Barriers and Unmet Need for Supportive Services for HIV Patients in Care in Los Angeles County,California

Introduction

P ersons infected with HIV frequently manage multiple challenges related to mental health, substance abuse, housing, transportation, and employment. ^{1 –3} The management of multiple health and social challenges concurrently can result in missed HIV primary care visits, suboptimal adherence to complex medication regimens and poor quality of life. ^{4 –7} Supportive or ancillary services were designed to help HIV-infected patients manage competing challenges and also facilitate the successful management of HIV disease. ^1,3,8 Supportive services for persons with HIV is considered any service other than primary medical care and typically includes case management, drug and alcohol treatment, housing support and transportation assistance.

The Ryan White HIV/AIDS Treatment Extension Act provided $2.29 billion in federal funding for the majority of the medical and support services for low-income persons with HIV infection in the United States in 2010. ⁹ Despite the resources allocated for supportive services for low-income persons with HIV, patients often have unmet service needs due to barriers that prevent access and service utilization. ^{1 –3,10,11} Barriers to receipt of HIV services may include language difficulties, active drug use, health beliefs, lack of case management, diminished mental status, inadequate housing, stigma and lack of disclosure of HIV status to social network members. ^{7,12 –15} Clients may also not know that they are eligible for services or may be unable to navigate the large and complex care delivery system. ^1,10,15

In Los Angeles County (LAC), a metropolitan area of the United States with the second highest number of people living with HIV, there are limited data on the unmet supportive service needs of persons with HIV and the barriers to needed supportive services. ¹⁶ The population in LAC most impacted by the HIV epidemic is men who have sex with men (MSM), African Americans and Latinos, many of whom have extensive supportive service needs due to high rates of poverty, unemployment, public assistance, and inadequate health insurance. ^{3,17 –20} In a study of 83 primarily Latino and African American low-income HIV patients who received their care at three LAC public county clinics from 2001 to 2004, all patients had a need for supportive services that included medical adherence support (16%), housing services (12%), and nutrition/food support (12%). ³ Data on supportive service needs from the HIV Cost and Services Utilization Study (HCSUS) that included a national probability sample of HIV-infected persons in care in 1996 also demonstrated that more than two thirds (67%) of patients had a need for at least one supportive service and almost a third (27%) had an unmet service need in the previous 6 months. ¹ Unmet need was greatest for benefits advocacy (35%) and substance abuse treatment (28%).

Detailed recent and representative local data are needed on the barriers and unmet needs of HIV patients in LAC to more effectively deploy limited and shrinking resources. Data are presented on the barriers to support services and factors associated with met and unmet services needs for a probability-based sample of HIV-infected patients in care in LAC in 2007 and 2008.

Methods

Data presented here were collected as part of the Medical Monitoring Project (MMP), a multisite study funded by the U.S. Centers for Disease Control and Prevention. MMP is a national supplemental surveillance system that monitors clinical and behavioral outcomes for HIV-infected persons in care in the United States. MMP uses a national, three-stage probability-based sampling scheme that includes 16 states, 6 jurisdictions and Puerto Rico. In LAC, 25 facilities were selected to represent the approximately 220 HIV medical care facilities in LAC in 2007 and 2008. A total of 21 facilities participated in 2007 and 16 facilities participated in 2008. The facility participation rate was 91% among eligible providers in 2007 and 84% in 2008 for an overall facility response rate of 88%. Among these, 52% of 2007 and 63% of 2008 sampled facilities received Ryan White funding.

The second sampling phase included the selection of patients. A total of 25,484 patients (14,601 in 2007 and 10,883 in 2008) were identified from the participating facilities as having received at least one HIV primary care visit from January 1 to April 30 in 2007 or 2008. A sample of 400 patients in 2007 and 400 patients in 2008 was randomly sampled from all of the patients with visits from January 1 to April 30 2007 and 2008. The patient response rate was 51% among all eligible patients (n=344) in 2007 and 41% among all eligible patients (n=391) in 2008 for an overall patient response rate of 45%. The overall study response rate was calculated by multiplying the overall provider response rate by the overall patient response rate for an overall response rate of 40% among eligible providers and patients.

Participant eligibility criteria included a confirmed HIV diagnosis, age 18 and older, ability to complete a survey in English or Spanish, ability to complete an informed consent, and an HIV primary care visit at one of the sampled HIV care facilities between January 1 and April 30 of either 2007 or 2008. Sampled patients were recruited through their HIV providers and referred to study staff for an interview and a medical record abstraction following completion of an informed consent. Participants were interviewed at the clinic, at home or at an agreed-upon location. Data were collected by trained interviewers using a handheld assisted personal interview device (HAPI) that was programmed using the QDS software. Data was collected on demographics, HIV testing, and care experiences, sources of care, met and unmet needs, reasons services were not obtained, treatment, medication adherence, sexual behavior, drug and alcohol use, prevention activities, mental health, and other health conditions.

