| |
HIV+ Minorities & Poor Get Poor Care
|
| |
| |
UCLA HIV Study Shows Minorities And Poor Less Likely To Have Good Care
October 27, 2006 7:33 p.m. EST
Linda Young - All Headline News Staff Writer
Berkley, California (AHN) - A first of its kind survey has revealed that HIV-positive minorities, the poor and substance abusers who do not have regular outpatient medical care are far less likely to get treatment for HIV than others.
Other studies had shown that minorities, the poor and substance abusers who were receiving routine outpatient medical care for their HIV infections still had worse outcomes than others and died more quickly. But in this study UCLA researchers looked at patients who do not have regular care.
The lead author of the study, Dr. William Cunningham, professor of medicine in the Division of General Internal Medicine and Health Services Research at the David Geffen School of Medicine at UCLA, noted that HIV-infected patients who did not receive regular care were more difficult to find and track. But they often show up in emergency rooms.
"As we expected, they are much less likely to get routine outpatient care but more likely to get acute care, when they are at their sickest," Cunningham said. "This is just the group that needs to get grassroots outreach service," Newswise reported.
The researchers compared two samples of more than 3,000 patients, one group was interviewed in 1998 and the other in 2001 to 2002. The study will appear in the November issue of the Medical Care journal.
For this study, to be published in the November issue of the journal Medical Care, the researchers compared socio-demographic, clinical and health care utilization characteristics of HIV-infected adults from two samples: 1,286 people from the 2001-02 Targeted HIV Outreach and Intervention Initiative (Outreach) and 2,267 who were interviewed in 1998 for the HIV Costs and Services Utilization Study (HCSUS).
Outreach, a multi-site program initiated in 2001 under the Ryan White Care Act, was intended to locate HIV-infected people who are hard to reach and connect them with medical care. HCSUS focused on HIV-positive patients who were receiving care. The 16 Outreach study sites, which were spread throughout the U.S., offered a variety of services such as HIV testing and counseling, social services, case management and direct medical care. They all provided outreach services linking HIV-positive patients with continuous outpatient care. The researchers examined the demographic data, the kinds of medical services the patients used and which services would be the most helpful to them.
The study group was 59 percent black, compared with 32 percent of HIV-infected people who were receiving routine care and were tracked. The study group was also 20 Hispanic versus 16 percent who received routine care. And 75 percent had annual incomes of $10,000 or less compared with 45 percent of those receiving routine care. Also nearly 60% were unemployed, homeless, had no insurance or used heroin or cocaine compared with less than half who received routing care.
"....This study found that participants in the Outreach sample, which aimed at engaging hard-to-reach persons, were overwhelmingly racial/ethnic minorities, substance users, and the poor. These are many of the groups in which the HIV epidemic is spreading most rapidly. Furthermore, participants in the Outreach sample compared with those in HCSUS generally had less favorable health status and health care utilization characteristics, as indicated by greater likelihood of using illicit drugs, less frequent ambulatory care and antiretroviral treatment, and greater likelihood of using acute care services...."
Health Services Utilization for People with HIV Infection: Comparison of a Population Targeted for Outreach with the U.S. Population in Care
Medical Care: Volume 44(11) November 2006 pp 1038-1047
Cunningham, William E. MD, MPH*; Sohler, Nancy L. PhD; Tobias, Carol PhD; Drainoni, Mari-lynn PhD; Bradford, Judith PhD; Davis, Cynthia MPH; Cabral, Howard J. PhD; Cunningham, Chinazo O. MD**; Eldred, Lois DrPH, MPH; Wong, Mitchell D. MD, PhD*
From the *Division of General Internal Medicine and Health Services Research, Department of Medicine, UCLA, Los Angeles, California; Department of Health Services, School of Public Health, UCLA, Los Angeles, California; City University of News York Medical School, New York, New York; Boston University School of Public Health, Bedford, Massachusetts; SERL/CHRI, Richmond, Virginia; Charles R. Drew University of Medicine & Science, Lynwood, California; **Montefiore Medical Center, Bronx, New York; and HRSA, Rockville, Maryland.
