Dentistry is witnessing a major paradigm shift in the measurement of oral health status.^1,2 The essential feature of this shift has been from the perspective of dentists to that of patients. Clinically derived indexes of disease that were based on the judgment of the dentist were the earliest oral health measures. Over time, these have been supplemented with measures that often are subjective and patient-centered.³ Self-reported oral health status is likely to focus on areas important to patients such as discomfort, ability to function and impact on socializing. Subjective health status measures are important in understanding how patients relate to the health care system and may be valid predictors of health behavior.⁴ While dentistry’s interest in the area of self-reported oral health research has been within the last decade,⁵ it is difficult to think of an area more crucial to determine outcomes of dental care than evidence that the outcomes are favorable to and valued by patients. For many aspects of dental care, patients are the only people competent to make the evaluation. The patients’ values should play an important part in selecting the appropriate care.⁶

A number of studies have examined the congruence between self-reported and clinically determined oral health. Studies that compared patient-reported oral conditions and clinical findings have had mixed results. Generally, patients are less likely to adequately assess their periodontal status and the presence of caries than they are to adequately assess their number of teeth, their number of restorations and the presence of fixed and removable prosthetics, which tend to be more consistent with the number recorded via clinical assessment.^7–10 Other studies have examined the relationship between self-reported oral health status and clinical findings and have concluded that patients’ subjective reactions to their oral conditions strongly influence their perceived oral health. Jones and colleagues¹¹ found that self-reported oral health status was better in people with more teeth and recent dental treatment and worse in those with tooth mobility, coronal caries and medical problems. They concluded that the clinical conditions and the impact of oral health on daily life are important determinants of self-reported oral health. Another study found that the discrepancy between dentists’ rating of oral health status and the patients’ self-assessed rating can be attributed to differences in the clinical and subjective factors that influence dentists’ and patients’ ratings.¹² These findings indicate that though patients have difficulty assessing their clinical status, their perceptions play an important role in the evaluation of the outcomes of dental care and understanding their health behaviors.

Economic markets for dental benefits vary considerably throughout regions in the United States.

In this article, we examine self-reported oral health status in relation to the characteristics of enrollees, dental markets, plan types, lengths of time in plans and out-of-pocket costs. We used responses to the question, "In general, would you say your overall dental health is: ‘excellent,’ ‘very good,’ ‘good,’ ‘fair,’ ‘poor’ or ‘very poor’?" to measure self-reported oral health status.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Economic markets for dental benefits vary considerably throughout regions in the United States; therefore, it was important for us to characterize the markets according to their amount of involvement with managed care before selecting the sample of plans to study. We developed an index to specify the degree of penetration of dental health maintenance organizations, or DHMOs, into the dental market,¹³ and identified four states that would ensure different levels of HMO penetration. The markets we selected were California (19.8 percent DHMO penetration), New Jersey (7.3 percent DHMO penetration), Michigan (4.6 percent DHMO penetration) and North Carolina (0.07 percent DHMO penetration).

We sought to identify companies that were in at least one market and offered more than one type of dental plan. We contacted 109 Fortune 500 companies whose operations appeared to include at least one of the four markets we selected. From these 109 companies, we recruited eight that agreed to participate. Four of these had dental plans in one market, and four offered plans in all four markets. The dental benefit plans of these eight companies included indemnity service or fee-for-service, or FFS, plans; preferred provider organization, or PPO, plans; and capitation, or CAP, plans. We classified plans on the basis of the method of payment to the dentist and who bore the financial risk. For the purpose of this article, we considered the one PPO plan to be an FFS plan, as the method of payment is on an FFS basis and the financial risk is borne by the plan. Thus, we categorized the plans as FFS or CAP.

The sample generated 42 different plans consisting of various combinations of company, market and plan types. We used the dual-party premium (an enrollee plus one dependent) as a measure of the premium cost of the benefit. There was considerable range in the premiums by market and by plan type as shown in Table 1. In both California and North Carolina, the lowest premium for FFS plans was higher than the highest premium for any CAP plan. In both New Jersey and Michigan, there were FFS plans with a lower premium than the highest CAP-plan premium.

