HIV prevalence

HIV prevalence has increased faster in the southern USA than in other areas, and persons living with HIV (PLWHIV) in the south are often rural, impoverished, or otherwise under-resourced. Studies of urban PLWHIV and those receiving medical care suggest that use of social services can enhance quality of life and some medical

outcomes, but little is known about patterns of social service utilization and need among rural southern PLWHIV. The AIDS Alabama needs assessment survey, conducted in 2007, sampled a diverse community cohort of 476 adult PLWHIV representative of the HIV-positive population in Alabama (66% male, 76% Black, and 26% less than high school education). We developed service utilization/need (SUN) scores for each of 14 social

services, and used regression models to determine demographic predictors of those most likely to need each service. We then conducted an exploratory factor analysis to determine whether certain services clustered together for the sample. Case management, assistance obtaining medical care, and financial assistance were most

commonly used or needed by respondents. Black respondents were more likely to have higher SUN scores for alcohol treatment and for assistance with employment, housing, food, financial, and pharmacy needs; respondents without spousal or partner relationships had higher SUN scores for substance use treatment.

Female respondents were more likely to have higher SUN scores for childcare assistance. Black respondents and unemployed respondents were more likely to have SUN scores in the highest quartile of the overall score distribution. Factor analysis yielded three main factors: basic needs, substance use treatment, and legal/medical

needs. These data provide important information about rural southern PLWHIV and their needs for ancillary services. They also suggest clusters of service needs that often occur among PLWHIV, which may help case managers and other service providers work proactively to identify important gaps in care.

Keywords: HIV; health services utilization; rural; south; need

Introduction

The prevalence of HIV infection has increased rapidly in the southern USA compared to the other areas of the country (Foster, 2007; Reif, Geonnotti, & Whetten, 2006) and southern states are among those with the highest AIDS-related death rates in the country (Reif, Geonnotti, et al., 2006; Whetten & Reif, 2006). For example, in 2006, Alabama had an age-adjusted HIV mortality rate of 4.2 per 100,000 persons, compared to 4.0 per 100,000 persons in the USA (Heron et al., 2009). Several issues in the south have been considered as possible contributors to the increased incidence of HIV and AIDS in the region, including high rates of sexually transmitted infections (STIs), racial disparities, the conservative culture of the south, poverty, and lack of access to health insurance and health care (Krawczyk, Funkhouser, Kilby, & Vermund, 2006; Luciano, 2002).

These factors create barriers to treatment for persons living with HIV (PLWHIV), but the major- ity of this research has been conducted in major

urban areas and these findings are not always

applicable to the southern population (Krawczyk

et al., 2006). For example, the HIV Cost and Service

Utilization Study (HCSUS) estimates that 46% of

PLWHIV in the USA have an annual house-

hold income of less than US$10,000, that 63% are

unemployed and 20% have no health insurance

(McKinney & Marconi, 2002). These estimates

are relatively consistent with observations of clinical

cohorts of PLWHIV, but may underestimate the

significant needs for social services among PLWHIV

in certain areas of the country, including the rural

south, where the proportion of uninsured adults

often approaches or exceeds 25%. A lack of economic and social resources compro-

mises a person’s ability to access health care services,

and research supports the finding that many southern

PLWHIV are disproportionally underresourced and

thus require considerably more assistance for every-

day activities and basic services, such as housing

assistance, emergency provisions, support groups,

*Corresponding author. Email: jfwalker@uams.edu

AIDS Care Vol. 23, No. 3, March 2011, 340�347

ISSN 0954-0121 print/ISSN 1360-0451 online

# 2011 Taylor & Francis DOI: 10.1080/09540121.2010.507743

http://www.informaworld.com

mental health-related counseling, legal assistance, and

financial assistance (Reif, Geonnotti, et al., 2006;

