The cost factor is complicated by issues raised by the disease itself because of its associated oral manifestations, as well as problems HIV-infected people have with access, disadvantaged status and coping with a host of other issues that may affect their use of and demand for dental care. Brown and colleagues¹ estimated that the U.S. per capita out-of-pocket cost for dental care in 1996 was $157 for a person at the poverty level according to the federal poverty guidelines and $229 for people with higher incomes (400 percent of the poverty level and greater). With respect to the population of people with HIV/AIDS, little has been reported regarding expenditures for dental care. Hsia and colleagues⁸ reported that children with AIDS averaged expenditures of almost $38,000 annually for both inpatient and outpatient services, including one dental visit. For adults, the annual expenditures for all services in 1996 were between $20,000 and $24,700. Bozzette and colleagues⁹ estimated that the cost of medical care for the same population of HIV-infected medical users presented in this article was about $20,300 per year in 1996. There is a dearth of information regarding expenditures for dental care among people living with HIV and/or AIDS.

The objective of the study we report here was to examine the factors associated with out-of-pocket expenses for dental care, based on self-reported data, among a national probability sample of HIV-infected people accessing medical care.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
This article presents data from the HIV Cost and Services Utilization Study (HCSUS), a research project funded by a cooperative agreement between the Agency for Healthcare Research and Quality (AHRQ) and RAND Corp., a private nonprofit research institution in Santa Monica, Calif.¹⁰ These data are from a nationally representative probability sample of HIV-infected adults accessing medical care in the contiguous United States in early 1996. The reference population was limited to people at least 18 years old with known HIV infection who made at least one visit for regular or ongoing care to a nonmilitary, nonprison medical provider other than an emergency department between Jan. 5 and Feb. 29, 1996. Full details of the design are available elsewhere.^11,12 We sampled patients from anonymous lists of all eligible patients who visited participating providers during January and February 1996. We sampled 4,042 eligible subjects, and 2,864 (71 percent) completed full interviews. For the purposes of national estimation, we used 68 percent of the complete interviews. National Opinion Research Center (NORC) interviewers contacted randomly selected subjects for interviews only after providers or their agents obtained permission for them to do so. The institutional review boards of RAND Corp. and the University of California, Los Angeles, reviewed all consent forms and informational materials before they were distributed to subjects.

There is a dearth of information regarding expenditures for dental care among people living with HIV and/or AIDS.

Baseline interviews using computer-assisted interviewing instruments designed for this study¹³ began in January 1996 and ended in April 1997. NORC conducted 91 percent of the 2,864 interviews in person and the remainder over the telephone. The data reported here are from the first follow-up interviews, which were conducted from December 1996 to July 1997. NORC interviewers conducted interviews with 2,466 respondents (86.1 percent of the baseline cohort). Interviewers asked respondents to report on the amount of money they paid out of pocket for dental care. For those who could not remember the exact amount, ranges of out-of-pocket payments were provided so that they could identify the most appropriate one. The oral health–related quality of life (HRQOL) measure used in this survey was derived from data elements in the study and includes seven items, one of which deals with social functioning, two with physical functioning, two with psychosocial concerns and two with pain.¹⁴ The time frame for these responses was "within the last four weeks."

The dependent variable used in the analysis was each subject’s response to the question, "How much have you spent on your dental care in the past 12 months? Please include payments made by you, your family or friends for your dental care. Do not include any payments made by insurance." The out-of-pocket costs refer only to payments for direct services, not travel costs, medications or opportunity costs such as lost income. The covariates used in the analysis included demographic characteristics (such as age, ethnicity and education), enabling characteristics (usual source of dental care, funding under the Ryan White Comprehensive AIDS Resources Emergency [CARE] Act, dental insurance, employment status), disease and health status (AIDS diagnosis, latest CD4 count, score on the oral HRQOL measure, perceived oral health status), and behavioral and living situation (exposure, smoking, living alone and employment status). We performed bivariate and multivariate analyses. We used weighted data to estimate national probabilities for each of the covariates. We performed analyses of variance to compare out-of-pocket costs with each categorical covariate. For multivariate analysis, we used a log transformation of the out-of-pocket data to ensure the normality of the dependent variable and to enable the confirmation of the statistical significance of the covariates.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Table 1 presents the findings of the bivariate analyses. These include the mean amount of out-of-pocket costs as well as the estimated number of patients and the total amount that they spent. We estimate that 135,000 HIV-infected patients who received medical care also accessed dental care in the 12 months before their interview. Their out-of-pocket costs totaled $20.5 million. The average annual out-of-pocket cost was $152 in 1996. Using the consumer price index, we estimate the average out-of-pocket cost would be $217 in 2004 dollars, a 43 percent increase over the period. Effectively, all out-of-pocket costs reported in this article are in 1996 dollars.

