Successful application of risk-based prevention strategies for dental caries and periodontitis depends on two assumptions. The first is that patients who are at heightened risk will be identified, and the second is that once identified, these patients will receive appropriate preventive treatment to reduce the likelihood that disease will occur or progress.

Despite the attention now being paid to risk-based disease prevention and management, neither of these two key assumptions has received much scrutiny. While evidence has been available for some time that patients at elevated risk of disease incidence and progression can be identified with a reasonable degree of accuracy,⁹ there is little indication that dentists do this. Similarly, while it is clear that when administered, preventive and management therapies will be effective in reducing the incidence and progression of caries and periodontitis,¹⁰ it is not clear that patients who are at elevated risk of developing disease are actually receiving the needed preventive or management therapy.

Working with a national insurance carrier, we conducted a pilot study to refine study procedures for a full-scale demonstration study of risk-based prevention in private dental offices. The pilot study—an initial step in the carrier’s efforts to tailor its dental plans to the needs of patients at higher levels of risk—tested methods of communicating risk assessment from the practitioner to the carrier, and examined practitioners’ reasons for assigning risk levels and their proposed preventive treatment plans. We report the results of the pilot study in terms of what was learned about how dental practices approach risk-based prevention.

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
From its network of dental providers, the insurance carrier identified a small number of dental practices in four urban areas that submitted at least 20 claims per month, and invited them to participate in the pilot program. The dentists and staff members of offices expressing interest were invited to attend evening meetings in the spring of 2001. At these educational meetings, we reviewed the concepts underlying risk-based prevention and explained the requirements and expectations for participation in the pilot study. Offices wishing to participate were then enrolled, with the study starting at the beginning of June 2001, and continuing for six months. The study concluded with debriefing videoconferences during which we reviewed and discussed the results with the participants.

Study participation entailed three elements:

– assessing the risk of new carious lesions developing and the initiation or progression of periodontitis;

– indicating the reason or reasons why a patient was assessed as being at elevated risk of developing either or both diseases;

– for patients assessed as being at elevated risk, indicating any treatments planned in addition to a patient’s usual care regimen to address the elevated risk.

The dental practice then communicated to the carrier the assessment and, if applicable, reasons (risk indicators) and planned treatment by means of a set of codes placed in the comment field of the claim form. To facilitate recording and intrapractice communication of the risk-related information, we provided dental offices with pads of preprinted stick-on notes that permitted efficient identification of risk levels, indicators of elevated risk and planned preventive treatments (Figure). For dental caries and periodontitis, the clinician assessing risk could indicate one of three risk levels (low, moderate, high). For patients assessed as being at moderate or high risk, clinicians were requested to indicate from a list of disease-specific risk indicators up to three reasons why risk was elevated. We also asked them to indicate from a list of disease-specific preventive treatments up to three that they planned for patients.

View larger version (61K): [in this window] [in a new window]   Figure. Stick-on note for recording and communicating risk information.

Analyses of the risk and claims data were exclusively descriptive (that is, no inferential statistics were used) and based on combined data from participating offices. We calculated basic distributions for risk levels, reasons for elevated risk and planned treatments because this information, which had not been reported previously, is important for designing risk-based dental care plans. To begin to address the two key assumptions underlying risk-based prevention, we examined past patterns of restorative and periodontal care for patients at low and elevated risk levels to determine if assigned risk was reflected by treatment history. In addition, we examined the extent to which planned treatments addressed the reasons for elevated risk.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Twenty-three dental offices agreed to participate in the pilot project. Because of transmission and reception problems with the coded risk-related information included in some electronic claims, data from only 15 of these offices were available for analysis. The data represented risk-related information for 813 adult patients (ranging from four to 147 per office) examined during the six-month period.

Caries risk and planned treatments. The proportion of patients assessed as being at high risk of developing dental caries was 4 percent (34 of 813 patients). The proportion ranged from 0 to 18 percent across offices, which was quite stable. The interquartile range among offices (the middle 50 percent) was from 1 to 6 percent. The proportion of patients assessed as being at moderate risk of developing dental caries was 29 percent. In contrast to the proportion for the high-risk patients, this proportion varied widely across offices, with a range from 7 to 88 percent and a broad interquartile range from 22 to 54 percent. Participating dental offices assessed the remaining two-thirds of patients as being at low risk of developing dental caries and, again, the range across offices was broad.