Supported by grant #H97HA00203 from the Health Resources and Services Administrations (HRSA), Special Projects of National Significance (SPNS) Program. Dr. Cunningham also received partial support from the NIMH (R-01 # MH69087), the UCLA-Drew Project Export, NIH, National Center on Minority Health & Health Disparities, (P20-MD00148-01), and the UCLA Center for Health Improvement in Minority Elders/Resource Centers for Minority Aging Research, NIH, National Institute of Aging, (AG-02-004). The HIV Cost and Services Utilization Study was conducted under cooperative agreement U-01HS08578 (M.F. Shapiro, PI; S.A. Bozzette, Co-PI) between RAND and the Agency for Health Research and Quality. Substantial additional funding for this cooperative agreement was provided by the Health Resources and Services Administration, the National Institute of Mental Health, the National Institute on Drug Abuse, and the National Institutes of Health Office of Research on Minority Health through the National Institute of Dental Research. Additional support was provided by The Robert Wood Johnson Foundation, Merck and Company, Inc., Glaxo-Wellcome, Incorporated, the National Institute on Aging, the Office of the Assistant Secretary for Planning and Evaluation in the U.S. Department of Health and Human Services.
Abstract
Background: Many persons with HIV infection do not receive consistent ambulatory medical care and are excluded from studies of patients in medical care. However, these hard-to-reach groups are important to study because they may be in greatest need of services.
Objective: This study compared the sociodemographic, clinical, and health care utilization characteristics of a multisite sample of HIV-positive persons who were hard to reach with a nationally representative cohort of persons with HIV infection who were receiving care from known HIV providers in the United States and examined whether the independent correlates of low ambulatory utilization differed between the 2 samples.
Methods: We compared sociodemographic, clinical, and health care utilization characteristics in 2 samples of adults with HIV infection: 1286 persons from 16 sites across the United States interviewed in 2001-2002 for the Targeted HIV Outreach and Intervention Initiative (Outreach), a study of underserved persons targeted for supportive outreach services; and 2267 persons from the HIV Costs and Services Utilization Study (HCSUS), a probability sample of persons receiving care who were interviewed in 1998. We conducted logistic regression analyses to identify differences between the 2 samples in sociodemographic and clinical associations with ambulatory medical visits.
Results:
Compared with the HCSUS sample, the Outreach sample had notably greater proportions of black respondents (59% vs. 32%, P = 0.0001), Hispanics (20% vs. 16%), Spanish-speakers (9% vs. 2%, P = 0.02), those with low socioeconomic status (annual income <$10,000 75% vs. 45%, P = 0.0001), the unemployed, and persons with homelessness, no insurance, and heroin or cocaine use (58% vs. 47%, P = 0.05). They also were more likely to have fewer than 2 ambulatory visits (26% vs. 16%, P = 0.0001), more likely to have emergency room visits or hospitalizations in the prior 6 months, and less likely to be on antiretroviral treatment (82% vs. 58%, P = 0.0001). Nearly all these differences persisted after stratifying for level of ambulatory utilization (fewer than 2 vs. 2 or more in the last 6 months).
In multivariate analysis, several variables showed significantly different associations in the 2 samples (interacted) with low ambulatory care utilization. The variables with significant interactions (P values for interaction shown below) had very different adjusted odds ratios (and 95% confidence intervals) for low ambulatory care utilization: age greater than 50 (Outreach 0.55 [0.35-0.88], HCSUS 1.17 [0.65-2.11)], P = 0.05), Hispanic ethnicity (Outreach 0.81 [0.39-1.69], HCSUS 2.34 [1.56-3.52], P = 0.02), low income (Outreach 0.73 [0.56-0.96], HCSUS 1.35 [1.04-1.75], P = 0.002), and heavy alcohol use (Outreach 1.74 [1.23-2.45], HCSUS 1.00 [0.73-1.37], P = 0.02). Having CD4 count less than 50 was associated with elevated odds of low ambulatory medical visits in the Outreach sample (1.53 [1.00-2.36], P = 0.05).
