported in four studies. Three studies did not describe other preventive dental exposures, either individual or community based. External generalizability was rated as limited in four studies, and internal validity was a concern in three. Conversely, three of the five studies reported appropriate blinding and also indicated that baseline assessments of treatment group equality were performed. All but one study had a duration of 2 years or more, and all but one included 50 or more lesions in the smallest analysis group. None of the studies reported a power analysis, however, and two did not analyze the reported results statistically.

Fluoride Varnishes, Gels, and Rinses

Seven studies^73-79 describe nine fluoride-based interventions, in all but one of which the preventive effect was evaluated on all permanent teeth. The percent reduction in caries increment ranged from 7 to 30 percent in the eight interventions where caries increments were compared with increments in placebo or nonintervention groups. However, only two of these reductions were statistically significant,^74,78 both involving fluoride varnishes.

Five evaluations of fluoride varnish interventions were included in the review. One study⁷⁷ compared thrice annual applications of full strength (2.26 percent fluoride) Duraphat varnish with a half-strength formulation (1.1 percent fluoride), finding no difference in caries increments after 36 months. Because this study offered no information on the efficacy of fluoride varnish, it will not be considered further.

Two studies evaluated full-strength Duraphat varnish against nil comparison groups. Over 36 months, a statistically significant (p<0.001) 30 percent reduction in caries increment was found in one study of twice yearly application for 11- to 13-year-olds (n=124) with "above average DMFS values."⁷⁴ This reduction amounted to 0.61 surfaces per year. In the other study in 12- to 14-year-olds (n=92) with mS levels greater than 10⁶ CFU/ml, a nonsignificant 7 percent reduction was found for applications every 3 months over 24 months, 0.23 surfaces per year.⁷⁵ Two aspects of these studies may help explain the difference in observed efficacy. The former study reported that subjects received "ordinary dental care" whereas the latter indicated that subjects received twice monthly fluoride rinses in schools. Additionally, the baseline DMFS score for subjects in the former study (17.7) was substantially lower than the 24.3 and 22.9 experimental and comparison group DMFS scores, respectively, in the latter study. The difference may reflect the 1-year difference in age, but may also reflect the differences in inclusion criteria (caries activity versus caries risk status) or the examination criteria, which differed in whether early signs of carious lesions were included in the visual criteria.

Two studies evaluated Fluor-Protector fluoride varnish, which was reported as containing either 0.7 percent⁷⁴ or 0.1 percent fluoride.⁷⁸ In both studies, subjects were identified as being caries active on the basis of baseline DMFS or dfs scores. In the former study of 11- to 13-year-olds (n=140), twice annual applications resulted in a nonsignificant 11 percent reduction in caries incidence on all permanent tooth surfaces over 36 months, 0.19 surfaces per year. In the latter study, a similar application scheme involving 4- to 5-year-olds (n=303) resulted in a statistically significant 25 percent reduction over 24 months on proximal surfaces of anterior and posterior primary teeth, amounting to 0.29 surfaces per year.

Two of these four evaluations of fluoride varnishes used a split mouth design that might have biased the estimates of the efficacy of the experimental intervention.⁷⁴ The fluoride ions in the varnish vehicle applied to the teeth on one side of the mouth could have migrated to the teeth on the other side of the mouth and conferred some level of caries protection. Thus, these two studies may underestimate the efficacy of the experimental intervention.

Taken together, the four studies of two fluoride varnishes provide some support for efficacy. Two studies show demonstrated statistically significant effects,^74,78 and the remaining two studies report nonsignificant reduction results.^74,75 To consider clinical significance, it is useful to express the results in terms of NNT, although this statistic must be interpreted carefully because when applied to caries prevention, it refers to the numbers of persons who must receive treatment to prevent one tooth surface from becoming decayed over the period of 1 year. Also, the number needed to treat depends on the incidence of carious lesions in the population studied, becoming lower as incidence rises. The NNTs are 1.6 and 4.3 for the statistically significant and nonsignificant Duraphat studies, respectively, and 3.5 and 5.4 for the significant and nonsignificant Fluor Protector studies, respectively. We rate the evidence for efficacy of fluoride varnish for caries prevention in permanent teeth in individuals considered to be caries active or at elevated risk for carious lesions to be fair.