The interview questionnaire had five major sections that addressed dental plans, use of dental plans, providers of care, enrollees’ oral health and enrollees’ demographic characteristics.

We contacted subjects by telephone until we either received a refusal or could not make contact after approximately 10 calls. The person we interviewed on the telephone was the primary person in the plan. The response rate was higher in the CAP-plan group (66 percent) than in the FFS-plan group (54 percent). The largest range in the response rate between CAP plans (63 percent) and FFS plans (48 percent) was in the New Jersey market. The highest response rate was for CAP plans in North Carolina (80 percent), and the lowest response rates were CAP plans in Michigan and FFS plans in New Jersey (both 48 percent).

We compared the demographic characteristics of the respondents, markets, plan types, premiums, enrollees’ length of time in plan and out-of-pocket costs with the enrollees’ self-reported oral health status. For categorical data (enrollees’ characteristics, markets, plan types and out-of-pocket costs), we conducted bivariate analyses using ² tests. For continuous data (dual-party premium), we conducted an analysis of variance. We used dual-party premiums to be able to make comparisons for bivariate analysis. We set the significance level at P < .05.

To determine the characteristics that are most important in self-reported oral health status, we conducted a multivariate analysis (multinomial logistic regression for categorical dependent variables) on all respondents in the sample. This type of analysis has the advantage of taking into account all of the characteristics and their interrelationships. For example, the bivariate analysis found that nonwhites were more likely to enroll in CAP plans than whites. The multivariate analysis took this association into account and allowed for the determination of the effects while controlling for this relationship. We treated the dependent variable—self-reported oral health status—as an ordinal outcome variable and used multinomial regression analysis. Since both education and income were significantly correlated (P < .001), we used only income in the analysis. We used coverage-specific premiums for multivariate analysis rather than dual-party premiums. Coverage-specific premiums use each enrollee’s coverage, whether it is individual, dual-party or family premium. The significance level was set at P < .05.

To examine the relationship between out-of-pocket costs and self-reported oral health, we conducted a second multivariate analysis (multinomial logistic regression for categorical dependent variables) on only the respondents who reported using dental care. This second model excluded nonusers because, though they would have had no out-of-pocket costs, they are different from those who use dental service with no out-of-pocket costs. Nonusers are likely to have different attitudes and beliefs about dental care than users with no out-of-pocket costs. Some may believe that they are in good oral health and do not require a dental visit, and others may not use dental care because they are dental phobic despite having dental care needs. Those who use dental services with no out-of-pocket costs are likely to believe in regular checkups and prevention and report better oral health status.

Nonusers are likely to have different attitudes and beliefs about dental care than users with no out-of-pocket costs.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Most of the 2,340 respondents rated their overall oral health positively; 20 percent rated it as excellent, 37 percent rated it as very good, and 31 percent rated it as good. Only 12 percent of respondents rated their oral health negatively as fair, poor or very poor (Figure). Due to the low number of respondents who rated their oral health negatively, we collapsed these three categories into one category for our analysis.

View larger version (20K): [in this window] [in a new window]   Figure. Distribution of self-assessed oral health status.

The distribution of self-reported oral health status of nonusers and users of dental services with varying levels of out-of-pocket expenditure varied significantly (P < .001). A total of 23.9 percent of nonusers reported fair, poor or very poor oral health, while between 7.6 and 14.8 percent of users with different out-of-pocket expenditures reported fair, poor or very poor oral health. In regard to users of dental services, those with higher out-of-pocket costs more frequently reported fair, poor or very poor oral health than those with lower out-of-pocket costs. For users with expenditures of $1 to $50 and $51 to $150, 7.6 percent and 9.1 percent reported fair, poor or very poor oral health, respectively. In comparison, users with expenditures of $151 to $350 and those with more than $350, 13.6 percent and 14.8 percent reported having fair, poor or very poor oral health, respectively. Other variables such as sex, having recently enrolled in the plan (a year or less) and premium were not significant.