The focus of the analysis presented here is on the unmet need for supportive services and barriers that prevented receipt of services. Participants were asked about the following supportive services: HIV case management services; mental health counseling; social services such as insurance assistance or financial counseling; assistance in finding a doctor for ongoing medical services; assistance in finding dental services; adherence support services; home health services such as home nursing care or assistance; chore or homemaker services (paid or volunteer); assistance in finding shelter or housing; assistance with finding meals or food; transportation assistance; childcare services; education or information on HIV risk reduction or other supportive services not listed.

Participants were asked if they needed and received each of the above services in the previous 12 months. Unmet need for a service was defined as a need for a service in the past 12 month that was not received. If a participant did not receive a needed service, they were asked to identify the main reason they did not receive a service. Answer options included: I did not know where to go or who to call; didn't complete application process; the system is too confusing; the waiting list is too long; it isn't available in my area; they charge too much; I don't have the money to pay; transportation problems; language barrier; not eligible/denied services; I am too sick to get out; or any other reason that was not listed. The reasons or barriers were combined into broader categories for analysis purposes and included an agency barrier (e.g., system too confusing; wait list too long; not available in area); lack of information (e.g., do not know where to go or who to call) and financial/practical barriers (e.g., they charge too much, transportation problems) and an other category. ¹⁰ Other reasons not listed as answer options that were considered agency barriers were “agency never responded,” and “waiting to hear from agency.” Other reasons not listed as answer options that were considered financial/practical barriers were “lack of proper forms” and “missed appointment, too busy.”

For participants reporting a need for at least one supportive service (n=236), socio-demographic characteristics are compared for participants with any unmet service need to those with no unmet service needs in the previous 12 months in a bivariate analysis using odds ratios (OR) and 95% confidence intervals (CI). Individual types of services needed are compared for the whole study group (n=333) by race/ethnicity (e.g., African Americans versus Latinos versus whites) using χ² tests. Frequency distributions for specific types of unmet service needs are presented for those needing at least one supportive service (n=236) and the reasons that services were not obtained are presented using the broader barrier to services categories.

A logistic regression model is presented for the whole study group (n=333) to identify predictors of service needs for whites, blacks and Latinos separately and combined, while controlling for confounding variables. A logistic model is also presented for the subset of participants needing at least one service (n=236) to examine predictors of unmet need that could not be stratified by race/ethnicity due to small numbers. All analyses were conducted using SAS version 9.1 (SAS Institute, Cary, NC). The project has received institutional review board (IRB) approval from all of the participating institutions that require IRB approval.

Results

Although not shown, 71% (n=236) of patients reported needing at least one supportive service and of these, 35% (n=83) reported at least one unmet need for a service. As shown in Table 1 in the total column, the majority of the study population with at least one service need were Latino (49%), male (78%), 30 years of age or older (91%), gay or bisexual (66%), born in the United States (54%), had less than a high school education (73%), were unemployed (54%), had an income over $10,000 per year (57%), were English-speaking (67%) and had no history of injection drug use (84%). The majority of the study group with at least one unmet service need were Latino (46%), male (83%), 30 or older (92%), gay or bisexual (77%), born in the United States (63%), had less than a high-school education (82%), and had an income over $10,000 (58%).

Homosexual or bisexual persons (OR=2.3, 95% CI: 1.2–4.3), those with more than a high school education (OR=2.1, 95% CI: 1.1–4.0), those on public assistance (OR=2.1, 95% CI: 1.2–3.7), those who were homeless in the previous 12 months (OR=2.6, 95% CI: 1.1–6.1) and persons with a history of injection drug use (IDU; OR=2.2, 95% CI: 1.1–4.6) were more likely to have at least one unmet need for services in the bivariate analyses shown in Table 1.

The types of services that were needed in the last 12 months for all study participants and for whites, African Americans and Latinos are shown in Table 2. Assistance finding dental services (39%), HIV case management (34%), mental health counseling (30%) and transportation support (23%) were the services most commonly needed by participants. A statistically larger percentage of blacks compared to whites and Latinos needed case management, homemaker services and shelter/housing services (p<0.05). A statistically larger proportion of African Americans needed transportation services compared to whites (p<0.05). In addition, a larger percentage of Latinos needed meal/food services compared to whites (p<0.05). A smaller percentage of whites needed assistance finding dental services and HIV education services compared to African Americans and Latinos (p<0.05).