The remaining four studies of fluoride therapies all demonstrated small, nonsignificant reductions in caries incidence. Two types of topical application, ferric aluminum fluoride (FeAlF) solution applied four times per year⁷⁵ and APF gel applied biannually⁷⁶ were compared with nil and placebo intervention groups, respectively. In 12- to 14-year-olds (n=97), the FeAlF topical solution achieved a 13 percent reduction in caries incidence on all permanent teeth; and in 6-year-olds (n=431), the APF gel achieved a 9 percent reduction. These reductions represented NNTs of 2.5 and 6.7, respectively. Two fluoride rinse therapies, a twice daily 0.044 percent NaF rinse⁷³ and a biannual 1 percent amine fluoride (AmF) rinse,⁷⁹ were evaluated in comparisons with placebo groups. In all permanent teeth of 11- to 15-year-olds (n=72), the NaF rinse achieved a 15 percent reduction in caries incidence; and in 6-year-olds (n=91), the AmF rinse achieved a 24 percent reduction. The NNT for the AmF rinse was 10.2. A comparable NNT for the NaF study could not be calculated because the outcome measure used was decayed and filled teeth, rather than surfaces. Because the number of studies is small, the quality of the individual studies is moderate, and the results of these studies all show nonsignificant small protective effects of application, the evidence for the efficacy of topical fluoride applications and fluoride rinses in the prevention of caries in caries-active and at-risk individuals is rated as incomplete.

Chlorhexidine Varnishes, Gels, and Rinses

Seven interventions evaluated the efficacy of CHX therapies.^75,80-84 These interventions examined gels, rinses and varnishes, resulting in very few evaluations of any one type of intervention.

The three interventions that evaluated varnishes stemmed from two studies.^83,84 We could not determine if the varnishes used in these two studies had the same CHX concentration. Both studies provided results for eligible study samples through subanalyses of the trial subjects stratified by mS level and one through stratification by baseline DMFS/dmfs score as well.⁸⁴ Application was biannual in one 30-month study⁸⁴ and occurred three times over the first 8 months of a 24-month trial period in the other.⁸³ Varnish outcomes were evaluated in one study on occlusal surfaces of first or second molars of 5- to 6-year-olds and 11- to 12-year-olds (n=47, mS subjects; n=115, DMFS subjects).⁸⁴ In the second study where results were reported separately for occlusal surfaces of first or second molars of 7- to 8-year-olds and 12- to 13-year-olds (n=18), the results for the younger children could not be used because of the team's inability to reconcile sample numbers in the relevant tables. The results in this latter study, which were not tested statistically, showed that 25 percent fewer of the test occlusal surfaces than the comparison surfaces (one each per subject) experienced carious lesions. An NNT could not be calculated from the available data. In subjects included in the stratified analysis because of high mS levels in the other study,⁸⁴ a statistically significant 33 percent reduction in the incidence of caries was found (NNT=2.8 to prevent one lesion in four first or second molar occlusal surfaces). When subjects in this study were included in the analysis because of their previous caries history, the caries reduction was a statistically insignificant negative 9 percent. The evidence for the efficacy of CHX varnishes in individuals with high mS levels is rated as suggestive but incomplete. In individuals selected for their caries activity, the evidence is also incomplete.

Three studies reported evaluations of CHX gel.^75,80,81 These studies all used 1-percent concentration of gels, although the patterns of application differed among the studies, ranging from every 3 months to eight treatments in 2 days whenever mS levels were higher than 2.5x10⁵CFU/ml. All studies involved approximately the same age children, used the same caries criteria but slightly different examination methods for posterior teeth, included subjects in the analyses on the basis of mS levels, and reported exposures of subjects to community and individual dental preventive procedures. The reductions in caries in these three studies ranged from 26 to 52 percent, with NNTs ranging from 0.6 to 2.0. However, only one of the studies reported that the reduction was statistically significant.⁷⁵ Statistical testing was not reported for one result, a subgroup analysis,⁸¹ and the reduction was not significant in the third study involving 31 subjects.⁸⁰ The evidence of efficacy for CHX gels in individuals with high mS levels is rated as suggestive but incomplete. There are no studies of gels in individuals selected because of previous caries activity.