Conclusions: Compared with HCSUS, the Outreach sample had far greater proportions of traditionally vulnerable groups, and were less likely to be in care if they had low CD4 counts. Furthermore, heavy alcohol use was only associated with low ambulatory utilization in Outreach. Generalizing from in care populations may not be warranted, while addressing heavy alcohol use may be effective at improving utilization of care for hard-to-reach HIV-positive populations.
"....Hispanic persons were more than twice as likely as whites to have low ambulatory medical visits (adjusted odds ratio [AOR] = 2.34, 95% confidence interval [CI] = 1.56-3.52) in the HCSUS sample but not in the Outreach sample (AOR = 0.81, 95% CI = 0.39-1.69, P = 0.02 for interaction term). Black subjects also were more likely than white subjects to have low ambulatory medical visits in the HCSUS sample; the odds ratio for black respondents was greater than one, but smaller in the Outreach sample, and the interaction was not statistically significant. Compared with those having more education, those having less than high school education were about twice as likely as college educated persons to have low ambulatory visits in the Outreach sample, but not in HCSUS (interaction NS). Compared with those with higher incomes, having an annual income of less than $10,000 was associated with a lower odds (AOR = 0.73, 95% CI = 0.56-0.96) of low ambulatory medical visits in the Outreach study but a higher odds in the HCSUS study (AOR = 1.35, 95% CI = 1.04-1.75, P = 0.002 for the interaction term). Having CD4 count less than 50 was associated with elevated odds of low ambulatory medical visits in the Outreach sample (AOR = 1.53, 95% CI = 1.00-2.36), but not in HCSUS (AOR = 0.93, 0.36-2.36; interaction NS). Also, use of heavy alcohol in the past 30 days was associated with about a 70% higher odds of low ambulatory medical visits in the Outreach study (AOR = 1.74, 95% CI = 1.23-2.45), but not in the HCSUS study (AOR = 1.00, 95% CI = 0.73-1.37, P = 0.02 for the interaction term)...."
Introduction
HIV disproportionately affects vulnerable populations, such as black and Hispanic communities, the poor, and substance users, who often have limited access to medical and social services.1,2 Numerous studies have demonstrated the difficulties that these groups have had in obtaining needed medical care.3-10 Most of these studies have concentrated on populations who were connected to medical care because they are the most convenient groups to study. For example, the HIV Costs and Services Utilization Study (HCSUS) examined a nationally representative sample of persons in care for HIV.11 Published manuscripts from HCSUS routinely acknowledge that persons who were not in regular care, or who received care in settings such as jails and emergency rooms, were unlikely to be included in the study. Nonetheless, several HCSUS articles observed suboptimal patterns of health care utilization among several vulnerable groups of persons with HIV.12 Suboptimal care consisted of having fewer than the recommended number of ambulatory care visits, emergency room visits that did not lead to hospitalization, any hospitalization, or the failure to receive antiretroviral therapy when clinically indicated.12,13
To address the problems with access to care among groups who are hard to reach and may receive suboptimal medical care, the Health Resources and Services Administration HIV/AIDS Bureau initiated the Targeted Outreach and Intervention Initiative, a multisite Special Project of National Significance (SPNS) under the Ryan White Care Act, in 2001. A central goal of this project is to use outreach strategies to find hard-to-reach HIV-positive persons and engage them in medical care. Grass-roots outreach programs were selected for their track record in identifying and providing services to HIV-positive populations who are hard to reach. Implicit in this Outreach Initiative was the objective of understanding the characteristics of persons with HIV who are less likely to be represented in clinical trials, cohort studies, or nationally representative studies of persons receiving care. However, it is not known whether the Outreach Initiative actually reached participants who are less likely to be represented in these studies or to receive suboptimal care.