Table 1 shows the proportions of high- and moderate-risk patients who were assigned the listed indicators of elevated caries risk. Clinicians assigned multiple risk indicators to most patients, with the most frequently cited being "multiple restorations," with "multiple carious lesions" assigned nearly as frequently in high-risk patients.

Assignment rates for these risk indicators varied broadly between offices, but the variation was due, in part, to unstable estimates resulting from the small numbers of patients in the medium- and high-risk groups in most offices. For this reason, we chose not to conduct analyses of interoffice variation for any assignment rates according to risk group. The next most frequent risk indicator was "other," and discussions during the debriefing sessions suggested that clinicians made this assignment primarily for white spot lesions, poor dietary practices, irregular visit patterns, chemotherapy or concern about marginal integrity.

Table 2 shows the proportions of high- and moderate-risk patients for whom specific caries-preventive treatments were planned. The relative frequencies were similar for both types of patients, with "oral hygiene and/or dietary counseling" planned most frequently, followed by "topical fluoride" application and "more frequent prophylaxis and/or recall" appointments. Clinicians used the "other" category to note specific oral hygiene interventions, such as use of an electric toothbrush, more frequent brushing or use of disclosing tablets. Patient-based preventive treatments such as use of prescription fluoride dentifrice or mouthrinse, prescription antimicrobial mouthrinse, over-the-counter fluoride rinse or calcium phosphate dentifrice were planned for small percentages of patients.

Table 3 shows the proportions of high- and moderate-risk patients who were assigned the listed indicators of elevated periodontal risk. Two indicators predominated among both groups of patients: "pocket depth" and "bleeding upon probing," with each assigned to just under two-thirds of high-risk patients and roughly between one-third and one-half of moderate-risk patients. Clinicians assigned more risk indicators, on average, to high-risk patients than they did to moderate-risk patients. Smoking was noted as a risk indicator for both high- and moderate-risk patients, as was "increasing pocket depth," while diabetes was noted only among moderate-risk patients. Clinicians indicated that they used the "other" category to note risk indicators such as family history, specific problems in plaque retention, histories of noncompliance with maintenance therapy, and previous periodontal surgery in the presence of current disease.

Evidence for key assumptions. The first key assumption is that clinicians will assess risk, and that their assessments will be reasonably accurate. The collected data support the first part of this assumption, and we tested the second part by examining previous restorative and periodontal treatment patterns for patients at low and elevated risk. The mean number of restorative procedures provided to patients at elevated (high and moderate combined) and low risk of developing caries in the three-year period before the pilot study was 1.62 and 1.04, respectively. The mean number of periodontal treatment procedures (not including prophylaxis) provided to patients at elevated and low risk of developing periodontal disease was 0.67 and 0.01, respectively. Both mean numbers were significantly higher for elevated-risk patients (unpaired t test = 2.80 for caries, 5.55 for periodontal disease; P = .006 for caries, P < .0001 for periodontal disease).

Our second key assumption is that patients at risk of developing disease will receive appropriate preventive therapy to reduce the risk. We examined the evidence for this assumption by comparing specific indicators of elevated risk with associated planned interventions. More specifically, we calculated the proportion of patients assessed with multiple carious lesions for whom clinicians planned fluoride treatment of any type, and the proportion of patients assessed with poor oral hygiene as either a caries or periodontal risk indicator for whom clinicians planned enhanced oral hygiene instruction. The proportions of patients were 51 and 73 percent, respectively.

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The data describing patient populations assessed as being at high and moderate risk of developing caries and periodontitis represent the first report of such distributions in private practices. The combined proportion of moderate- and high-risk patents is somewhat greater than the proportions reported to be at elevated risk of developing caries in three large closed-panel group practices in which dentists did assess caries and periodontal risk among their adult patients.¹¹

In the present study, clinicians classified 33 percent of adult patients as being at elevated risk of developing caries, compared with 17 percent in the group practices.¹¹ For periodontal risk, the proportions were more alike, with 37 percent of patients classified as being at elevated risk in our study compared with 31 percent in the study involving group practices.¹¹ Some external validation is available for the periodontal risk proportion in that the prevalences of moderate and severe disease have been estimated at 28 and 8 percent, respectively.¹² No such disease estimates exist for patients at high risk of developing caries, perhaps owing to the variety of risk indicators used to determine caries risk.

