PrEParing women to prevent HIV: An integrated theoretical framework to PrEP Black women in the United States

Abstract

Pre-exposure prophylaxis (PrEP) with optimal adherence has demonstrated efficacy in reducing HIV incidence in women. Black women are disproportionately burdened by the HIV epidemic, accounting for more than half of all new HIV cases in women, thereby making PrEP an ideal prevention strategy for this group. However, to date, PrEP uptake by women in the United States has been slow. Further domestic research is needed to understand the multilevel factors related to PrEP awareness, uptake, and implementation in Black women. Our purpose was to review the current status of HIV prevention in Black women. We summarize clinical trials germane to federal approval of PrEP, discuss important PrEP studies focused on women including non-oral options, and review multilevel barriers to PrEP uptake. Lastly, we discuss the use of an integrated theoretical framework to organize multilevel factors related to PrEP uptake by Black women in order to guide intervention development.

An estimated 40,000 people were diagnosed with HIV in the United States (Centers for Disease Control and Prevention [CDC], 2015). Advances in antiretroviral therapy (ART), as well as behavioral (e.g., risk reduction counseling, condom use, and needle exchange programs) and biomedical (e.g., pre-exposure prophylaxis [PrEP]) interventions have played a role in the recent decline in HIV incidence rates. Despite these advances, racial and gender disparities have persisted. Black men and women account for less than 12% of the U.S. population, but continue to account for nearly half of all new HIV cases (CDC, 2017a). Furthermore, Black men who have sex with men (MSM) account for the highest number of new HIV diagnoses. In fact, it is estimated that half of all Black MSM will be diagnosed with HIV in their lifetimes (CDC, 2016a). Many Black men do not self-identify as MSM (gay, bisexual) for fear of interpersonal and structural-level stigma, discrimination, and homophobia, thereby placing Black women at increased risk of acquiring HIV (Paxton, Williams, Bolden, Guzman, & Harawa, 2013; Tieu et al., 2012). The high lifetime risk of HIV acquisition for Black MSM has led to an increase in targeted prevention efforts; however, in doing so, Black women’s risks for acquiring HIV and the need for effective prevention strategies has been largely overshadowed. Therefore, effective prevention efforts must not only target Black MSM but must also target Black women, who are at substantially increased risk for HIV.

Black women continue to be burdened by the HIV epidemic accounting for approximately 60% of new HIV cases among women and for the highest HIV incidence rate attributed to heterosexual transmission among all new HIV diagnoses in the United States (CDC, 2017b). The most common mode of HIV transmission for women is heterosexual contact (87% overall, 92% for Black women, 88% for Hispanic women, and 71% for White women; CDC, 2017b). Several multilevel behavioral and structural risk factors contribute to women’s risks for HIV acquisition, including sex with high-risk contacts such as MSM and male partners of unknown risk/status, transactional sex, mass incarceration of Black men, racial discrimination and gender inequality, traumatic life experiences, depression, intimate partner violence (IPV), and sexual coercion (CDC, 2014; Cooper et al., 2015; Garfinkel, Alexander, McDonald-Mosley, Willie, & Decker, 2017; Klein, Elifson, & Sterk, 2008; Rosenthal et al., 2014). These data highlight the need for woman-centered approaches to HIV prevention strategies, such as PrEP.

PrEP is a promising, contemporary biomedical strategy for HIV prevention commonly offered to MSM and the uninfected partners of people living with HIV (i.e., serodiscordant partners). Consistent use of PrEP has been shown to be up to 90% effective in reducing HIV acquisition, according to several studies published since 2010 (Baeten et al., 2014; CDC, 2011; Donnell et al., 2014; Heffron et al., 2014; Murnane et al., 2014) with efficacy ranging from as high as 91% in heterosexual men and women (Donnell et al., 2014), and up to 85% for studies done exclusively in women (Heffron et al., 2014; Murnane et al., 2013). Low efficacy observed in studies done primarily in women may have been due to biologic factors including pharmacokinetic properties and/or behavioral factors, such as suboptimal medication adherence (Hanscom et al., 2016). In one such study, low adherence to oral PrEP among African women was noted with only 12% of participants having tenofovir disoproxil fumarate (TDF) drug concentration levels consistent with high adherence throughout the study (Corneli et al., 2014). However, PrEP has been shown to be highly effective in women when adherence is optimal; therefore, we cannot eliminate this biomedical option from the prevention armamentarium.