It is important for studies to target those who are hard to reach and may receive suboptimal care for several reasons. First, it is estimated that only about half of all persons infected with HIV in the United States either know that they are infected14 or receive regular medical care.11 Second, the HIV epidemic is spreading most rapidly among vulnerable groups, such as the black and Hispanic communities, substance users, and the poor, who are also generally less likely than others to receive medical care for other health conditions.15-25 Third, most studies of persons with HIV typically focus on persons receiving regular medical care from a provider or select participants from the investigators' or providers' own patient population.4,10 To address the gap in knowledge about hard-to-reach HIV-positive populations, this study compares sociodemographic, clinical, and utilization characteristics of 2 HIV-positive populations across the United States. The index sample is from the SPNS Outreach Initiative, a multisite evaluation of programs designed to target hard-to-reach HIV-positive persons (Outreach sample). The comparison sample is from the well-known HIV Costs and Services Utilization Study (HCSUS), a nationally representative cohort of persons with HIV infection who were receiving care in the United States. The goal of this study is to examine whether the characteristics of the Outreach sample are consistent with a population that is out of care relative to a well-known, nationally representative population that is in care. In addition, we compared the 2 samples after stratifying by whether ambulatory medical visit frequency met guidelines versus less frequent utilization.12,26 Finally, we compared the correlates of ambulatory utilization that were observed in the 2 samples.
METHODS
SPNS Outreach Sample
The SPNS Outreach sample consisted of 1286 HIV-positive persons recruited from 16 Ryan White Care Act-funded study sites. This national project was designed to evaluate service delivery for people living with HIV who were hard-to-reach and may receive suboptimal HIV care. The 16 study sites are located in Anniston, AL; Boston, MA; Chicago, IL; Detroit (2), MI; Los Angeles, CA; Miami (2), FL; New York, NY; Portland, OR; Providence, RI; Sacramento, CA; San Antonio, TX; Seattle, WA; St. Louis, MO; and Washington, DC. The Outreach study sites were the settings for a variety of different kinds of programs that provided a range of services, including HIV testing and counseling (8 sites), case finding and linkage to HIV clinics (10), case management (10), and direct medical care (7). Venues for outreach services included client neighborhoods (12), churches or other community organizations (11), public medical clinics in low-income areas (7), bars (6), mobile vans that traveled to different neighborhoods (5), homeless shelters or single resident hotels (4), jails (4), and parks (3). The hard-to-reach populations served were defined locally by each program and included active substance users (5), medically indigent (5), commercial sex workers (4), incarcerated persons (4), and homeless (2). Previous research guided the selection of these services, venues, and populations as likely ways to find hard-to-reach HIV-positive persons.27-33 Most sites offered services in several types of venues (n = 16), served multiple populations (n = 16), and provided multiple services (n = 12). Despite the range of services, venues, and populations, all of the programs provided supportive outreach services to link HIV-positive persons who were not in any care, not in consistent care or were at risk for dropping out of care with continuous ambulatory medical care. Further details about the outreach programs are available at: http://hab.hrsa.gov/special/%5CSPNS05RPT%5Coutreach.htm .
Across all programs, eligible persons were 18 years of age or older, HIV-positive, and able to complete the interview in English or Spanish. A purposive sample of eligible, hard-to-reach persons was selected during the study period by each of the outreach service programs. Trained interviewers obtained written informed consent and collected sociodemographic, clinical, and utilization data using a survey instrument covering the 6-month period prior to the interview. Baseline interviews were conducted with 1286 HIV-positive persons between 2001 and 2003. All participating institutions obtained institutional review board approval for the study.
HCSUS Comparison Sample
Full details of the HCSUS probability sampling design are presented elsewhere.12,34 Eligible persons included those who were HIV-positive and aged 18 years or older recruited from a representative sample of HIV provider practices throughout the contiguous U.S. Baseline and 2 follow-up interviews were conducted between 1996 and 1998. At baseline, the HCSUS completed IRB approved interviews with 2864 persons. The present study uses HCSUS data from the 2267 (79% of those completing the baseline survey) respondents who completed the second follow-up interview in 1998 because it was the most recent interview available for comparison on this topic. More than 99% of HCSUS respondents reported having a usual source of care. HCSUS data were weighted to permit inference to the population of HIV-positive persons in care in the United States.