A single study evaluated the effect of a CHX rinse⁸² on all permanent teeth in 11-year-olds identified as having at least one new carious lesion in the past 2 years. Experiment subjects rinsed twice per day for 5 days and repeated the pattern every 3 weeks. Subjects in the comparison group received biannual NaF varnish treatments. A nonsignificant 3 percent reduction in caries experience was found, with the NNT being 27.5. The evidence for the efficacy of CHX rinses in caries-active individuals is incomplete.

Combination Treatments

Six studies examined the effects of combined treatments, either CHX and fluoride^82,85-88 or CHX and sealants.⁸⁹ This group of studies is quite varied in terms of the strategies employed in combining the preventive methods.

Three of the interventions were rinses, including a 1 percent CHX/NaF solution used daily in conjunction with a toothpaste with a similar composition,⁷³ and two 0.05 percent CHX/0.04 percent NaF solutions, one with and one without 500 ppm strontium, used twice per day.⁸² When evaluated against nil comparison groups but with a background of annual professional fluoride exposure, the first of these interventions resulted in a statistically significant 43 percent reduction in caries incidence, NNT=0.9. The other two interventions resulted in statistically nonsignificant reductions of 34 percent (NNT=2.1) and 8 percent (NNT=9.2) for the strontium-free and strontium-containing rinses, respectively.

A comparison of semiannual application of a combined 1 percent CHX, 1 percent NaF against NaF varnish alone on permanent posterior proximal surfaces resulted in a nonsignificant negative 26 percent reduction in caries incidence.⁸⁸ A combined 0.2 percent NaF/1 percent CHX rinse administered to mothers of 1-year-old children three times per day for 2 days twice per year resulted in a nonsignificant 13 percent reduction in dentinal caries incidence after 3 years⁸⁷ (NNT could not be calculated because of tooth-level outcome measure). One study examined effects of combined therapy in adults age 50 to 60.⁸⁶ Twenty-five subjects who met inclusion criteria involving salivary flow rates and Lactobacillius levels as well as mS levels received either no treatment or CHX gel treatments at unspecified intervals in addition to topical fluoride treatments and home fluoride rinses or gels, again at unspecified intervals.⁸⁶ After 1 year, a statistically significant 89 percent reduction in new decayed and filled surfaces was found, NNT=0.7. Finally, subjects with high mS levels included in a subgroup analysis of an RCT participated in daily CHX gel therapy for 14 days, repeated as necessary to maintain mS levels, and received occlusal sealant treatment for all unfilled occlusal fissures.⁸⁹ After 3 years, a statistically significant 81 percent in caries incidence was recorded, NNT=0.2, compared with a group receiving bimonthly NaF rinses.

These combination studies did not furnish sufficient evidence for determining the efficacy of any one approach. Although three CHX/NaF rinses all resulted in caries reductions, two of the reductions were statistically not significant. No conclusions can be based on the results of the remaining single evaluations of other approaches. The evidence for caries prevention through combined methods is rated as suggestive but incomplete.

Other Methods

The remaining six studies addressed five specific therapies, including alum,⁹⁰ chewing gum,^91,92 kanamycin,⁹³ and occlusal sealant,⁹⁴ and a sixth intervention designed to change dentists' provision of preventive therapy by informing them of individual subjects' mS scores.⁹⁵ All of these interventions resulted in reductions in caries incidence, although two of the reductions were small and not statistically significant (alum and dentist behavior), and two others were substantial, but not analyzed statistically (kanamycin and sealant). The two interventions that resulted in statistically significant reductions in caries incidence (one using a one-tailed test) involved the use of "sugar-free" chewing gum, in one instance containing xylitol⁹¹ and in the other containing sorbitol, manitol, and aspartame.⁹² In neither trial did the comparison group chew a placebo gum, so the possible effects of chewing gum on salivary flow and plaque disturbance could not be separated from effects of the sweeteners. The evidence for use of gum is suggestive, but for each of these methods the available evidence is rated as incomplete.

Special Populations

The team examined the results of the literature search to identify studies conducted in two types of subjects who are considered to be at high risk for dental caries: patients receiving orthodontic treatment who have bands and/or brackets placed on their teeth and patients who have received radiotherapy to the head and neck. Evidence Tables 5 and 6 describe the seven studies of orthodontic patients and six studies of radiotherapy patients that were included in the review.

