Barriers to HIV Care for Women of Color Living in the Southeastern US Are Associated with Physical Symptoms,Social Environment,and Self-Determination

Introduction

I n 2009, women represented 24% of all HIV diagnoses in the United States. ¹ Of women diagnosed with HIV, African-American (AA) women were the largest proportion of those infected and accounted for two-thirds of new AIDS diagnoses in 2010, ² while Hispanic women were 4.2 times more likely to be infected than White women from 2007 to 2010. ³ HIV-infected women belonging to racial and ethnic minorities (women of color, WOC) are more likely to report infrequent antiretroviral therapy (ART) use, ⁴ to be socio-economically disadvantaged, miss medical appointments, ⁵ and have late access to, and more frequent discontinuations of ART ^{6 –8} than other women. Related work suggests unique trajectories for the process of engaging and remaining in HIV care among WOC compared with other populations. ^{9 –11} The process by which people living with HIV enter and are retained in HIV medical care is influenced by a variety of factors that include medical, social-environmental, and personal characteristics.

Medical characteristics, like disease severity and overall physical health, are associated with care-seeking behavior and service utilization. Poor health is associated with increased risk of hospitalization among HIV-positive patients in previous studies. ¹² Patients with low CD4 counts are at greater risk of more outpatient visits and hospitalizations than patients with higher CD4 counts. ¹³ In a review of the literature, ⁸ illness severity, physical symptoms, and health related quality of life, have all been shown to be associated with increased health care usage.

Besides physical symptoms, the social environment is another influence on HIV care engagement, as both supportive and abusive relationships impact health service utilization among chronically ill patients. ¹⁴ Intimate partner abuse towards HIV-infected women may constitute a barrier to outpatient HIV care, particularly if the HIV-infected woman is dependent on the abuser for affection, money, or transportation. ^15,16 The presence of general social support from formal and informal networks has been shown to contribute to engagement in HIV care, particularly among AAs. ^{17 –19} Social expectation, norms, and pressure from one's peer network may facilitate engaging in HIV care. ²⁰ Psychological aspects of the patient, such as self-determination, may also play an important role in engaging in HIV care among WOC.

The concept of self-determination, which can be characterized as a personal characteristic, is used to help understand and predict the adoption of a variety of health related behaviors, including smoking cessation, exercise, and diabetes care. ^{21 –24} According to self-determination theory, the inherent psychological needs for competence, relatedness, and autonomy must be fulfilled for the adoption of new behaviors to be internally-motivated and sustained. ²⁵ Limited use has been made of this theory for understanding HIV care behaviors, but autonomy support, autonomy motivation, and competency have been correlated with ART adherence. ²¹ Thus, higher levels of self-determination, via competency, relatedness, and autonomy, could predict fewer perceived barriers to accessing HIV care among WOC.

There are also other potential barriers and facilitators to accessing HIV care, such as demographic characteristics, medical and psychiatric co-morbidities. Yet, information regarding the extent to which any of these patient-level factors is associated with logistical and environmental barriers-to-care has not been analyzed for HIV-infected WOC. In light of this limitation and prior literature, the research reported here will answer the following questions: (1) To what extent do HIV-infected WOC report logistical and social-environmental barriers to care? (2) Are reports of barriers to care associated with demographic characteristics, mental or physical co-morbidities? (3) Are positive social environments or higher levels of self-determination characteristics associated with fewer barriers-to-care?

Methods

Results

Discussion

In the context of better understanding the unique gender-specific needs of HIV-positive women, ³⁵ this study highlights the perceived psychological and environmental barriers and facilitators to HIV care among WOC. The data suggest that a high number of barriers are present when HIV-infected women of color attempt to access HIV care. Most women reported four of 11 possible barriers, with community stigma and financial issues being most commonly cited. Access to mental health treatment and support were the two least frequently cited barriers but presented some difficulty to 18–25% of the women. The remaining seven barriers were cited by 27–56% of women. The cumulative experience of accessing healthcare by the women interviewed is one of multiple barriers.

Co-morbidities of all types were consistently associated with barriers to care. Women with poorer physical health, moderate to severe levels of depression, anxiety or higher levels of psychological abuse were more likely to have five-or-more barriers. This is consistent with the literature supporting the effects of intimate partner abuse on delayed access to HIV care and missed appointments ¹⁷ and with other studies illustrating the relationship between mental health and service utilization. ^36,37 Additionally, poorer physical health was associated with a greater risk for more barriers to care. Treatment-specific and general social supports were associated with fewer reported barriers, though the association with treatment specific support was not significant in the adjusted model. Additionally, women reporting high levels of autonomy, competence, and relatedness reported fewer barriers.