Comparison Variables
To conduct this analysis, we constructed a set of identical variables collected in both samples, reflecting health care use, sociodemographic, and clinical characteristics. These variables are described in detail herein.
Measures of Health Care Use
The main dependent variable is respondent reported number of ambulatory medical visits in the prior 6 months. Having fewer than 2 ambulatory visits in 6 months is a well-established indicator of poor access to HIV care.5,12,13 In addition, we report data on emergency room visits and hospitalization because higher utilization of acute care may indicate inadequate ambulatory medical care for the degree of illness.12 Use of a case-manager35 and mental health care visits36 are included in this analysis because studies have shown that availability of these services help to improve health outcomes, particularly for poor, minority, and drug using persons with HIV infection. Finally, the current use of antiretroviral therapy, which improves health outcomes, is included.37 To account for the change in recommended guidelines for the use of antiretroviral therapy over time, we compared the rates of antiretroviral use after stratifying for CD 4 cell count less than 500 cells/mm3 and less than 350, corresponding to the threshold for initiating antiretroviral therapy in different guidelines.26,38,39
Measures of Sociodemographic and Clinical Characteristics
We restricted measures in this analysis to items common to both HCSUS and Outreach samples. We included mutually exclusive measures of age (18-34, 35-49, 50 and older), gender (male and female), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic/Latino, non-Hispanic other), language of interview (English or Spanish), educational attainment (less than high school degree, high school degree, some college, and college degree or more), annual income (less than $10,000 vs. $10,000 or greater), employment status (currently working vs. not working), unstable housing (living other than in one's own home or apartment; yes or no), homelessness during the past 6 months (yes or no), and insurance status (private, Medicaid, Medicare, no insurance).
To provide data on clinical characteristics that might affect the comparisons, we assessed the most recent (prior 6 months) self-reported CD4 cell count (<50, 50-199, 200-349, 350-499, 500 or more)40; self reported health status (single item, overall health: excellent, very good, good, fair, poor); and yes/no items on heroin or cocaine use ever, and more than 5 drinks of alcohol in 1 day during the previous 30 days.41
Analysis Plan
We first compared the sociodemographic characteristics, clinical characteristics, and health care use of the 2 samples, using χ2 tests for detecting differences in categorical variables and t-tests for detecting differences in continuous variables. We hypothesized that the Outreach sample would include higher proportions of persons from traditionally vulnerable backgrounds, with worse clinical status, with lower ambulatory medical use and with higher acute medical care use (emergency department and hospital visits). Next, we compared the sociodemographic and clinical characteristics mentioned previously between the 2 samples after stratifying on use of ambulatory medical care (fewer than 2 vs. 2 or more visits in 6 months).12 This cut-point for stratifying ambulatory medical visits was chosen because guidelines have recommended medical visits at least every 3 months or 2 in 6 months.26 Although the HCSUS sample recruited participants through clinical providers and mainly included persons receiving medical care, the study also sampled some individuals receiving infrequent or sporadic HIV care, allowing for the comparison between the 2 samples stratifying on number of ambulatory visits. We hypothesized that the Outreach sample in the lowest utilization strata would have the highest proportions of traditionally vulnerable persons, and would have lower clinical status indicators.