To date, uptake of PrEP among U.S. women has been very low, despite almost 500,000 U.S. women having an indication for PrEP based on risk (Wu et al., 2016). PrEP implementation efforts will require researchers and clinicians to solicit women’s knowledge, attitudes, beliefs, and preferences in the context of other individual, interpersonal, environmental, and structural influences to identify salient areas for targeted intervention development. Therefore, our purpose was to (a) review the current status of HIV prevention needs and (b) to present an overarching integrated theoretical framework to organize multilevel factors related to PrEP uptake to facilitate the development of evidence-based and theory-informed strategies to promote awareness and individual-level information, motivation, and behavioral skills to improve PrEP uptake by Black women in the United States.

Prep and HIV Prevention

Co-formulated TDF/FTC (FTC= emtricitabine), antiretroviral (ARV) medications used in combination with other ARVs to treat HIV, is currently the only medication approved by the Food and Drug Administration for use as PrEP. PrEP is recommended for use by uninfected individuals in combination with other evidence-based behavioral risk-reduction tools such as condom use to prevent the acquisition of HIV (Gilead Science, 2017; U.S. Food and Drug Administration, 2015). Approval of PrEP was predominantly based on data from two key clinical trials, the Pre-exposure Prophylaxis Initiative (iPrEx) and Partners PrEP trials (U.S. Food and Drug Administration, 2015). iPrEx was a randomized, double blind multinational (Peru, Ecuador, South Africa, Brazil, Thailand, and the United States) trial that evaluated the safety and efficacy of once-daily oral TDF/FTC versus placebo for the prevention of HIV acquisition in 2,499 MSM including transgender women (N = 29; Grant et al., 2010). There was a 44% relative reduction in HIV acquisition; however, the relative reduction in HIV acquisition improved to 92% for individuals with detectable study drug levels when compared to individuals without a detectable level (Grant et al., 2010). Lower efficacy than what was originally anticipated was attributable to suboptimal study-drug adherence. In 2008, recruitment begin for the Partners PrEP trial, another randomized, double-blind, three-arm, multisite (2 West African Countries) trial that evaluated the efficacy of once-daily TDF/FTC (n= 1579 couples), TDF (n= 1584 couples), and placebo (n= 1584 couples) in 4,747 serodiscordant couples of which more than50% were male (Baeten et al., 2012). There was a 62% relative reduction in HIV in the TDF arm and a 75% relative reduction in HIV in the TDF/FTC arm (Baeten et al., 2012). The relative reduction in HIV acquisition improved to 86% (TDF) and 90% (TDF/FTC) with detectable study-drug level when compared to individuals without a detectable level (Baeten et al., 2012). As evidenced by these studies, optimal adherence to PrEP was essential for prevention of HIV.

More recently, the PROUD (PRe-exposure Option for reducing HIV in the UK: an open-label randomization to immediate or Deferred daily Truvada for HIV negative gay men) and IPERGAY (the Intervention Préventive de l’Exposition aux Risques avec et pour les Gays or Action to Prevent Risk Exposure by and for Gay Men) trials provided real-world clinical data regarding the effectiveness of PrEP (Marrazzo, 2017; McCormack et al., 2016; Molina et al., 2015). The PROUD trial was an open-label randomized, multisite study in London, England that assessed the efficacy of PrEP in routine clinical practice as well as the influence of PrEP on risk reduction behaviors such as condom use, number of sexual partners, and prevalence of sexually transmitted infections. The study randomized MSM to receive TDF/FTC immediately (N = 275; immediate group) or defer PrEP for 12 months (N = 269; deferred group). Due to the increased risk of HIV in the deferred group (N = 3 in the immediate group and N = 20 in the deferred group), the data monitoring and safety board requested that all participants receive PrEP, which corresponded to a relative reduction in HIV acquisition of 86% (McCormack et al., 2016).

The IPERGAY trial randomized 400 HIV-uninfected MSM in France and Canada who reported having unprotected anal intercourse with at least 2 partners during the 6-month period prior to the study to receive TDF/FTC (n = 199) versus placebo (n= 201) to assess the efficacy and safety of intermittent or “on-demand” PrEP: 2 pills 2 to 24 hours before sex and then 1 pill 24 and 48 hours after sex, for a total of 4 pills (Molina et al., 2015). Similar to the PROUD trial, there was an 86% relative reduction in HIV acquisition in the TDF/FTC group, with 2 incident cases of HIV in the TDF/FTC group and 14 in the placebo group (Molina et al., 2015). The IPERGAY investigators were cautious about interpretation of the results, indicating that study results cannot be generalized to individuals taking less than 15 pills per month (Molina et al., 2015). Recent data presented at the 9th International AIDS Conference, however, concluded equivalent efficacy of on-demand PrEP in men who used fewer than15 pills per month (Antoni et al., 2017). These trials provided evidence that PrEP could be successfully implemented in routine, clinical settings and further evidence that PrEP was a highly effective, biomedical HIV prevention tool in MSM and male serodiscordant partners of people living with HIV. However, data regarding the utilization of PrEP by women was limited.