Enabling characteristics—such as source of dental care, dental insurance status and receiving medical care from a site where Ryan White CARE Act funding is available—all were statistically significant. Those whose usual source of care was private practice spent an average of $232 annually; those without a usual source of care spent an average of $150. Those whose usual source of care was either public or HIV clinics averaged $25 and $41 in out of pocket costs, respectively, while those who received care in dental school clinics averaged $65 in out-of-pocket costs. Those who reported having private dental insurance spent $213 annually, while those without insurance and not eligible for Medicaid spent $246. Those eligible for Medicaid fell into two categories. The first category, people living in states that included adult Medicaid dental benefits, spent, on average, $47 out of pocket. The second category, those who lived in states in which there were no dental benefits for adults, spent $84 out of pocket. An indirect enabling characteristic is whether the site at which the patient received medical services also obtained dental funding under the Ryan White CARE Act. We found that receiving care at sites with less Ryan White CARE Act funding resulted in higher out-of-pocket expenses. Patients who received medical services from medical sites that received no Ryan White dental funding averaged $169 in dental costs, while those in sites with less than and more than $100,000 in such funding paid $124 and $54, respectively. Finally, those who were working full- or part-time spent $207 compared with $108 spent by those who did not work.

The HIV-related measures and oral health quality-of-life characteristics such as CD4 counts, AIDS diagnosis or oral HRQOL scores were not associated with out-of-pocket expenditures for dental care in the bivariate analyses. Smoking showed a significant association with out-of-pocket expenditures. Those who never had smoked spent $183, while those who were previous smokers spent $187 annually. However, current smokers spent only $114, significantly less than their more health-oriented counterparts. With regard to risk behaviors, men who had sex with men spent $215, IV-drug users spent $90 and heterosexuals spent $51.

The bivariate analyses provide us with the national estimates and average annual out-of-pocket costs; however, they do not ensure the strength of the association for two reasons. First, each bivariate comparison cannot be viewed in isolation. These characteristics interact; without taking the other characteristics into account, we may be getting a distorted view. The second problem is that these statistical tests assume that the distributions are normal. We know that out-of-pocket cost is not distributed normally. To address both of these issues, we performed a log transformation of the dependent variable and conducted a multivariate regression analysis, which is shown in Table 2 (page 1411). This approach allowed us to examine the strength of the association for all the covariates examined in Table 1. However, the one disadvantage of log transformations is that it is difficult to interpret the coefficients. In addition, we included HRQOL as a continuous variable in the regression analysis.

IV drug users, heterosexuals and patients with other risk factors had significantly lower out-of-pocket expenditures than did men who had sex with men. CD4 count levels were not statistically significant in the bivariate analysis. However, in the multivariate analysis, compared with people whose CD4 counts were above 500, those with lower CD4 counts had lower expenditures. Finally, as scores on the oral HRQOL measure increased, out-of-pocket costs for dental care decreased.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The findings reported in this article add an economic dimension to our previous research that focused on access to oral health services and perceptions of oral HRQOL among a national probability sample of people living with HIV.

Out-of-pocket costs are not related solely to the degree of dental disease present but are influenced by a number of different elements. These include a number of service-related factors, such as the desire for and value placed on dental care, the type and mix of services provided, the ability to pay for dental services, and the degree to which services provided are subsidized by third-party payers or programs. For non–HIV-infected populations, Vargas and Manski¹⁵ found that Hispanics had the highest mean expenditure in 1987 ($304, which translates to $584 in 1996 dollars), followed by whites ($283, or $538 in 1996 dollars), with blacks having the lowest ($242, or $450 in 1996 dollars). Although these differences were not statistically significant, they provide a reference for exploring this issue in other populations. Vargas and Manski also found that 53 percent of expenditures were paid out of pocket, with blacks having a lower percentage of out-of-pocket costs (46 percent). Brown and colleagues¹ compared 1987 data with 1996 data and found that dental expenditures for white, black and Hispanic adults had declined by 1996 in real terms. In 1996, blacks continued to have the lowest per capita expenditures.

