A Longitudinal Analysis of Unmet Need for Oral Treatment in a National Sample of Medical HIV Patients

Abstract

This longitudinal study examines perceived unmet dental need in a nationally representative probability sample of HIV-infected persons in medical care. A logistic regression analysis modeled the relationship between unmet need and explanatory variables. We estimate that 40% of HIV/AIDS patients report an unmet need associated with being male, being unemployed, injecting drugs, being heterosexual, lacking dental insurance, and having less education. Disparities in unmet need are related to socioeconomic status rather than to disease stage or ethnicity.

Perceived unmet need for oral health care is a useful measure of potential demand in that it represents whether people feel their “wants” for dental services are being fulfilled. In the general adult population, 16.7 million people (11%) reported needing oral health care but not receiving it. African Americans reported higher rates of unmet need for oral health care than other groups.1 Reports of perceived unmet need for oral health care among persons living with HIV/AIDS range from 5% to 52%.2–6 These studies occurred prior to the wide use of highly active antiretroviral therapy (HAART).

The current study expands on previous work in 2 ways. We use longitudinal, rather than cross-sectional, data from the first national probability sample of patients receiving care for HIV in the United States.7 Second, this study reports the prevalence of perceived unmet need for oral health care from 1996 through 1998, the period during which HAART was introduced and became the standard of treatment. These medications alter the course of the disease and appear to reduce the frequency and severity of oral opportunistic lesions.8–12

METHODS

The HIV Costs and Services Utilization Study (HCSUS) created a representative sample of HIV-infected adults in medical care in the contiguous United States. The 2864 initial study participants in HCSUS represent approximately 231400 people with HIV disease in the United States. The interviews were conducted in 3 waves from January 1996 through January 1998. Full details of the design are available elsewhere.13,14 Data analysis is based on a weighted HCSUS patient cohort that completed all 3 interviews (2109 subjects). Reported unmet oral health need was measured by a binary outcome variable (1=yes, 0=no). Respondents were asked if there was a time when they needed dental treatment but could not get it. The analysis consists of bivariate comparisons between unmet need and each of the covariates and multivariate modeling using a generalized estimating equations (GEE) approach. The GEE method enables us to account for the dependence of repeated observations from the same subject over time with a binary outcome.15,16

RESULTS

We estimate that 88 000 (40%) of HIV-infected medical patients reported unmet need for oral health care during at least 1 of the 3 HCSUS interviews. In the multivariate analyses presented in Table 1 ▶, among demographic variables, gender and education showed significant differences. Over the course of the study, women were 25% less likely to have perceived an unmet oral health need than men. Compared with college graduates, persons with some college reported a 44% increase in perceived unmet need, high school graduates a 45% increase, and persons with some high school education an almost twofold increase.

TABLE 1—

Multivariate Longitudinal Analysis Predicting Unmet Need for Oral Health Care Among HIV Patients

Variable	OR	SE	P	95% CI
Gender
Male			Reference
Female*	0.75	0.10	.033	0.58, 0.98
Ethnicity
White			Reference
African American	1.00	0.12	.988	0.79, 1.26
Hispanic	1.02	0.16	.902	0.75, 1.38
Other	1.56	0.37	.061	0.98, 2.48
Age, y
18–34			Reference
35–49	0.92	0.10	.412	0.74, 1.13
≥50	0.80	0.15	.229	0.55, 1.16
Education
BA/BS degree			Reference
Some college*	1.44	0.25	.035	1.03, 2.02
High school degree*	1.45	0.26	.040	1.02, 2.06
Some high school*	1.91	0.38	.001	1.30, 2.81
Exposure
Males having sex with males			Reference
Intravenous drug use*	1.56	0.21	.001	1.20, 2.03
Heterosexuals*	1.75	0.28	.000	1.28, 2.39
Other (transfusions, etc.)*	1.65	0.37	.028	1.06, 2.57
Smoking
Never			Reference
Past	0.84	0.13	.256	0.63, 1.13
Current	1.23	0.15	.100	0.96, 1.57
Dental insurance
Private insurance			Reference
No dental coverage*	2.02	0.30	.000	1.51, 2.70
Medicaid + dental coverage	1.20	0.19	.250	0.88, 1.65
Medicaid, no dental coverage*	2.51	0.42	.000	1.81, 3.47
Drug therapy
HAART			Reference
Polydrug therapy	0.95	0.08	.539	0.80, 1.13
Monodrug therapy	1.16	0.16	.267	0.89, 1.52
No ARV therapy	1.26	0.18	.107	0.95, 1.66
Living arrangement
Living with someone			Reference
Living alone	0.88	0.09	.193	0.72, 1.07
Employment status
Employed (full-/part-time)			Reference
Not employed*	1.30	0.14	.013	1.06, 1.60
Latest CD4 count
≥500			Reference
200–499	1.13	0.11	.215	0.93, 1.38
50–199	1.09	0.14	.536	0.84, 1.41
0–49	1.04	0.16	.794	0.77, 1.42

Note. OR = odds ratio; CI=confidence interval; HAART=highly active antiretroviral therapy; ARV=antiretroviral.

*P < .05.

Compared with males who had sex with males, all other subgroups were more likely to report unmet oral health need during the course of the HCSUS study. The reported rate of unmet need was 56% higher among intravenous drug users, 75% higher among heterosexuals, and 65% higher among those exposed through hemophilia or blood transfusion (or not defined). Respondents without any dental insurance were twice as likely to report a perceived unmet need for oral health care as those with private dental insurance. Those enrolled in a Medicaid program that did not have a dental benefit were 2.5 times more likely to report a perceived unmet oral health need. Another consistent finding was that those who were employed were less likely to have a perceived unmet need. None of the disease stage or clinical measures showed any significant differences.

DISCUSSION

This report expands our previous research and models by examining perceived unmet oral health needs during the course of the entire HCSUS study using data from all 3 interviews. Neither CD4 count levels, nor being diagnosed with AIDS, nor taking HAART were significant determinants even though we had expected that more severe effects of HIV might lead to more unmet dental need. Not being employed, which in this population might be a proxy for not only income but also the effects of HIV, was the only measure suggesting that disease severity might have an effect on unmet dental need. One might also argue that employment status can be an indirect measure of well-being or disease stage.

Although the rate of unmet need decreased as the population increased their use of HAART, comparisons between those not taking HAART and those taking HAART were explained by other factors in the multivariate analysis, namely, that the most economically vulnerable persons were less likely to have access to this treatment. Factors associated with lower socioeconomic status (e.g., education, employment, and availability of dental insurance) were more important than ethnic minority status. Perceived unmet need for oral health care for this population is considerably higher than the general US population.

More than medical care, dental care presents greater social and economic barriers for working poor and indigent individuals living with a chronic disease. Our findings, therefore, have implications for the organization and delivery of dental care for the most economically vulnerable patients with HIV, particularly in states that do not include dental care in their public insurance programs. Injecting drug users and ethnic minorities, the groups among which the epidemic is spreading most rapidly, are also the most likely to be uninsured or inadequately insured. Providing dental coverage to those on Medicaid would probably reduce unmet need for dental care for these more vulnerable groups. Such coverage is becoming even more critical when states are seeking to eliminate adult dental benefits from Medicaid as a cost-saving measure in their current budgetary crisis. Besides greater generosity with respect to state-sponsored dental Medicaid coverage, there is also need for a more coordinated system linking the medical and dental care—perhaps at the same site—of patients with HIV. This comprehensive approach would require concerted efforts by clinicians, health care administrators, and public health practitioners to increase access to oral health care through coordinated case management and patient referral strategies.