Key PrEP Studies Focused only on Women

While numerous trials in MSM and serodiscordant partners (men and women) demonstrated efficacy for prevention of HIV acquisition, studies exclusively enrolling women have not demonstrated similar efficacy of oral PrEP (Table 1). In the Pre-exposure Prophylaxis Trial for HIV Prevention among African Women (FEM-PrEP), 2,120 women were recruited from four sites in Kenya, South Africa, and Tanzania, and randomized to receive TDF/FTC or placebo. The study was concluded early due to low efficacy, subsequently determined to have been due to poor adherence (<40% of women enrolled had evidence of recent study drug use based on drug blood tests; Van Damme et al., 2012). Further analysis of the FEM-PrEP study suggested that less than 12% of the women had adequate drug levels congruent with high adherence throughout the study (Corneli et al., 2014). Similar to FEM-PrEP, the Vaginal and Oral Interventions to Control the Epidemic (VOICE) trial was closed early at interim analysis due to lack of efficacy. It was determined that suboptimal adherence drove these equivocal outcomes with study drug detected in only 30% of women assigned to oral tenofovir (TFV), 29% of those assigned to oral TDF/FTC, and 25% of women assigned to intravaginal TFV gel condition (Marrazzo et al., 2015). Lack of perceived social support, lack of perceived susceptibility to HIV acquisition, skepticism about the use of ARV for HIV prevention, researcher mistrust, and fear of HIV-related stigma were cited as reasons for suboptimal adherence to TFV-based study drugs (van der Straten et al., 2014). These early results prompted considerable attention to the need for biomedical prevention strategies that were alternatives to oral PrEP.

Table 1.

PrEP Trials with Efficacy Data for Women

Trial/project	Sample size	Location(s)	Study drug(s)	Efficacy	Adherence
Partners PrEP	4,758 couples; 1,164 women: 598 randomized to TDF arm and 566 randomized to TDF/FTC arm	Kenya Uganda	Daily Oral TDF or TDF/FTC	TDF: 71% (95% CI = 37% – 87%) TDF/FTC: 66% (95% CI = 28% – 84%)	Detectable plasma drug level 31% in women who acquired HIV compared to 82% in women who did not acquire HIV
				Efficacy with detectable study drug level: TDF: 86% (95% CI = 57% – 95%) TDF/FTC: 90% (95% CI = 56% −98%)
Fem-PrEP	2,102; 1,062 randomized to TDF/FTC arm	Kenya South Africa	Daily Oral TDF	6% (95% CI = 52% – 41%)	Detectable plasma drug level <40%
				Study stopped early due to lack of efficacy
TDF-2	1,219; 362 women: 331 randomized to TDF/FTC arm & 331 to placebo arm	Botswana	Daily oral TDF/FTC	49% (95% CI = 22% – 81%)
				Efficacy with detectable study drug level: 76% (95% CI = 24% – 94%)
VOICE	5,029; randomized to TDF = 1,007; TDF/FTC = 1,003; &TFV gel = 1,007	South Africa Uganda Zimbabwe	Daily Oral TDF or TDF/FTC or Intravaginal TFV 1% gel	TDF: −49% (95% CI = −129% – 3%) TDF/FTC: −4% (95% CI = −50% – 27%) TFV: 15% (95% CI = −21% – 39%)	Detectable plasma drug level 30% in the TDF arm & 29% in TDF/FTC arm. 49% detectable drug level in vaginal sample
				Efficacy with detectable study drug level: TDF/FTC: −25% (95% CI = −1.58% −39%) TDF: 34% (95% CI = −44% −69%) 1% TFV gel: 66% (95% CI = 13% −87%)
CAPRISA 004	889; 445 randomized to TFV gel	South Africa	Intravaginal TFV 1% gel	39% (95% CI = 6% to 60%)	An estimated 40% of women had <50% gel adherence; efficacy was 38% when adherence was 50% −80%.
				Efficacy with detectable study drug level (>80% gel adherence): 54% (95% CI 6% to 80%)
FACTS-001	2,059	South Africa	Intravaginal TFV 1% gel	0% (95% CI = −40% −30%)	50–60% based on returned applicators. Detectable drug level in 22% of women
				Efficacy with detectable study drug level: 52% (95% CI 3% to 77%)
The RING study	1,959; 1,307 randomized to DPV intravaginal ring	South Africa Uganda	Monthly DPV intravaginal ring	31% (95% CI = 1%−51%)	Detectable plasma drug level in 84% of women.
				Efficacy with detectable study drug level: 29% (95% CI = −22% −58%)
ASPIRE	2,629; 1,313 randomized to DPV intravaginal ring	Malawi South Africa Uganda Zimbabwe	Monthly DPV intravaginal ring	27% (95% CI = 1%−46%); 37% (95% CI = 12%−56%) after extracting 2 sites with low adherence	Detectable plasma drug level in 82% of women. Lower adherence was seen in women 18 −21 years of age.
				Efficacy with detectable study drug level: 37% (95% CI = 12% −56%)