To examine whether the HCSUS and Outreach samples also differed in other ways that have implications for the types of strategies that are most appropriate for improving utilization of HIV care, we conducted further multivariate analysis. We conducted 2 logistic regression analyses using the same model, first for the Outreach sample, and then for the HCSUS sample to examine potential differences in the relationship between ambulatory visits and the sociodemographic and clinical variables. For both models, the dependent variable was having fewer than 2 ambulatory visits in the prior 6 months (vs. 2 or more visits), and the independent variables were age, sex, race/ethnicity, educational attainment, annual income, insurance status, CD4 count, self-reported health status, heroin or cocaine use ever, and heavy alcohol use in the prior 30 days. To formally test whether the predictors significantly differed between the 2 samples, we pooled the data from both samples and conducted another logistic regression analysis, including a variable for study sample (Outreach versus HCSUS) and interaction terms for the sample and each of the independent variables listed previously. Differences in significant multivariate correlates between the 2 samples would provide suggestive evidence that approaches to improving care might need to be different for hard-to-reach populations like those in the Outreach sample than for populations in care represented by HCSUS. We imputed missing data for some essential variables (eg, insurance status, income, and CD4 count) in both samples using standard hot-deck methods.42,43 We examined alternative models that excluded imputed data and found very similar results. All analyses involving the HCSUS sample incorporated weights to account for the complex sampling employed in that study. For the Outreach sample, each site was treated as if it were a primary sampling unit identical to those in HCSUS; the results for the Outreach sample are equivalent to unweighted results. Linearization methods were used in all models to account for clustering, stratification, and weights, and to estimate odds ratios, standard errors, and levels of significance, using the survey functions of SUDAAN.44,45
RESULTS
Comparison of Outreach and HCSUS Sample Characteristics
Consistent with our hypothesis, the Outreach sample had a higher proportion of black respondents (58.6% in Outreach vs. 32.3% in HCSUS), Hispanic (19.7% in Outreach vs. 15.5% in HCSUS), and Spanish-speakers (8.8% in Outreach vs. 2.5% in HCSUS) than the HCSUS sample (Table 1). In addition, more of the Outreach respondents had no high school degree, had less than $10,000 in annual income, did not work, had unstable housing or were homeless, and were uninsured or publicly insured. In terms of clinical characteristics a higher proportion of Outreach participants had CD4 counts greater than 500 cells/mm3, and had higher levels of ever using heroin or cocaine.
Comparison of Outreach and HCSUS Samples' Utilization of Services
A greater proportion of the Outreach sample had fewer than 2 ambulatory visits in the prior 6 months than in HCSUS (25.9% in Outreach vs. 15.6% in HCSUS; see Table 1, bottom). A higher proportion of the Outreach sample had an emergency room visit and a hospital visit in the prior 6 months. The proportion that reported having a case manager was greater in the Outreach sample than in HCSUS. Correspondingly, a greater proportion of Outreach than HCSUS respondents reported mental health care visits in the prior 6 months. Finally, a substantially greater proportion of HCSUS than Outreach respondents were receiving antiretroviral treatment, even when the comparison was restricted to those with CD4 counts less than 500 cells/mm3 or less than 350 (62.1% in Outreach vs. 89.7% in HCSUS, P = 0.0001).
Comparison of Outreach and HCSUS Characteristics, Stratified for Level of Ambulatory Utilization
In the stratum with fewer than 2 ambulatory medical visits in 6 months, compared with the HCSUS sample, the Outreach sample had significantly higher proportions of black participants (64.0% vs. 38.1%, P = 0.004), those with less than high school degree (49.6% vs. 28.1%, P = 0.03), those with annual incomes less than $10,000 (71.2% vs. 51.1%, P = 0.006), the unemployed (74.8% vs. 59.2%, P = 0.01), those without housing (31.2% vs. 6.1%, P = 0.0001), the uninsured (34.5% vs. 25.5%, P = 0.0003), those with CD 4 counts 500 or greater (30.6% vs. 23.6%, P = 0.02, those who had a case-manager (80.5% vs. 57.0%, P = 0.003), and those who were not taking antiretroviral medications, including among those with CD4 counts <350 (54.2% vs. 86.1% received antiretrovirals, P = 0.0001; see Table 2). In the stratum with 2 or more ambulatory medical visits, the differences between samples were very similar, and in some cases were more pronounced.