The studies of orthodontic patients were of two general types: short-term studies on banded teeth extracted after 4 to 5 weeks to measure depth of demineralization and longer term studies evaluating the number of lesions or percent of sites that became demineralized per subject. Among these longer term studies, six interventions evaluated the effects of daily rinsing with APF⁹⁶ and NaF,⁹⁷ twice daily brushing with an SnF gel,⁹⁷ CHX rinses every 3 weeks with and without prophylaxis,⁹⁸ and periodic prophylaxis alone.⁹⁸ The APF rinse returned a statistically significant 53 percent reduction in the percentage of sites with detectable demineralization on two index teeth upon visual examination over a 20- to 28-month period compared with no treatment. Use of both the SnF and NaF rinses was associated with a statistically significant reduction in the number of initial lesions detected visually on all facial surfaces, 72 and 30 percent, respectively, when compared with subjects in a nil comparison group. Finally, the combination of four CHX rinses and a prophylaxis every 3 weeks resulted in a 95 percent reduction in the number of initial lesions detected visually on six index teeth compared with placebo rinsing. In this same study, prophylaxis alone every 3 weeks reduced initial lesions by 67 percent, and four CHX rinses alone every 3 weeks reduced initial lesions by 14 percent.

In short-term studies (4 weeks), titanium tetrafluoride (TiF) topical solution⁹⁹ and fluoride varnish¹⁰⁰ reduced the mean depth of demineralization on facial surfaces by a statistically significant 37 and 48 percent, respectively, compared with no treatment. Similarly, a combination of daily NaF rinsing and twice daily CHX rinsing yielded a statistically significant 69 percent reduction in mean demineralization depth when compared with daily NaF rinsing alone.¹⁰¹ Finally, two different approaches to cleaning the plaque accumulation under orthodontic bands, removing the band and cleaning the revealed surface either through pumice prophylaxis or wiping with a cotton pellet, prevented any visually detectable demineralization on the smooth surfaces of first molars compared with groups where no band removal and cleaning occurred, in which all surfaces had detectable lesions.¹⁰² As these short-term studies did not have carious lesions as outcomes, no conclusion about the effectiveness of these methods for caries prevention in individuals undergoing orthodontic treatment was warranted. The longer term studies suggested that a variety of interventions in this population may well be effective, but the evidence is clearly incomplete.

The interventions aimed at individuals who were receiving, or had recently received, radiotherapy generally tested daily regimens involving fluoride or CHX, or both, against alternative fluoride regimens. Sample sizes in most of these studies were small, with intervention group numbers usually fewer than 20. In the only comparison against a placebo, daily use of 1 percent NaF gel reduced the incidence of new lesions 95 percent over a mean of 20 months.¹⁰³ When sucrose restriction was added to the regimen, the reduction was 98 percent. Both reductions were statistically significant. Rinsing with a combination of NaF and calcium phosphate (CaP), first twice daily and then daily, together with NaF gel treatments reduced the incidence of new decay a nonsignificant 28 percent compared with the same regimen without the CaP over 12 months.¹⁰⁴ A regimen that began with APF gel for 4 weeks and then switched to NaF gel sweetened with xylitol resulted in a nonsignificant 22 percent increase in caries incidence in a 12-month study.¹⁰⁵ Daily brushing with SnF gel was no more effective than daily use of NaF gel for 3 months followed by twice daily NaF rinses in a 12-month study of coronal surfaces, but on root surfaces, the SnF regimen reduced the caries increment by a statistically significant 70 percent.¹⁰⁶ Fluoride made available by means of an intraoral slow release device attached to molar facial surfaces resulted in a nonsignificant 67 percent decrease in caries incidence compared to daily application of an NaF gel over 6 months.¹⁰⁷ In these latter two studies, which used NaF as comparison agents, comparison group incidence (0.06 and 0.03 DFS/mo, respectively) was much lower than in any of the other studies included in the review. Finally, a 6- to 10-month comparison of a combined CHX-NaF regimen (daily rinsing plus weekly topical applications for the first month) with daily NaF rinsing and APF topicals for the first month showed a significant 177 percent decrease in caries incidence (the experimental group DFS decreased).¹⁰⁸ In a third experimental group in this study, a nonsignificant 94 percent reduction was achieved by the CHX-NaF daily rinses alone, without the introductory gel treatment. The evidence for the efficacy of fluorides with or without CHX in the prevention of carious lesions among individuals receiving head and neck radiotherapy is ranked as fair. The evidence for other interventions is judged to be incomplete.