Note. TDF = tenofovir disoproxil fumarate; FTC = emtricitabine; TFV = tenofovir; DPV = dapivrine; CI = confidence Interval.

Non-Oral Options for Women

In addition to oral PrEP, alternative formulations of PrEP and methods of drug delivery have been explored. In addition to the VOICE trial, other trials have evaluated the efficacy of intravaginal TFV gel for PrEP delivery with varying efficacy. TFV gel was determined to be effective in preventing HIV acquisition in CAPRISA-004 (The Centre for the AIDS Program of Research in South Africa; Abdool Karim et al., 2010); however, TFV 1% intravaginal gel was not effective in reducing HIV acquisition in the FACTS-001 (Follow-on African Consortium for Tenofovir Studies), which, again, was attributed to suboptimal adherence (Hess et al., 2015).

In the CAPRISA-004 study, rural and urban South African women (n = 889) were randomly assigned to receive TFV 1% gel (n= 445) or placebo gel (n = 444). The gel was intended for on-demand insertion with women inserting one dose of gel 12 hours before sex and a second dose within 12 hours after sex, not to exceed two doses in a 24-hour period. At 30 months, new HIV cases were reported in the TFV arm (n = 38) and placebo arm (n = 60), with a 39% effective rate. However, in sub-analyses a 54% reduction in HIV acquisition was seen in high gel adherers (> 80%; Abdool Karim et al., 2010). The FACTS 001 study expanded the CAPRISA-004 study by increasing the sample size (n = 2,059) across South Africa. In that study, TFV gel was not effective in preventing HIV acquisition and low adherence was noted to be a potential contributing factor. In fact, only 1 in 5 women were high adherers (>80% use of gel with sex; Hess, 2015; Hess et al., 2015).

In addition to TFV-containing intravaginal gel, two trials assessed the efficacy and safety of another PrEP delivery system, intravaginal rings containing dapivirine (an ARV). The RING and the ASPIRE (A Study to Prevent Infection with a Ring for Extended Use) trials each had a reduction in HIV acquisition for women of 31% and 27%, respectively (Baeten et al., 2016; Nel et al., 2016). In the ASPIRE study, 2,629 women were recruited from 15 research sites in Malawi, South Africa, Uganda, and Zimbabwe. Women were randomly assigned to receive a dapivirine vaginal ring (n = 1313) or a placebo vaginal ring (n = 1316). Vaginal rings were worn continuously and replaced every 4 weeks. Incident HIV cases occurred in the dapivirine group (n = 71) and the placebo group (n = 97), with a 27% reduction in HIV acquisition in those receiving the PrEP containing intravaginal ring. After excluding data from two sites because of poor protocol and product adherence, the relative HIV reduction improved to 37% in the dapivirine group when compared to the placebo group. Furthermore, in a subgroup analysis based on two age groups (<25 years versus ≥25 years), efficacy was higher in women 25 years and older (61%) than women younger than 25 years (10%). In subsequent post hoc analysis, efficacy against HIV acquisition was even lower for women 18 to 21 years (27%) when compared to women older than 21 years (56%; Baeten et al., 2016).

The RING study had similar findings. For this study, women were recruited from seven research centers in South Africa and Uganda and randomly assigned to receive either a dapivirine vaginal ring or a placebo vaginal ring. Vaginal rings were worn continuously and replaced every 4 weeks. A 31% reduction in HIV acquisition was seen in the dapivirine group when compared to the placebo group (Nel et al., 2016). While a difference by age existed in the ASPIRE study data, in a subsequent analysis based on two age groups (≤21 years versus >21 years), no significant difference in efficacy was found in the RING study (Nel et al., 2016).