Reif, Whetten, Lowe, & Ostermann, 2006; Whetten &

Reif, 2006). Reif and colleagues studied infectious

disease clinics throughout the south, and found that

84.5% of the patients with HIV who were served by

these clinics needed at least one support service, with

47% reporting that at least one need was not met

(Reif, Whetten, et al., 2006). Given that, in urban samples of PLWHIV, lack

of access to social services is not only related to

general quality of life (Chin, Botsko, Behar, &

Finkelstein, 2009), but is also closely tied to medical

adherence, access to medical treatment, and health

outcomes (Messeri, Abramson, Aidala, Lee, & Lee,

2002; Reif, Whetten, et al., 2006), it is crucial to

understand more about the social service needs of

people living with HIV/AIDS in the south. Since a

sizable proportion of southern PLWHIV is not

engaged with a medical care provider (McKinney

& Marconi, 2002), surveys of social service needs

that are limited to samples of HIV-positive patients

or only urban PLWHIV may provide an incomplete

picture of the experience of PLWHIV in rural or

southern communities. Community-based surveys may provide a more

comprehensive description of the patterns of social

service utilization and need in this population. One

particularly valuable approach to reaching this com-

munity is through AIDS Service Organizations

(ASOs) which are staffed by individuals who are

well known and trusted by the community of

PLWHIV, and provide social services that are other-

wise limited or unavailable (Tolle, 2009). Under-

standing the patterns of service needs and utilization

among these community populations may be helpful

to providers and case managers in working proac-

tively with their clients. AIDS Alabama is an ASO located in Birmingham,

Alabama that has been funded since 1988 by HRSA,

CDC, HUD, state, local, and private resources. It

serves a catchment area that includes approximately

16,000 PLWHIV, and coordinates with other ASOs

throughout the state. To understand the priority

needs of its client population, AIDS Alabama con-

ducted statewide surveys of PLWHIV in 1999, 2003,

and 2007. We examined the 2007 AIDS Alabama Needs

Assessment Survey data to explore patterns of service

utilization and needs, expecting that a high propor-

tion of respondents would report utilization of or

need for services across multiple domains, and that

several services would ‘‘cluster’’ together.

Methods

Participants

AIDS Alabama and partner agencies recruited PLWHIV to be as representative as possible of the adult PLWHIV population, in terms of geographical distribution and racial and gender balance; partici- pants were recruited from most Alabama counties. However, in some instances, certain counties re- mained underrepresented in the sample due to a low known HIV-positive population. Attempts were made to include individuals who both were and were not well-connected to health care, by working through ASOs, case managers, and other HIV- positive individuals, who were able to identify and refer sporadic users of ASO services and those not receiving medical care.

Measures

This survey was created specifically to assess the needs and concerns of the adult PLWHIV population for the use of AIDS Alabama staff in program planning, and as such was not originally developed for research purposes. It was jointly developed by AIDS Alabama and Columbia University School of Public Health faculty in 1999. The survey content was slightly edited by consultants at the University of Alabama at Birmingham for 2003 and again for 2007 but similar collection methods were used in all survey administrations. In addition to standardized mea- sures on aspects of living with HIV, the interview covered demographics; basic needs such as food, transportation, and housing; and a wide range of other ancillary services. A total of 14 basic and ancillary service needs were included: financial, legal, pharmaceutical, employment, and housing assistance; substance use, mental health, and alcohol treatment; medical services, transportation, dental care, case management, food, and childcare. Respon- dents were asked to indicate whether they had received services or had needed assistance in the past 6 months in each of these areas.

Procedures

Survey questionnaires were administered by trained interviewers who were HIV-positive Alabamians. Interviewers traveled to meet and conduct the survey at a convenient location for the participant, usually at an ASO or HIV clinic office or meeting room where privacy could be assured. All information collected was confidential, assigned a unique identification number, and free of identifying personal information. Some interviews were video- or audio-taped with the

AIDS Care 341

participant’s written permission. Each recording was securely stored without personal information. The interview took approximately one hour to administer. Participants did receive US$15 compensation for their time, travel, and effort.