The significant results of the multivariate analysis comparing self-reported oral health status with the different independent variables for all respondents are presented in Table 4. The multivariate analysis showed that race/ethnicity was a significant factor in the rating of oral health status even when taking the other variables into account. The odds ratios indicate that nonwhites were less likely to report good, very good or excellent oral health than fair, poor or very poor compared with their white counterparts. Plan type was also a significant factor in self-reported oral health status. When comparing plan type, the odds ratios indicate that respondents who were enrolled in CAP plans were less likely to rate their oral health as good, very good or excellent compared with those enrolled in FFS plans.

Users with higher out-of-pocket cost are less likely to report very good or excellent oral health compared with fair, poor or very poor oral health.

Table 5 presents the multivariate analysis results that compare self-reported oral health status with the various independent variables for only those respondents who use dental services. The results from this analysis model were similar to our analysis of all respondents, and the results indicate that race/ethnicity, plan type and income are significant factors in self-reported oral health status. Again, the odd ratios show that nonwhites were less likely than whites to report good, very good and excellent oral health compared with fair, poor or very poor. Also, those enrolled in CAP plans were less likely than FFS-plan enrollees to report good, very good and excellent oral health compared with fair, poor or very poor oral health. In regard to income level, again there were no significant findings for the comparisons of good and very good oral health versus fair, poor or very poor oral health. However, for the excellent versus fair, poor or very poor comparison, there were significant findings showing that those in income ranges earning less than $100,000 were less likely to report excellent oral health.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
This study’s finding that nonwhites were more likely to rate their oral health poorly compared with whites reinforces the well-documented problem of racial/ethnic disparities in oral health. The surgeon general’s report states that one approach to reducing these disparities is to increase access by making dental benefit plans available to more Americans.¹⁵ However, this study found that despite having dental insurance and income—key enabling factors in access to dental care—nonwhites continue to perceive to their oral health more poorly than whites did. This highlights the need to not only remove the economic barriers to access to care but to address other noneconomic barriers that nonwhites may be facing such as language, diet, cultural beliefs and health behaviors. It is important to tailor oral health promotion and services to specific racial/ethnic groups because nonwhites’ health beliefs and behaviors are likely to be different from those of whites.

The result that CAP-plan enrollees rated their oral health status poorer than FFS-plan enrollees raises the question: Was the CAP-plan enrollees’ perceived oral health status poor when they selected the plan? A limitation of our study is that it is cross-sectional such that the data reflect only one point in time. We did not measure the respondents’ perceived oral health status before enrollment in the plan but rather when the people already were enrolled in the plan. Therefore, we know only how they perceived their oral health at the time they were interviewed.

Assuming that CAP-plan enrollees’ perceived oral health was worse than FFS-plan enrollees’ when they selected their plan raises the issue of the relationship between plan selection and perceived oral health status. Does adverse selection occur in which people who perceive their oral health as poor select dental CAP plans? There is a lack of literature regarding bias selection of dental plans; however, a number of studies of medical plans have shown favorable selection of medical HMOs. Adverse selection of dental CAP plans would be contrary to the findings from medicine. In medical plans, people who are in better health tend to enroll in CAP plans, while those in poorer health tend to be enrolled in FFS plans.^16–18 A major consideration in the selection between CAP and FFS medical plans is the anticipated costs related to health status. FFS medical plans typically require significantly higher monthly premium costs and have considerably higher out-of-pocket costs for specialty care compared with CAP plans. Enrollees in poorer general health are willing to pay more for FFS plans that provide more access to specialists and maintain their relationship with their current providers.¹⁹ Those in good general health do not anticipate using expensive care and are likely to select a CAP plan to minimize their costs. The selection of dental plan type may have a different relationship to oral health status. People in poorer oral health may opt for CAP plans because they anticipate requiring more dental care and think CAP plans would be less costly. In dental CAP plans, there is a tendency for the covered benefits to appear to be more generous and have lower copayments, fewer deductibles and no caps on benefits compared with FFS plans. Another incentive for enrolling in CAP plans is the anticipated use of specialists. Unlike in medicine, general dentists, not specialists, provide most of the dental care; therefore, access to specialists is a less important incentive to enroll in FFS plans.