PrEP Safety

Available data have supported the safety of TFV and dapivirine-based PrEP. There were no statistically significant differences in incidences of death, serious adverse events, or laboratory abnormalities (i.e., renal and liver dysfunction) in the experimental groups when compared to control groups and reports of adverse effects were low (Baeten et al., 2016; Marrazzo et al., 2015; Nel et al., 2016; Van Damme et al., 2012). The study medications were well tolerated, with nausea being more commonly reported in participants taking oral TFV-based study medication; however, the symptom usually resolved after the first month of administration (Baeten et al., 2012; Van Damme et al., 2012). Decreased bone mineral density (BMD) is a concern with the use of TFV-based medications. In the iPrEx trial BMD decreased by about 1% in the hip and 2% in the spine, but BMD returned to baseline within 6 months after discontinuing the medication (Grant et al., 2016). There were no differences in trauma-related bone fractures in the groups receiving TFV-based study medication compared to placebo, and no pathologic fracture occurred (Baeten et al., 2012; Grant et al., 2010; Marrazzo et al., 2015). TFV-based PrEP has been found to be safe during pregnancy and to impact hormonal contraception (Mugo et al., 2014; Murnane et al., 2014). Although safe, safety data has been sparse for women, warranting more longitudinal studies in diverse samples of women.

PrEP State of the Science Summary

The data discussed above provide evidence that PrEP is safe and can prevent HIV acquisition by men and women, but suboptimal adherence to oral PrEP in clinical trials in women may have been particularly consequential to negate prevention outcomes. As PrEP rolls out in different communities, added efforts and careful attention to promote PrEP awareness and use among women at risk for HIV are essential. To improve PrEP uptake by Black women in the United States in particular, various multilevel factors associated with PrEP uptake in this marginalized population must be considered.

Multi-Level Correlates of PrEP Uptake in Women

Clinical trial data suggest that several factors, including suboptimal adherence may become challenges to PrEP effectiveness for women. Individual-level barriers to PrEP uptake have been shown to include insufficient or negative social support, low perception of HIV risk, confusion about ARV as prevention, forgetfulness, fear of side effects, and pill characteristics (Corneli et al., 2016; van der Straten et al., 2014). Several interpersonal factors were identified as barriers to PrEP uptake, including healthcare provider stigma and gender norms in relationships, which allow men to make decisions regarding women’s reproductive health (Montgomery et al., 2015; van der Straten et al., 2014). Stigma was a salient barrier to PrEP uptake. In addition, women feared being labeled as having HIV infection, isolated from their families and communities, and the target of discriminatory behavior from friends, family, male partners, and community members (Corneli et al., 2016; Montgomery et al., 2015; van der Straten et al., 2014). However, these barriers were all reported from non-U.S. clinical trial data. PrEP uptake data is limited in the United States, with most studies assessing the potential for women to adopt PrEP with little evaluation of individual and interpersonal-level barriers for PrEP uptake. Black women, in particular, remain under-represented in PrEP research and PrEP dissemination in real-world settings despite disproportionately higher risks for HIV acquisition.

Gauging Awareness among U.S. Black Women

A limited number of U.S studies, mostly qualitative, have evaluated potential barriers for PrEP uptake by women. Awareness of PrEP was found to be a critical challenge for Black women in the United States. In an exploratory qualitative study, women were asked about their attitudes, knowledge, and likelihood of using PrEP. Some women were angry that they had not heard of PrEP, but, once aware, many were interested in and willing to use PrEP (Auerbach, Kinsky, Brown, & Charles, 2015). Several trusted sources for information have been suggested to boost PrEP awareness in Black women such as media, including social media and more traditional media such as commercials, billboards, and magazine ads; healthcare provider delivery; and peer recommendations and sharing of personal experiences with PrEP use (Auerbach et al., 2015). Once informed about PrEP, Black women were more likely to report willingness to adopt PrEP regardless of sexual risk, when compared to their White counterparts (Wingood et al., 2013). More specifically, Black women were more likely to report willingness to adopt PrEP if recommended by a friend or healthcare provider (Auerbach et al., 2015; Wingood et al., 2013).