The interoffice variation in the proportion of patients assessed as being at high and medium risk of developing disease was similar for caries risk and periodontal risk assessments. All dental offices identified small proportions of patients who were at high risk of developing disease. Evidently, the risk indicators that distinguish high-risk patients are interpreted similarly across dental offices, despite the absence of any definitive criteria provided by the project to the participants. The same cannot be said for making determinations between medium and low risk of developing dental caries or periodontitis. The interoffice variation in the proportion of patients assigned to these categories was extensive, although somewhat less so for periodontitis than for dental caries. This variation may reflect both the lack of specific criteria for determining moderate risk and a general lack of clinical experience in working with trichotomous as opposed to dichotomous classification systems.

Possible anomalies. Overall, the frequencies with which the risk indicators were cited as the basis of elevated risk of developing caries and periodontitis seem to be appropriate. However, a few possible anomalies merit mention. The indicators of caries risk listed on the reporting form were compiled from several reported risk assessment systems.^1,3,4,13,14 Multiple restorations, the most frequently cited indicator of both high and moderate caries risk in this study, is not a component of all of these risk assessment systems. While restorations are an indicator of past caries activity, their presence may not signal current elevated risk. Some patients classified as being at moderate caries risk were reported as having multiple carious lesions. Multiple carious lesions probably should be considered an indication of high caries risk, with the possible exception of arrested lesions.^1,4

Similarly, some patients who smoke or have diabetes were classified as being at moderate risk of initiation or progression of periodontal disease. Because both smoking and diabetes are true risk factors for periodontal disease, these patients should be considered at high risk.^7,15

The preventive treatments that clinicians reported they planned to administer to patients classified at elevated risk are almost entirely appropriate, but they may be incomplete. The most notable example of a preventive treatment that was not always indicated by the clinician was exposure to fluoride for patients at risk of developing caries. Although the various fluoride interventions in Table 2 seem to indicate that 83 percent of high-risk patients and 46 percent of moderate-risk patients were scheduled to receive fluoride, multiple treatments for some patients reduce these proportions to 63 and 32 percent, respectively.

Overall, for those patients designated as having multiple carious lesions regardless of the assigned risk level, just 51 percent were to receive any fluoride intervention. This possible underutilization of an effective preventive agent in adults at risk has been noted previously,¹⁶ and may reflect clinicians’ reluctance to recommend preventive treatments not covered by patients’ dental insurance plans. Similarly, intensified oral hygiene interventions may be underutilized when poor oral hygiene is a risk indicator (although for dental caries, the evidence regarding the effectiveness of this approach to prevention is weak).

We should note that in a few instances, the caries-active patients who were already following enhanced preventive regimens may not have been designated as recipients of planned enhanced treatment, because of differences in the interpretation of instructions to clinicians. Thus, the figures we report may be slight underestimates of treatment actually planned for moderate- and high-risk patients.

Assumptions underlying risk-based prevention. Some support for the two assumptions underlying risk-based prevention is evident in the results of this pilot study. The volunteer dental offices did perform risk assessments and provide associated reasons and preventive treatment plans for eligible patients. Although technical problems in communicating this information to the insurance carrier interfered with data collection, exit interviews with participants indicated that they submitted the risk-related information for a large majority of eligible patients.

With respect to the accuracy of these risk assessments, patients at elevated risk of developing caries had received approximately 56 percent more restorations in previous years than did patients not at elevated risk. Moreover, patients at elevated risk of developing periodontitis were virtually the only patients in the participating offices to have received periodontal therapy. Although this is relatively weak evidence for classification accuracy, it is, in fact, the only assessment possible, because prospective analyses of treatment receipt will be confounded by administration of enhanced preventive therapy. Provision of that therapy is the second assumption upon which risk-based prevention depends. The results of the pilot study suggest that while appropriate heightened preventive therapy clearly is planned by participating clinicians, the completeness of the planned treatment might be questioned.

	CONCLUSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The results of this study are encouraging from the perspective of promoting risk-based prevention. We believe that application of the concept in private dental practices appears feasible. The volunteer clinicians reported generally appropriate reasons for assigning patients to elevated risk categories, and these patients had received more treatment in the immediate past than did other patients, supporting the assumption that they are at elevated risk of developing caries or periodontitis.

Further, the volunteer clinicians reported generally appropriate approaches to enhancing the preventive and maintenance treatment planned for patients at elevated risk. The results do suggest, however, that attention needs be paid to clarifying criteria used to define the classification of moderate disease risk, especially for caries. In addition, clinicians should consider the inclusion of more intensive exposure to fluorides in preventive treatment for patients at elevated risk of developing caries. Clearly, the results suggest that clinicians are quite capable of providing the clinical information that insurers would need to administer benefit plans tailored to patients’ risk levels.