Harms

Only three studies contained any information regarding harms, in all instances commenting on side effects associated with the use of CHX. In one study,⁸⁷ 4 of 59 subjects dropped out complaining of "bad taste" or "a burning feeling on mucosal surfaces." In two others, staining was at issue. "Mild staining on a small proportion of subjects" was noted in one or negligible,⁷³ and "yellowish-brown staining of some dentitions…mostly very mild" in the other.⁸²

Limitations of the Evidence Base

The literature on the management of dental caries in individuals considered to be caries active or at risk for caries has several limitations. The principal shortcomings are the number of available studies for any given intervention, the variety of experimental protocols among any set of studies, the lack of studies including adult subjects and root surfaces, the meager number of studies examining effects on primary teeth, and several study design issues.

For any given management strategy, the number of available studies was small. Most preventive protocols were tested in general populations, with few investigations limiting their samples to individuals with elevated caries experience or with known risk factors for caries. The lack of a focus on such caries-active individuals is understandable because interest in "targeting" preventive procedures at the individual level has grown only within the past decade. Nevertheless, the number of available studies limited conclusions that can be drawn about the efficacy of specific preventive approaches and in some instances obviated them. For example, we found no studies describing the results associated with fluoride supplements.

Our ability to draw conclusions about any specific preventive approach was further limited by the variation in experimental protocols. For example, fluoride varnish was one of the few interventions for which we were able to rate the evidence at a level other than "incomplete." Yet in five fluoride varnish studies, three different concentrations of two different varnishes were evaluated using three different application frequencies. Comparison groups in these five studies received three different types of treatment, and all subjects received five different patterns of additional community and individual preventive procedures during the courses of the trials.

The literature focused almost exclusively on children, and in children, on permanent teeth. Only two studies of primary teeth were included in the literature and only one of adults. Efficacy of preventive approaches on root caries in caries-active individuals was examined only in one study of radiotherapy patients. The generalizability to adults of the outcomes in children is unknown, particularly in populations with relatively high rates of decay. Similarly, the extent of generalizability to primary teeth is unknown.

The variation in the methods for identification of caries-active subjects reflected the developmental nature of the literature on risk assessment methods for dental caries. Several approaches were advocated, all had some support from the literature on risk factors for dental caries, but none was validated. The two principal methods for selecting subjects were mutans streptococci levels and previous caries experience. When mS levels were used to identify subjects, the cutoff was generally 10⁶ CFU/ml, although both 10⁵ and 5 x 10⁵ cutoffs were also used. Criteria for cutoffs for previous caries experience were more variable, and depending on the age of the sample, included consideration of caries experience in both primary and permanent teeth. Presumably the cutoff was set in each study to capture individuals in the upper proportion of the distribution of disease (i.e., carious lesions and/or filled teeth and/or teeth missing as a result of caries). The mean cutoff for either admission to the study or for inclusion in subgroup analyses was the upper 35 percent of the study population or sample, but this value ranged widely, 7 to 87 percent, among the 18 studies reporting this information. For mS-based criteria, the mean cutoff was 33 percent (range 11 to 74 percent); and for previous caries activity, the mean cutoff was 38 percent (range 7 to 87 percent). One study⁸⁴ used both mS (>10⁶) and caries experience (DMFS/dmfs>0) to form two separate post hoc analytical groups. The proportions of the entire sample population included were 15 and 36 percent, respectively, and the outcomes of the analysis were markedly different, a significant 33 percent reduction in 30-month caries incidence and a 9 percent increase, respectively. This observation heightens our general concern about the comparability of outcomes in subjects selected using differing criteria for caries activity.

The quality scores for the studies included in the review were similar to those included in the noncavitated lesion review, with a wider range from 25 to 80 and a similar mean of 55. Quality scores reflected limitations in fewer than one-half of the studies with respect to subject blinding, small sample sizes, high or unreported loss to followup, and insufficient information or selectivity about identification of caries-active subjects. More frequent problems occurred in terms of analytical designs (only three studies used intention-to-treat designs), examiner reliability (in 17 studies incomplete or no information was provided), and the rater's subjective assessment of internal validity. Finally, a majority of studies did not report compliance estimates for subjects in the study, which was not a component of the quality score. Although in a few instances compliance could be assumed to be 100 percent because only "full participants" were included in the final analyses, compliance was not ascertainable in the remainder.