Lack of perceived risk for HIV was a barrier, especially for women with higher levels of education (Wingood et al., 2013). However, more than half of the women who seroconverted during the FEM-PrEP trial reported that they were not at risk. Therefore, HIV prevention strategies are needed to boost knowledge of HIV risk and indications for PrEP use in those who are at risk for HIV acquisition (Corneli et al., 2014). While some women are unaware or underestimate their risks, other women are aware of their risks but are unable to protect themselves against HIV because of power dynamics in relationships, which influences condom negotiation and consensual sex (Hartmann et al., 2016; Montgomery et al., 2015). Black women in the United States also reported stigma as a barrier for willingness to use PrEP, fearing being labeled as having HIV and isolated by sexual partners and family members, as well as, coming to terms with their personal perceptions of people at risk for HIV (Auerbach et al., 2015). Other barriers to potential PrEP uptake include cost of the medication, medical mistrust, and housing instability (Auerbach et al., 2015; Wingood et al., 2013).

A Conceptual Model of PrEP Awareness for U.S. Black Women

PrEP is a potential patient-centered approach that could provide opportunities for women to take control of their sexual health and decrease the risk of HIV. Despite Black women’s increased risks for HIV acquisition and evidence in support of PrEP efficacy in women, research is limited regarding the multiple factors that influence PrEP uptake in Black women in the United States. The impact of PrEP in Black women will remain limited unless systematic approaches are used to identify key multi-level factors influencing PrEP uptake in Black women that can be used to direct future intervention development. We propose the use of an integrated conceptual framework, Anderson’s Behavioral Model (ABM) and the situated information, motivation, behavior (sIMB) model, to assess multiple level factors that influence PrEP uptake by Black women.

To evaluate potential barriers and facilitators for uptake of PrEP by Black women living in the United States, a conceptual model including constructs from the ABM with subsequent individual level intervention driven by the sIMB framework can be used (Amico, 2011; Andersen, 1995; Bradley et al., 2002). Each of these theoretical frameworks has a strong basis in evidence evaluating health outcomes in people living with HIV and in HIV prevention research.

Situated Information, Motivation, Behavior (sIMB) Model

The sIMB model is based on the IMB model, which has been used to explain HIV risk reduction, adherence, and many other behaviors applied specifically to engagement in HIV care (Fisher, 2012; Fisher, Fisher, Amico, & Harman, 2006; Fisher, Fisher, Bryan, & Misovich, 2002; Kalichman, Picciano, & Roffman, 2008; Robertson, Stein, & Baird-Thomas, 2006; Smith, Fisher, Cunningham, & Amico, 2012). The sIMB model is an adaptation of IMB, situated to key psychosocial constructs (e.g., stigma, disclosure) as factors inherent to the model and the development of information, motivation, and behavioral skills, rather than as moderators external to the process, as conceptualized in the traditional IMB framework. sIMB is currently being used as the theoretical basis for a multi-site randomized control trial of an intervention to increase adherence to HIV care and ART to promote rapid and sustained viral load suppression for patient initiating HIV care (R01AI103661; NCT1900236)

Anderson’s Behavioral Model (ABM) of Health Services Utilization

The ABM model is one of the most widely used frameworks in research, primarily conducted in the United States and the United Kingdom, and has been used to understand health care utilization. It posits that the interplay of individual, societal, and contextual factors impacts health service utilization. As a multi-level model, it incorporates individual, as well as contextual factors as determinants of health services uptake.

ABM has been used in studies of factors associated with engagement in HIV care, ART adherence, and disparities in HIV service utilization by vulnerable populations (Andersen et al., 2000; Anthony et al., 2007; Cunningham et al., 1999; Kilbourne et al., 2002; Smith & Kirking, 1999). The model has also been applied to understanding uptake of health services in highly vulnerable U.S. populations, including minorities, children and adolescents, people with mental health issues, disabled individuals, and impoverished persons (Gelberg, Andersen, & Leake, 2000).

Population Factors

Individual-level factors are broken down into constructs focused on predisposing characteristics, enabling resources and perceived need for health services (Andersen, 2008). Predisposing factors include individual level factors, such as demographics, mental health, attitudes, values, and health beliefs, as well as contextual factors, such as community demographics, social structures, cultural norms, and political views. In our conceptual model, intersectional stigma, defined as the synergistic, mutually constitutive associations between social identities and inequities, is also included as an individual-level predisposing factor (Simien, 2007). Enabling factors may facilitate or impede health service utilization. Financial and organization factors are included in our construct of the model. Financial, individual, and contextual-level equivalents may include health insurance status, transportation, community income, and health insurance coverage, as well as pricing for goods and other services. Lastly, perceived need encompasses an individual’s perceived and actual need for services, which may reflect respectively an individual’s perceived risk of HIV and need for PrEP services, as well as sexual risk factors used to determine indications for PrEP service referral. Environmental and population assessment of need would include health-related conditions in the environment and epidemiologic indicators of population health indices (in this case potentially including sexually transmitted infection/HIV incidence rates).