One finding of this study is that HIV-infected blacks had significantly fewer out-of-pocket expenditures than did whites and Hispanics. This is consistent with findings for non–HIV-infected populations. For blacks living with HIV/AIDS, this may be because they received less care, as evidenced by previous work¹⁶ showing that they had significantly fewer dental visits than did whites. The type of dental services received tended toward emergency or episodic care, which usually costs less than comprehensive care. The combination of less care with less expensive care may be the reason for lower out-of-pocket spending in this population. Since no clinical assessment was performed in this study, it is not clear as to the clinical need for dental care in this population. However, previous research indicates that blacks in general have higher rates of periodontal and untreated dental caries.¹⁷ In this study population, we found that a higher percentage of blacks reported an unmet need for oral health services than did whites.⁷ This would indicate that they probably required more service of higher complexity. If they had access to care and the needed services were provided, we would expect greater out-of-pocket expenditures. However, we do not know the actual services provided to this population.

In the general population, blacks and His-panics are less likely to have dental insurance than are whites.¹⁷ The lack of dental insurance should increase out-of-pocket costs. However, in the HIV/AIDS population, these ethnic groups had lower out-of-pocket costs. As discussed above, one explanation is that these groups received less care and less expensive care than did whites. Another related reason for this may be that they are likely to use public dental clinics rather than private practices. There are programs specifically geared toward the dental needs of people with HIV. These programs are a form of subsidization that replaces private dental insurance. Our data show that blacks living with HIV were more likely to have a public or HIV dental clinic as their usual source of care. Because many of these clinics are funded through the Ryan White CARE Act, this may explain why blacks in this study paid significantly less out of pocket than did whites and Hispanics. Hispanics living with HIV also are eligible for Ryan White CARE Act benefits or other programs but had higher out-of-pocket expenditures than did blacks. This may be due in part to language and cultural barriers that render them unaware of their eligibility for insurance or other benefits.

Income, of course, was an important predictor of out-of-pocket expenditures. People in the highest income group paid almost five times more for their dental care than did the lowest income group. It is obvious that the availability of discretionary funds is an overriding factor even when enabling characteristics and other covariates are controlled for. Those with private dental insurance paid more out of pocket ($213) than those who had Medicaid ($47) and even those who were eligible for Medicaid but lived in a state where there were no benefits for adults ($84). In the former case, the state covered costs, while in the latter case the patients had to limit their dental expenditures.

For patients who accessed medical care in a site that also received Ryan White CARE Act dental funding, there were lower out-of-pocket costs for dental care even when we took other covariates into account. Patients whose medical care sites received no Ryan White dental funds paid an average out-of-pocket dental cost of $169; patients whose medical site received less than $100,000 paid $124 and those whose site received $100,000 or more had out-of-pocket costs of $54. Apparently, some used the dental service at the site, and that had a significant effect on their costs of care.

The mode of exposure to HIV was a significant factor in out-of-pocket costs. Men who had sex with men had the highest expenditure ($215). IV-drug users spent $90 and heterosexuals only $51. The heterosexuals in this population tend to be the most disadvantaged. Even though the analysis takes into account their sex and ethnicity, these people tended to have the fewest options.

Compared with patients who had CD4 counts of 500 or greater, those with lower counts have lower out-of-pocket costs. There does not seem to be a trend with regard to CD4 counts. The P values do not show increasing levels of significance as the CD4 count declines. In fact, the mean out-of-pocket costs decrease, from $151 for CD4 counts of 500 and above, to $143 and $146 for CD4 counts 200 to 499 and 50 to 199, respectively. However, for counts below 50, the mean cost is $172, which is higher than that for those with CD4 counts of 500 and above.

The oral HRQOL measure used in this analysis was associated with out-of-pocket expenditures.¹⁴ The oral HRQOL is a continuous variable with a mean of 50, ranging from 7 to 60 in this sample. The model showed that this variable was significant in predicting out-of-pocket expenditures at the P = .002 level. The relationship between these two variables is inverse, in that an increase in oral HRQOL score decreased the out-of-pocket expenditure. From this, one may conclude that a person who reports having a high oral HRQOL score, all other things being equal, will have lower out-of-pocket costs.

	CONCLUSION

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
This article examines the complex issue of out-of-pocket expenditures for dental care in a population that is likewise complex. Not only is there a range of demographic characteristics in this population, but also a disease that imposes itself on behavior as well as oral health. Our approach has worked through this complexity to identify key factors in how out-of-pocket expenditures for dental care operate in this population; our approach also helped us identify these key factors’ relative importance. The study is limited by the sample’s being derived from users of medical care and therefore not reflecting the entire population with HIV/AIDS. Furthermore, the out-of-pocket costs were determined by the patient reports rather than actual payments made to practices or clinics. Future research should involve an in-depth analysis of both out-of-pocket and reimbursement by third-party and government sources. It would be important to establish the actual cost of dental care for this population.