The types of services that were needed but not received (unmet need) among only those who reported needing a service in the previous 12 months (n=236) were combined for all race/ethnicities due to small numbers and are presented in Table 3. The most common service needs that were not met were for shelter/housing (42%), assistance finding dental services (30%), social services (23%), homemaker services (21%), and mental health counseling (19%). The combined main reason that any of the services were not received includes lack of information (47%); an agency barrier (33%); and financial/practical barriers (18%).

As shown in Table 4, while controlling for race/ethnicity, gender, age, health insurance, language of interview and years since HIV diagnosis, the major predictor of any service needs for the whole study group combined was African American race/ethnicity (OR=3.1, 95% CI: 1.1–8.7) and an annual income less than $10,000 (OR=3.5, 95% CI: 1.3–9.3). The main factors associated with any service needs for whites was lack of health insurance (OR=8.1, 95% CI: 2.2–30.3) while controlling for gender, age, education, income, language, and years since HIV diagnosis. There were no major predictors of need for services for African Americans. Finally, among Latinos, the major predictor of any service need in the previous 12 months was low income (<$10,000; OR=4.0, 95% CI: 1.1–14.6).

As shown in Table 5, among only those needing at least one service (n=236), the major predictor of having an unmet need in the previous 12 months for all racial/ethnic groups combined was gay or bisexual sexual orientation (OR=2.8, 95% CI: 1.3, 6.1) while controlling for age, race/ethnicity, education, public assistance, homelessness and injection drug use.

Due to the low response rate, the demographic characteristics of the study sample was compared to that for all HIV patients reported to the Los Angeles County HIV/AIDS surveillance system for 2007 and 2008. The two datasets were very similar with the exception of the under-representation of African Americans (χ²=8.9; p=0.01), patients age 18–29 (χ²=4.926; p=0.03), and the overrepresentation of Latinos (χ²=6.9, p=0.01) in the study sample. In addition, MMP participants were compared to nonparticipants with respect to basic demographic characteristics and the study participants were statistically more likely to be Latino (χ²=6.87; p=0.009); over 50 years of age (χ²=7.05; p=0.008); and have an AIDS diagnosis (χ²=5.74; p=0.017). The major reason for study nonparticipation among the sampled MMP participants included that they were not reachable (62%) due to: a nonworking phone number, the provider recommended against calling, the patient was lost to the clinic, in rehab or a locked psychiatric facility, or there was no response from the provider.

Discussion

These data underscore the continuing and substantial need for supportive services for HIV-infected people in care in a large metropolitan area of the United States and highlight the sizable proportion of patients with a need for services that is not met. In fact, the percentage of HIV-infected patients in care in the current study who need supportive services and have unmet need is almost identical to a national estimate from 15 years ago from the HCSUS study. ¹ Although the HCSUS supportive service need definition referred to the previous 6 months and the MMP definition refers to the previous 12 months, 67% of the HCSUS patients in 1996 and 71% of the LAC MMP patients in 2007 and 2008 reported a need for at least one supportive service. In addition, 27% of the HCSUS sample and 35% of the LAC MMP sample reported at least one unmet need for a service. ¹ These data suggest that HIV-infected patients in care continue to have a large array of competing needs and that current service levels may not be not sufficient to address those demands.

The study findings highlight the self-reported barriers that prevent HIV-infected patients from accessing needed services and point to poor publicity regarding service availability as well as systemic problems with the delivery of services. The findings on lack of information as a barrier to services in the current study (47%) are consistent with findings from a study in Northern California in which 44% reported the same barrier to services. It should be noted that the time period in the Northern California referred to the past 4 months and the current study refers to the previous 12 months. ¹⁰ Clearly, information dissemination regarding supportive service availability for HIV-infected patients is not only an issue in LAC.

The finding that agency barriers are one of the most common reasons that HIV-infected persons can not access supportive services is also one of the main barriers to services reported in the Marx study. ¹⁰ More effective and efficient delivery of health care services in general is not a new problem but highlights the continuing importance of the need for improvement in the area of the delivery of HIV supportive services.