Results

Data analysis approach

The 2007 survey was administered to 525 community- dwelling individuals. A total of 29 individuals who identified with a racial/ethnic group other than Black or White or had missing data for racial/ethnic group and 20 individuals who identified themselves as transgender or had missing gender data were ex- cluded for comparison purposes, which resulted in 476 individuals included in subsequent analyses. Descriptive analyses were completed to obtain the characteristics of this sample.

Binary service utilization/need (SUN) variables were created for each of the services included in the survey. If the respondent indicated either that s/he was receiving services of a specific type or had needed that service then a value of 1 was assigned to the service/ utilization variable; if the respondent indicated that they had not received the service and had not needed it, then a value of 0 was assigned. This combined SUN variable approach reflected the overall demand for each service by including those who have already gained access to the service as well as those who needed the service but were not able to obtain it.

Descriptive univariate analyses were completed to determine the number and percentage of respondents who reported using or needing each service, and demographic characteristics were compared for those who had an individual SUN variable score and those who had missing data for that variable. No significant differences were noted for any of these comparative analyses. Subsequently, multivariate linear regression analyses described the characteristics of individuals most likely to need or use each service. Logistic regression models included each SUN as the outcome variable. Predictor variables in each model included: race, gender, age, relationship status, education, having children in the home, and employment status.

Next, values for the individual SUN variables were summed to create an overall SUN score for each respondent. A total of 367 respondents had complete data for each of the SUN variables and thus a calculated overall SUN score. Descriptive univariate analyses were completed to characterize the pattern of overall SUN scores; again, demographic character- istics were compared for those with and without an overall score. No substantial differences between

groups were identified. Multivariate linear regression analyses described the characteristics of individuals who were evidenced the highest and lowest SUN scores (defined as those respondents in the highest and lowest quartiles of overall SUN scores, respec- tively). The outcome variable was the overall SUN score and predictor variables were included as described above.

Then SUN variables were subjected to an explora- tory factor analysis using squared multiple correla- tions as prior communality estimates. The principal factor method was used to extract the factors. We used both oblique and orthogonal rotation of factors; because the results were essentially identical, the orthogonal rotation results are reported. A three- factor solution was hypothesized; factor selection was based on eigenvalues (]1.0), scree plot, and inter- pretability were considered in confirming factors. Items were considered to load on a factor if the rotated factor loading was greater than or equal to 0.35 on one factor and less than 0.35 on other factors.

Demographics of the sample

The characteristics of the analysis sample are sum- marized in Table 1. The majority of survey respon- dents were black males between the ages of 40 and 64 years. The average age was 42.5 years (standard deviation (SD), 9.84). The youngest was 18 years and the oldest 76 years of age. A majority were single or never married. Approximately, one in four respon- dents had children in the home, one in three had some college education, and one in four reported having less than a high school diploma. Just fewer than 70% reported being employed either full or part time.

Frequency of service utilization/need (SUN)

Services used or needed by the survey respondents are summarized in Table 2. The most commonly used or needed services included case management, medical assistance, and financial counseling. The least needed or used services were childcare, legal assistance, alcohol treatment, and employment assis- tance. Approximately, half of the respondents in- dicated that they either needed or were receiving mental health services, transportation assistance, and housing assistance.

Demographic group associations with service utilization/need (SUN)

Table 3 summarizes findings regarding the demo- graphic subgroups most likely to report using or needing services, by type of service. Regression

342 K.E. Stewart et al.

analyses indicated no significant differences among

groups in their likelihood of needing or using medical

or legal assistance. Black respondents were more

likely to be users of alcohol treatment, employment,

financial, food, housing, and pharmacy assistance

than were their white counterparts. Individuals not

living with a partner or a spouse were more likely to

be using or need substance use treatment services

than respondents in partnered relationships. Women

were more likely than men to be users of childcare

assistance, as were individuals with children in the

home, compared to individuals without children in

the home. Furthermore, compared to respondents

who were employed, respondents who were not

working were more likely to be users of a number

of services, including: case management, mental