Environmental Factors

Healthcare delivery systems, external environmental factors, and health policies are environmental factors that can influence healthcare access and delivery of healthcare services for many. These can be stratified into two major dimensions: resources and organization. Resources refer to the revenue and labor requirements needed for the healthcare system to function. This can be further extrapolated into understanding how resources are geographically distributed, the personnel needed to service a population, and supplies necessary to deliver services. By examining the extent of services needed for health service delivery, intervention development can focus on populations with limited resources for service delivery and health policies needed to drive change for more equitable service distribution.

The manner in which a healthcare system uses its resources describes organization. This is further divided into how a population accesses an organization and the structure of the organization. For healthcare to be delivered effectively, there are certain steps a patient must take to enter and traverse through the system to access treatment. Potential barriers may include co-pays associated with visits, insurance deductibles that must be met, and the length of the queue waiting to receive treatment. The characteristics of the healthcare system, including, but not limited to, processes for referral, indications or guidelines for treatment delivery and the utilization of ancillary staff, are all examples of structures that determine what happens to patients once they enter the system (Andersen & Newman, 1973).

Health Behavior: Integration of the sIMB Model

The IMB model originated as a tool to understand determinants of HIV risk and prevention behavior to aid in targeted intervention development (Fisher, 1997). Specifically, it conceptualizes health promotion interventions through causal relationships between constructs. Information, as it relates to specific health-related facts or simple rules, is a health behavior determinant. Health-related behavior is also influenced by individual (personal attitudes about the health behavior) and social (social norms related to the health behavior) motivations. Lastly, behavioral skills objectively neededrelated to implementation of the health behavior and perceived self-efficacy also are key components included in the IMB model. Health promotion behavioral skills directly influence behavior through health promotion information and motivation. When conducting research using the IMB approach, it is recommended that existing individual-level information, motivation, and behavioral skills be assessed. Intervention development can then be targeted to deficits found within these constructs and, finally, evaluation of overall impact can be conducted (Fisher, Fisher, & Harman, 2003). The sIMB model adopts these core determinants and situates them in the context of the socio-environmental and psychosocial constructs in which they occur. Rather than consider these factors as external moderators, or modifiers of IMB efficacy, they are viewed as inherent to the model and essential to the development of information, motivation and behavioral skills situated in real-world contexts (Amico, 2011). By including sIMB constructs within the ABM theoretical framework, individual-level health behavior core determinants, in the context of the broader environmental and population-level characteristics influencing health behavior, may be better understood (Figure 1).

Figure 1.

Conceptual Model – Grounded in ABM with sIMB Constructs for Intervention Development.

Note. Conceptual model adapted from Andersen Health Care Utilization Model (ABM), incorporating constructs from situated, Information, Motivation, Behavioral Skills model (sIMB). Items in boxes are constructs of ABM and sIMB, with arrows relating their interactions. PrEP = pre-exposure prophylaxis; IMB = information, motivation, and behavioral skills model.

Development of Multilevel Interventions Informed by the Conceptual Model to Increasing PrEP Uptake by Black Women

Application of our conceptual model to understanding and developing health prevention interventions for uptake of PrEP by Black women could be highly effective and would help evaluate factors that vary by geographic location and/or rurality in research in specific U.S. regions to ensure appropriate, effective design of targeted interventions (Cene et al., 2011). It would also allow a comprehensive evaluation of environmental and population factors that contribute to changing health behaviors. Interventions grounded in this model would be multi-level, by definition, and focus on the individual and community-level information, motivation, and behavioral skills needed for uptake found with contextualized predisposing, enabling, and need factors. The need for modification of a behavioral intervention could be based on epidemiologic population trends in HIV incidence rates.

To ensure appropriate intervention development, it will be necessary to use the IMB approach to conduct research, by first gathering adequate data from the population in which the intervention will be targeted, then developing an intervention influenced by the information gathered, and, finally, evaluating overall impact of the intervention in the community. Full understanding of Black women and their individual and contextual-level characteristics related to the uptake of PrEP will likely require a qualitative and quantitative approach. Using a mixed method approach would provide a more in-depth and richer understanding of potential barriers and facilitators of PrEP than using either quantitative or qualitative methods alone (Ivankova, Creswell, & Stick, 2006; Tashakkori & Teddlie, 1998). Mixed-method studies would need to be followed by another evaluation of the environmental factors faced by Black women that may impede uptake of PrEP, which might include gathering information about healthcare coverage, local healthcare providers, and facilities that provide PrEP services. Rigorous research might also need to focus on the information, motivation, and behavioral skills of providers, community advocates, and other key stakeholders, who may need education to support and provide health prevention messaging about PrEP uptake. By including both individual and structural level data findings to inform intervention development, a multi-level intervention will have a higher likelihood of being effective and sustainable past the duration of the study. Lastly, by using multiple measures of impact, such as biological markers, overall impact of the intervention can be assessed.