Our finding that CAP-plan enrollees perceived their oral health to be worse than that of FFS-plan enrollees adds to the concern regarding the quality of care in CAP plans. The relationship between the quality of care and capitation is not self-evident.^1,20 A major difference between FFS plans and CAP plans is that CAP-plan forms of insurance shift the financial risk for providing services from the insurance company to dentists. The logic behind this type of insurance is that if the provider shares some of the risk for escalating costs, it will be more aggressive in restricting the costs. The inherent incentives/disincentives might result in the dentist’s denying treatment or providing cheaper services to control costs, but it also might increase preventive services so the dentist later might avoid more costly restorative services. Marcus²¹ noted that while abuse of FFS plans can lead to overtreatment, excessive use of expensive services and materials, with underuse of less-expensive services and materials, in CAP plans, abuse can result in lack of access to care, under-treatment and reduced time with patients. CAP plans are attractive to consumers and employee benefit managers because they often have lower premiums and better benefit structure. To be attractive to providers, however, there must be reasonable compensation for the services rendered. Marcus and colleagues²² have shown that the providers in many CAP plans can obtain reasonable compensation only by lowering the number of patient visits per year or the percentage of enrollees using the plan. They also may provide more services that are not covered by the plan. The degree to which these practices occur has not been established through a body of research.

Previous studies have examined the relationship between plan type and the quality of care. Atchison and Schoen²³ found that neither FFS-plan nor CAP-plan programs routinely provided good dental care. In a controlled trial in England and Scotland^24–26 comparing CAP-plan and FFS-plan dental care for children, the results showed that those under the CAP plan had fewer restored teeth, had more carious untreated teeth, had fewer radiographs taken and were recalled less frequently. However, the CAP-plan dentists provided more preventive care (advice on prevention, prescriptions for fluoride supplements, dietary advice, hygiene advice) than did dentists in the FFS-plan group. Beazoglou and colleagues²⁷ presented conflicting results for the provision of preventive care and found that patients in the FFS plan received twice the volume of preventive services as those in the CAP plan did. They also found that the CAP plan consistently substituted less expensive procedures for more expensive ones. The existing research is unclear regarding the effect of plan type on the quality of care and on oral health status.

As we reported previously, even though CAP-plan enrollees used dental services as much as FFS-plan enrollees, CAP-plan enrollees were less satisfied with their plans, gave the plans lower ratings, were less satisfied with their dentists and more often perceived their oral health as "fair to very poor."¹⁴ These findings raise important policy issues. In terms of having access to dental care, there were no statistically significant differences between CAP-plan and FFS-plan enrollees. But in terms of patients’ perceptions toward the quality of care received and the effect of the plan on their oral health, CAP-plan enrollees were at a disadvantage. Determining whether these various indicators of patient perception and attitude accurately, but indirectly, measure quality of dental care will require more studies.

	CONCLUSIONS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
We conducted this study to examine the impact of differences in type of dental plan, dental insurance market, premiums, out-of-pocket costs and enrollee demographics on enrollees’ perceived oral health status. Those enrolled in CAP plans were less likely to rate their oral health excellent, very good or good relative to those enrolled in FFS plans. This finding raises the question of whether those who perceive themselves to have poor oral health select CAP plans, as well as concerns about the quality of care in CAP plans. Further studies are needed to determine if adverse selection of CAP plans occurs and if the quality of care provided in CAP plans is inferior to that in FFS plans.

Another finding of this study was that non-whites were less likely than whites to rate their oral health as excellent, very good or good oral health. Disparities in perceived oral health status among racial/ethnic groups persist despite all having potential access to dental care. It is important for dentistry to address the noneconomic barriers to care, including cultural beliefs and health behaviors that may lead to poorer perceived oral health among racial and ethnic groups.

Financial factors are another major concern that has a significant influence on enrollees’ perceived oral health status. We found that enrollees with high incomes (greater than $100,000) were more likely than those earning less than $100,000 to report excellent oral health versus fair, poor or very poor oral health. In addition, we found that as enrollees’ out-of-pocket costs increased there was a decreasing likelihood of their reporting excellent oral health.

Enrollees who used dental care services were more likely than nonusers to report having excellent, very good or good oral health. It would seem that nonuse was not driven by a belief that they were in excellent oral health, and it would seem some other barrier is operating to prevent enrollees seeking care. Additional outreach and promotion of dental services may help encourage nonusers to use dental services and improve their oral health.