Multivariate Analysis of Utilization in Outreach and HCSUS
Using separate multiple logistic regression models for the Outreach and HCSUS samples, we examined the relationship of sociodemographic and clinical characteristics with low use of ambulatory medical visits, which was defined as having fewer than 2 visits in the last 6 months; the pooled sample multivariate analysis was used to formally test for interactions observed in the stratified models (Table 3 reports ORs and 95% CIs from each separate regression model and the P-values from the pooled sample interaction analysis). The odds of having a lower number of ambulatory medical visits for those lacking health insurance were similar in the 2 samples, but insurance status was a statistically significant correlate in the HCSUS sample only. The association of many of the other patient characteristics with low ambulatory medical visits, however, differed between the 2 study samples. Compared with persons younger than 35 years, Outreach subjects at least 35 years of age were less likely to have low ambulatory medical visits, whereas HCSUS subjects ages 35-49 were more likely to have low ambulatory medical visits. The different relationship of age and ambulatory medical visits between the 2 samples was statistically significant (P = 0.0005) based the interaction effect between study sample and age in the multivariate analysis pooled across both studies.
Hispanic persons were more than twice as likely as whites to have low ambulatory medical visits (adjusted odds ratio [AOR] = 2.34, 95% confidence interval [CI] = 1.56-3.52) in the HCSUS sample but not in the Outreach sample (AOR = 0.81, 95% CI = 0.39-1.69, P = 0.02 for interaction term). Black subjects also were more likely than white subjects to have low ambulatory medical visits in the HCSUS sample; the odds ratio for black respondents was greater than one, but smaller in the Outreach sample, and the interaction was not statistically significant. Compared with those having more education, those having less than high school education were about twice as likely as college educated persons to have low ambulatory visits in the Outreach sample, but not in HCSUS (interaction NS). Compared with those with higher incomes, having an annual income of less than $10,000 was associated with a lower odds (AOR = 0.73, 95% CI = 0.56-0.96) of low ambulatory medical visits in the Outreach study but a higher odds in the HCSUS study (AOR = 1.35, 95% CI = 1.04-1.75, P = 0.002 for the interaction term). Having CD4 count less than 50 was associated with elevated odds of low ambulatory medical visits in the Outreach sample (AOR = 1.53, 95% CI = 1.00-2.36), but not in HCSUS (AOR = 0.93, 0.36-2.36; interaction NS). Also, use of heavy alcohol in the past 30 days was associated with about a 70% higher odds of low ambulatory medical visits in the Outreach study (AOR = 1.74, 95% CI = 1.23-2.45), but not in the HCSUS study (AOR = 1.00, 95% CI = 0.73-1.37, P = 0.02 for the interaction term).
DISCUSSION
This study found that participants in the Outreach sample, which aimed at engaging hard-to-reach persons, were overwhelmingly racial/ethnic minorities, substance users, and the poor. These are many of the groups in which the HIV epidemic is spreading most rapidly. Furthermore, participants in the Outreach sample compared with those in HCSUS generally had less favorable health status and health care utilization characteristics, as indicated by greater likelihood of using illicit drugs, less frequent ambulatory care and antiretroviral treatment, and greater likelihood of using acute care services.
It is particularly noteworthy that we found many large differences between the Outreach sample and HCSUS, which is the only available national probability sample of person in care for HIV in the United States. Previous studies have reported wide variations in care within the HCSUS sample, and found that minorities, women, drug users and the poor often received fewer services.5,11,12,46,47 Thus, we hypothesized that the Outreach sample would resemble the low utilization group from HCSUS. However, we found that the Outreach sample had an even less favorable pattern of utilization (fewer ambulatory visits, more acute care visits), and an even higher proportion of blacks, Hispanics, women, and drug users when compared with the low utilization strata of HCSUS. The Outreach sample appears to be a highly vulnerable group of people, many of whom face barriers in obtaining ambulatory medical care and are largely different from persons described in the HCSUS study in ways that may require a different approach to improve their care. The good news may be that this study shows that it is possible to identify a population of hard-to-reach persons with HIV whose characteristics put them at greater risk of poor outcomes, and who generally received suboptimal medical care.