Management of Noncavitated Carious Lesions

A useful advance for research concerning noncavitated lesions would be the development and use of a standard set of valid criteria for their diagnosis and assessment of progression. The extant studies relied on either visual or radiographic methods depending on the location of the lesion, with different criteria employed within these methods. From this review of diagnostic methods, it would seem that on occlusal surfaces visual methods may offer better sensitivity, and this method could possibly be applied to proximal surfaces through tooth separation. However, to ensure generalizability to dental practice, studies should use criteria likely to be employed by clinical dentists, which suggests that radiographic criteria for proximal surfaces may be necessary. If so, the effects of nonsurgical management methods on their progression must be evaluated separately.

It is possible that developing technologies such as laser fluorescence may offer better diagnostic performance on all tooth surfaces. To be sure, when laser fluorescence methods become more widely available, it is likely that a great deal more "caries" will be diagnosed. Thus, research that determines the efficacy of nonsurgical strategies to treat lesions detected and assessed using this diagnostic method will be of paramount importance in any campaign intended to forestall a likely wave of surgical intervention that will accompany the adoption of the new imaging technology.

This review did not reveal any treatment approach that merits particular attention to the exclusion of others. Too few studies were available to be helpful in suggesting potentially fruitful or unrewarding areas of investigation.

Management of Caries-Active Individuals

Research needs to be directed toward techniques for predicting which individuals will develop new carious lesions in the absence of professional intervention. These techniques, variously know as "risk assessment" and "caries prediction" are at the heart of the question addressed in the review. To date, no set of "predictors" or "risk indicators" has been identified that offers documented satisfactory performance in identifying individuals who will experience new carious lesions within some future time interval. Nevertheless, several risk assessment instruments have been reported and are in use in a variety of settings. Basic work is needed to establish the validity of existing risk assessment instruments, as well as to identify more effective predictors.

The review suggests that well-designed, adequately powered trials of antimicrobial regimens and combined fluoride and antimicrobial regimens may be fruitful in managing caries-active individuals, but these strategies should not be pursued to the exclusion of other approaches such as gum or fluoride regimens. Finally, for individuals experiencing special high-risk conditions such as radiotherapy, work should continue to refine existing efficacious regimens.

TEAG Members

• Craig W. Amundson, D.D.S.

• Clinical Expert

• Health Partners

• Minneapolis, MN

• Kenneth J. Anusavice, D.D.S.

• Dental Researcher

• University of Florida

• Gainesville, FL

• (Representative of the American Dental Association)

• Brian A. Burt, D.D.S.

• Dental Researcher in Public Health

• University of Michigan

• Ann Arbor, MI

• John D.B. Featherstone, D.D.S.

• Clinical Expert

• University of California at San Francisco

• San Francisco, CA

• David Pendrys, D.D.S.

• Dental Researcher

• University of Connecticut

• Farmington, CT

• Nigel B. Pitts, D.D.S.

• Clinical Expert

• University of Dundee

• Scotland, UK

• Jane Weintraub, D.D.S.

• Dental Researcher in Public Health

• University of California at San Francisco

• San Francisco, CA

Peer Reviewer List

• Clinical Expert

• Steve Matteson, D.D.S.

• University of Texas at San Antonio

• San Antonio, TX

• Dental Professional Associations

• Burton Edelstein, D.D.S.*

• Representative

• American Dental Education Association

• Cynthia Hodge, D.D.S., M.P.H.

• Representative

• National Dental Association

• Charles Poland, III, D.D.S.*

• Representative

• American Academy of Pediatric Dentistry

• Patient, Consumer, and Public Health

• Irwin Mandel, D.D.S.

• Representative

• Consumers Union

• Rebecca King, D.D.S.

• Incoming President

• American Association of Public Health Dentists

• Evidence Report Users

• William J. Costello, D.D.S.

• Representative

• Dental Manufacturers of America

• Claude Padgett, D.D.S.

• Representative

• National Association of Dental Plans

• Richard John Hastreiter, D.D.S., M.P.H.

• Representative

• National Association of Dental Plans

• Thomas Gotowka, D.D.S.

• Representative

• Health Insurance Association of America