The finding that gay and bisexual HIV-infected patients are more likely than heterosexual patients to have at least one unmet need for HIV support services is an important finding for resource planning in LAC. Although the small numbers prevented in-depth analysis by race/ethnicity, HIV-infected Latino and African American MSM experience stigma related to both their sexual orientation and their HIV disease and it is possible that they may also experience more barriers to access and receipt of HIV support services. ^21,22 Other research has suggested that stigma can be an impediment to consistent health care-seeking behavior due to lack of disclosure of HIV and/or MSM status to social network members which can result in hidden and often substandard care-seeking behaviors. ^5,7,22,23 In fact, a recent study in LAC found that Latino MSM who report more MSM stigma were less likely to be retained in consistent HIV care. ⁷ Although HIV and MSM stigma were not measured in the current study, the findings point to the need for additional research to more fully understand why HIV-infected gay and bisexual men in LAC do not receive the support services that they need. Previous research in LAC suggests that stigma reduction programs could help Latino MSM in particular more openly pursue their HIV-related health care needs. ⁷

The finding that African American and low-income HIV-infected patients are more likely to have at least one service need is consistent with the national 1996 HCSUS data in which those with public or no insurance and those with lower incomes were more likely to need at least one supportive service and also have unmet oral health needs. ^1,24 Other research on residents of New York City found that African Americans were more likely to have inconsistent HIV care and low education and insurance status were associated with underutilization of HIV health care services. ^25,26 These data are not surprising and in fact the Ryan White system was set up specifically to provide services for low-income people and these data validate the profile of low income HIV-infected patients who continue to need Ryan White-funded supportive services. ⁹ The finding that African Americans are more likely to have at least one service need is also not surprising given what is known about the disproportional numbers of HIV-infected African Americans in LAC who are unemployed, have low-income, or are on public assistance and also have a high prevalence of substance abuse and mental health challenges. ^{11,17 –19}

The specific need for shelter/housing services for the current study group of HIV-infected African Americans older than 18 compared to whites and Latinos is consistent with other LAC data on 18- to 24 year-old HIV-infected African American men who have sex with men (MSM) who have been shown to be more likely than young Latino MSM to need housing services. ¹¹ Unstable housing for HIV-infected persons has been associated with difficulties remaining in consistent HIV care and poor physical health status. ^12,27 The greater need for case management services among African Americans compared to whites and Latinos is also supported by what is known about the low socioeconomic status and high prevalence of substance abuse and mental health challenges for African Americans with HIV and AIDS in LAC and is consistent with HCSUS data showing that non-whites were more likely to have contact with a case manager. ^1,11,18,19

The finding of a statistically greater need for meal and food services for HIV-infected Latino patients is consistent with an earlier study of a largely Latino (65%) HIV patient study group in LAC in which basic subsistence challenges related to nutrition/food support and housing were among the most commonly reported needed support services. ³ These data point to the continuing need for basic subsistence support for Latinos with HIV in Los Angeles.

The finding of less need for assistance finding dental services and HIV education services for HIV-infected whites compared to Latinos and African Americans is also consistent with demographic data for persons with HIV and AIDS in LAC in which whites tend to be more employed, with higher incomes and more educated. ¹⁸ In addition, in LAC white HIV-infected persons tend to be MSM who test for HIV earlier in the course of their infection, are less likely to have unrecognized infection, are more educated, have private health insurance, receive HIV care from private doctors and are thus more likely to have the resources to find dental services on their own. ^{17,28 –30}

The low study response rate is a limitation to the generalizability of these data. However the comparison to surveillance data and nonparticipants that shows African Americans and persons ages 18–29 as underrepresented and Latinos and those with an AIDS diagnosis as overrepresented, allows for an interpretation of the study data in light of the low response rate. Previous research has shown that Latinos and African Americans in LAC test late in the course of their HIV infection, therefore delaying HIV care which may mean that African Americans are under-represented in HIV care generally in LAC, not just MMP. ^15,29 The only inconsistency in this interpretation is that if this theory is true, Latinos should also be underrepresented in MMP which is not the case. Alternatively the obstacles to HIV care may be worse for HIV-infected African Americans than Latinos. In addition, some of the effect estimates in the logistic regression analyses that are presented separately by race/ethnicity have wide confidence intervals and should be interpreted with caution. Another limitation is that these data are not directly generalizable to the population using Ryan White funding or to those out of care but instead includes a sample of all persons in HIV care in Los Angeles County, including those treated by private and other non-Ryan White funded providers.

These data on the barriers, needs and unmet needs of a population in care for HIV in LAC can be used to reduce barriers to service utilization and also guide funding decisions for HIV support services. The findings point towards improving access and reducing barriers for certain subgroups including HIV-infected patients who are African American, low income, uninsured, and gay and bisexual men.