This study provides substantial evidence that the strategies needed to improve care might differ between the populations represented by the Outreach and HCSUS samples. We found that the associations of age, Hispanic ethnicity, income, and heavy alcohol use with low ambulatory care utilization were significantly different (interacted) between the 2 populations. These differences in associations suggest that very different approaches may be needed to improve care for hard-to-reach groups who can only be identified through programs such as those included in the Outreach sample. For example, while low income was associated with low utilization in HCSUS, it was associated with high utilization in the Outreach sample. This finding suggests that while targeting low income groups is important, existing public programs may be already effective in maintaining utilization for this very low income group.
Data limitations only allow us to look at those with annual incomes of less than $10,000; therefore, many poor persons are also represented in the greater than $10,000 annual income comparison group. However, it is striking that in the Outreach sample 75% had annual incomes less than $10,000 compared with only 45% of the HCSUS sample. Another example is that recent heavy alcohol use was strongly associated with lower utilization in the Outreach sample but not in HCSUS, which suggests that providing alcohol treatment services may be particularly beneficial in increasing ambulatory medical care use among persons reached through outreach efforts. Recent studies have emphasized the role of heavy alcohol use in HIV risk behavior,48,49 but little study has been directed toward alcohol as a barrier to the use of HIV services in underserved populations. Having a CD4 count less than 50 was associated with elevated odds of low ambulatory medical visits in the Outreach sample, a potentially alarming finding since this group is at greatest risk of poor health outcomes and is most in need of medical services. Although the Outreach sample received more mental health care and case-management (services that could help to address heavy alcohol and drug use problems) than the HCSUS sample regardless of level of ambulatory medical visits, Outreach participants had markedly lower antiretroviral therapy use. Lower antiretroviral therapy use, despite similar CD4 cell count levels could also results in worse health outcomes.50 Our findings suggest important directions for future interventions that address improved utilization of medical care among vulnerable populations with HIV infection.
Despite the great policy concern about persons who are positive for HIV but who are undiagnosed or not in care, it is not possible to conduct a national survey of HIV-positive persons whose HIV infection is not known. Similarly, there are no ready means to identify a national sample of persons who are known to be HIV-positive but have not received any medical services. To identify potentially underserved HIV-positive persons, we recruited from various Ryan White Care Act-funded organizations that provide outreach services to hard-to-reach persons who were not in any care, not in consistent care or were at risk for dropping out of care across a wide range of geographic regions in the United States. Although this study allowed us to include persons not well-represented in HCSUS, we do not know the extent to which the study participants represent the universe of persons who are HIV-positive, but are not continuously engaged in care. Thus, it is very informative to compare the Outreach sample to a nationally representative sample of persons who were sampled from providers, and who were receiving care. Our study is also limited by the lack of detailed measures of antiretroviral use that would have enabled the examination of HAART use in the Outreach sample. Although we chose the last survey wave of HCSUS to minimize the difference in time of data collection with the Outreach study, the comparisons between samples would have been even stronger if the data had been collected contemporaneously. Still, it is unlikely that the 3-year time difference between the 2 studies could account for very much of the differences we observed between the 2 study samples.
Despite these limitations, we found substantial differences between samples in the proportions of traditionally vulnerable populations, in the utilization of services, and in the characteristics associated with low utilization. These findings suggest that generalizing results to hard-to-reach groups from a population identified because they were in care may not be warranted. The differences we found indicate that the Outreach sample we identified is in extensive need of additional services to improve their care. Furthermore, our findings suggest that the types of services that the Outreach participants could benefit from most may be different from those that might be directed to a population in more regular ambulatory care. In addition to well-known factors such as insurance coverage, our findings also suggest that addressing heavy alcohol use could have a substantial impact on improving utilization for hard-to-reach HIV-positive persons identified by supportive outreach programs.
|
|
| |
| |
|
|
|