Discussion

Black women are disproportionately burdened by HIV. PrEP is a promising biomedical strategy to reduce HIV in this vulnerable population. Therefore, PrEP implementation in this disproportionately-impacted group is an important public health priority, but data regarding PrEP uptake by Black women are limited. Novel frameworks are needed to understand contextual multi-level factors associated with PrEP uptake and to guide PrEP research in populations such as Black women (McMahon et al., 2014). Our proposed integrated conceptual framework provides a multi-level approach to systematically evaluate PrEP uptake in vulnerable populations such as Black women.

Our framework could be used to gain a deeper understanding of modifiable barriers and facilitators to PrEP uptake, which has the potential to foster targeted intervention development and evaluation of theory-guided interventions in real-world settings. The integrated framework can be used to (a) explore potential barriers and facilitators to PrEP uptake; (b) once identified, assess the strength and magnitude of relationships between factors; and (c) assess the efficacy an intervention to improve PrEP uptake. The integrated framework has not yet been tested. However, researchers can use the framework as a guide to elucidate individual- and environment-level factors associated with PrEP uptake and then use the findings to develop or adapt existing interventions with the goal of improving PrEP uptake. Healthcare providers can use information gained from researchers to identify patients at risk for suboptimal PrEP uptake and to initiate appropriate referrals. For example, if depression were determined to be a salient factor that was impacting PrEP uptake and social support was determined to be a mediator of depression and PrEP uptake, then healthcare providers could refer the patient to a support group or peer educators who are taking PrEP.

Due to the time-intensive nature of applying this conceptual framework to understanding facilitators and barriers of PrEP uptake by Black Women, there is a drawback in time to effective implementation of intervention strategies to increase uptake of HIV pre-exposure prophylaxis in this at-risk population. Therefore, it might be prudent to simultaneously apply already-existing behavioral interventions in this particular population while research is being conducted using our theoretical framework. For example, CDC has published and publically provides literature on HIV prevention interventions that have increased condom use with main partners by at-risk women (Belgrave, Corneille, Nasim, Fitzgerald, & Lucas, 2008; Harshbarger, Simmons, Coelho, Sloop, & Collins, 2006). The efficacy of these interventions is currently being seen in the declining overall rates of HIV infection among women. Another limitation of applying Our conceptual model is that it has not been yet been validated in research. The framework is currently being used to evaluate the uptake of PrEP by young Black MSM in the Deep South with the goal of developing a tailored intervention. However, the constructs used in our model have been validated separately in HIV prevention research and in research evaluating health outcomes for people living with HIV. ABM has been used to study engagement in HIV care, ART adherence, and disparities in HIV service utilization (Andersen et al., 2000; Anthony et al., 2007; Cunningham et al., 1999; Kilbourne et al., 2002; Smith & Kirking, 1999). Because we incorporated the sIMB theoretical framework into our conceptual model, investigators will also be able to evaluate core determinants of initiating the health behaviors related PrEP uptake by Black women. IMB has been used to study engagement in HIV care, adherence to ART, and HIV risk behavior change in adolescents (Fisher, 2012; Fisher et al., 2006; Fisheret al., 2002; Kalichman et al., 2008; Robertson et al., 2006; Smith et al., 2012). However, in light of the persistent racial and ethnic disparities seen in Black women as well as geographic variability in infection rates, research using innovative conceptual models to tailor intervention development is urgently needed to understand and implement new tools to curb the epidemic in these communities. This is especially needed for research studying uptake of this relatively new and highly effective biomedical prevention tool.

Conclusion

We proposed a novel framework to explore multi-level factors associated with PrEP uptake in Black women. The framework will provide a systematic approach to identifying potentially modifiable, multi-level factors. Providing data to guide intervention development and implementation of PrEP for Black women could reduce HIV-related disparities.

Key Considerations.

• Pre-exposure prophylaxis has been shown to be effective in reducing HIV acquisition in individuals with optimal adherence.

• Black women in the United States are disproportionately impacted by the HIV epidemic, and pre-exposure prophylaxis offers a contemporary strategy for HIV prevention.

• Many Black women are unaware of PrEP. Women may encounter multilevel barriers related to PrEP utilization when they become aware.

• An integrated model for …‥