In this analysis of barriers-to-care reported by HIV-infected women, the theory of self-determination was used to expand the literature of engagement and retention in HIV care. All three basic needs—relatedness, competency, and autonomy—were associated with fewer barriers to care. Previous SDT-related research has focused on ART adherence, and show that autonomy support and/or autonomous motivation are associated with higher levels of adherence, and a greater sense of well-being. ^21,38,39 Men with higher levels of adherence were more likely to report that “taking medications for others” was part of an intrinsic value system whereas men with lower adherence reported this as an external obligation. ⁴⁰ Autonomy with motivation based on personal values was associated with higher levels of well-being in persons adjusting to a new diagnosis of HIV. ³⁹ This quantitative work complements previous qualitative findings showing that autonomy, competency, and relatedness influence adherence to HIV medical care. ¹¹

Severity of physical symptoms was associated with multiple barriers-to-care in our sample. This is in contrast to earlier work reporting that health service utilization, including outpatient care, increased with increased severity of illness. ^8,13 Gender, urban versus rural environments, age, or other factors are possible explanations for the differences between our findings and previous work. These studies examined actual use of care and here we are reporting perceived barriers. Indeed, in two studies with rural HIV-infected women, women described physical symptoms as barriers to care. We reported earlier that women in our mixed geographic population describe missing clinic appointments when they feel sick or have medication side effects. ¹⁰ Others reported that California rural women (50% white, 29% Latina, 15% African-American) reported more missed appointments in the presence of physical symptoms. ⁴¹ Not accessing outpatient care increases the use of hospital services. ¹³ Chronically ill women experiencing multiple barriers to care may be at risk for poor health and subsequent hospitalizations or emergency room visits. Poor health, in turn, may increase barriers to care. Severity of illness influences utilization of services in a complex manner depending on the setting, population, and service type.

Treatment-specific social support positively influences engagement in HIV care, especially among AAs. ^{17 –19} However, this effect was not observed among our sample. Rather, general social support in the form of emotional and practical support was associated with fewer reported barriers to care, confirming that social support may buffer barriers to care. ^{17 –19,42} The difference in these two findings may be explained by the heterogeneous effects of social networks on access to care. In a prior study on HIV-positive African-American women, ¹⁹ social support from other women, and emotional support, were associated with outpatient service use, whereas instrumental support in the form of direct financial and physical assistance was not. A recent pilot study suggests that peer-support among HIV positive African-American women reduces stigma ⁴³ and may have implications for engagement in HIV care. Treatment-specific social support is more instrumental and may not represent the type of social support that is needed to address barriers to care.

These findings, while promising, are not without limitations. Twenty-three respondents completed only part of the Barriers to Care Scale and there were women who declined to participate in the data collection. While the resulting sample of 141 women represents one-third of the clinic's population of minority women, and have similar demographics, respondents may differ from non-respondents in meaningful ways. Responses to the Barriers to Scale represent perceived barriers to HIV care. Further study is needed to capture the direct experience of barriers to care among this population. Additionally, our dependent variable is specified as a dichotomous indicator of multiple barriers to care, which may limit the interpretability of our findings. Since our sample included only WOC of one geographic area, the results may not be generalizable to other populations. Our use of validated instruments, however, facilitates comparison with other populations. Further, the distribution of responses to the covariate items suggests representation of broad range of experiences and perspectives, increasing the generalizability of the findings. Finally, these data are cross-sectional, which means the causal direction of the relationships between barriers to care and our independent variables cannot be determined.

HIV-infected women attending a medical clinic report the presence of multiple barriers to accessing HIV care. These findings have important policy implications. Since the sample was drawn from women already accessing care, the results may understate the effect of these predictors on barriers to HIV care for non-engaged women. Furthermore, as described above, a cycle between poor physical health and increases in barriers to care may play a role for these women. Thus, engaging and tracking women at the point of care is important. There is some evidence that implementing rapid HIV testing in emergency rooms can further our understanding of the unique characteristics of women of color not in HIV care, ⁴⁴ while potentially addressing perceived barriers to care. Interventions that target improved self-determination and social support among HIV positive women of color may also help improve access and retention into HIV care.