Rapid or Immediate ART, HIV Stigma, Medical Mistrust, and Retention in Care: An Exploratory Mixed Methods Pilot Study

Rapid or immediate antiretroviral therapy (iART) after HIV diagnosis improves linkage to care and time to viral suppression. However, iART may affect or be affected by HIV-related stigma and medical mistrust. In this mixed-methods pilot study, we examined the bi-directional role of HIV stigma, medical mistrust, and visit adherence (VA) in the context of iART in a diverse, newly diagnosed patient population. Participants were recruited from an HIV clinic in New York City and we utilized a convergent parallel design integrating quantitative data from demographic surveys, the HIV Stigma Survey (HIVSS), the Medical Mistrust Index (MMI) and electronic medical records, and qualitative data from in-depth interviews. Among the sample (N = 30), 26% (N = 8) initiated ART same-day or within 3 days, while the majority (N = 17) initiated between 4 and 30 days, and 17% (N = 5) initiated ART > 30 days. The median (range) age was 35, and most were English-speaking, Black or Hispanic men and identified as gay. Time to ART initiation was associated with time to linkage to care and time to viral suppression. Day 0–3 group’s major theme was iART as stigma prevention, and they had the highest mean HIVSS, lowest MMI score, and a visit adherence of 0.86. Day 4–30 group’s major theme was alleviation of internalized stigma, and they had the lowest mean HIVSS score, and highest visit adherence of 0.91. Day > 30 group’s major theme was exacerbation of perceived or anticipated stigma, had the highest MMI score and a visit adherence of 0.85. iART implementation requires equitable strategies that address HIV-stigma and mistrust. Supplementary Information The online version contains supplementary material available at 10.1007/s10461-023-04058-4.


Introduction
Rapid or immediate antiretroviral therapy (iART) upon HIV diagnosis has become standard of care globally, including the United States (US) [1][2][3]. iART is defined as the offer of ART on the same day or as soon as possible after HIV diagnosis without waiting for laboratory testing, and includes various time frames but generally considered within 30 days. Global randomized controlled trials and observational studies in various clinical settings in the U.S have established that iART improves linkage to care and shortens time to viral suppression [4][5][6][7]. The HIV epidemic in the US continues to be marked by persistent inequities by race, ethnicity, gender, and sexual minority status; it is critical to understand the psychosocial costs and benefits of iART among these groups during a highly vulnerable time immediately postdiagnosis [8,9].
HIV-related stigma has influenced care engagement on multiple levels for the most affected populations living with HIV in the US [10][11][12]. Whether internalized, anticipated, or enacted stigma from prior lived experiences, studies have shown that racial, gender and sexual minorities experience higher rates of various forms of HIV-related stigma, leading to suboptimal HIV care outcomes [13,14]. Medical mistrust is another barrier to HIV care engagement and ART adherence [15]. Studies have shown that people with HIV (PWH) believe there is an overemphasis on ART at the expense of other non-HIV related priorities-a consequence of distrust in providers and the health system [16]. Therefore, it is vital to understand perceptions and experiences of receiving iART, and how receiving iART may impact HIV-related stigma and medical trust/mistrust, and how that, in turn, influences HIV care engagement.
Few qualitative studies have explored the impact of iART during the post-diagnosis period for those newly diagnosed with HIV in the US. A recent qualitative study from San Diego among 19 men who have sex with men (MSM) with HIV showed that patients' desires to remain virally suppressed and prevent ongoing viral transmission were powerful motivators in the acceptance of iART [17]. However, no studies examine more deeply how iART impacts important psychosocial factors such as HIV-related stigma and medical trust/mistrust at the time of diagnosis and the subsequent effect on HIV care engagement. This mixed-methods pilot study sought to understand the psychosocial costs and benefits of iART for a newly diagnosed predominantly Black, Latino, and MSM patient population in New York City and explore associations between iART, HIV-related stigma, medical trust/mistrust, and retention in care.

Study Design
We utilized a convergent parallel mixed-methods research design, combining quantitative data from a demographic survey, stigma and mistrust surveys, and electronic medical record (EMR); and qualitative data from semi-structured indepth interviews (IDIs). Theoretical considerations regarding the role of HIV-related stigma and medical mistrust were grounded in a conceptual framework encompassing the socio-ecological model and the theory of planned behavior, underlining the multi-level influences that determine the impact of iART, HIV stigma and medical trust/mistrust on retention in care (Fig. 1).

Participants and Eligibility Criteria
Participants were recruited from an HIV care clinic in an academic medical center in New York City, serving a safety net population between October 2020 and June 2021. The Fig. 1 Conceptual framework integrating social-ecological model and theory of planned behavior inclusion criteria were: [1] aged 18 years or older; [2] diagnosed with HIV on or after January 1, 2018; [3] offered iART; [4] enrolled in the clinic at the medical center for at least 6 months; [5] reside within one of the NYC boroughs; and [6] English or Spanish speaking.

Procedures
Participants were recruited in-person, by phone, or by responding to flyers posted in clinic waiting areas. Participants were electronically consented via REDCap. EMR data were abstracted. The interview guide and surveys were developed to last 60-90 min, translated by a native Spanish speaker, and backtranslated into English. The interview and surveys were administered either in-person or virtually while maintaining participant privacy and confidentiality. All interviews were audio-recorded, professionally transcribed verbatim and professionally translated from Spanish to English. Participants received $50 gift cards.

Qualitative Interview Guide
IDIs were conducted using a semi-structured interview guide, iteratively developed by the research team and grounded in the conceptual framework. The interview guide was piloted with the first two study participants (included in the final analysis), and adjustments were made to ensure the meaning and intention of questions were understood. The five domains of the guide were [1] Experiences and perspectives from new HIV diagnosis to iART initiation; [2] HIV care engagement; [3] Medical trust/mistrust; [4] Stigma; and [5] COVID-19.

Quantitative Measures iART
The cohort was divided into three groups according to time to ART initiation: 0-3 days, 4-30 days, and > 30 days. This categorization is based on the New York State Department of Health AIDS Institute definition of rapid ART initiation, which states "ideally on the same day or within 72 hours" however never beyond 30 days of diagnosis [18].

Participant Demographic Survey
Using a structured instrument developed by the research team, participant demographics collected included age, country of birth, primary language spoken, race/ethnicity, gender identity, sexual orientation, the highest level of education, current employment status, current housing situation, current annual income, health insurance and relationship status.

HIV Stigma Survey (HIVSS)
A validated 40-item instrument was used to examine participants' internalized, interpersonal and community-based HIV stigma [19]. Each item was scored on a 4-point Likerttype scale: 1 = Strongly Disagree, 2 = Disagree, 3 = Agree, 4 = Strongly Agree. The HIVSS is comprised of 4 subscales addressing different dimensions of stigma: negative selfimage, personalized stigma, public attitudes about HIV and diagnosis disclosure. We interpreted the type of HIV-related stigma (e.g. internalized, anticipated, enacted, community) according to the framework laid out in Turan et al. [12]. Negative self-image corresponds to internalized stigma, while personalized stigma corresponds to anticipated, perceived, or enacted stigma. Public attitudes correspond to community stigma.

Medical Mistrust Index (MMI)
A validated 17-item instrument designed to measure individual trust/mistrust in the health care system was used. The Medical Mistrust Index (MMI) is not particular to one provider at a healthcare facility (also known as "provider trust"), but rather healthcare institutions [20]. Each item was scored on a 4-point Likert-type scale: 1 = Strongly Disagree, 2 = Disagree, 3 = Agree, 4 = Strongly Agree.

Medical Chart Review
We conducted an EMR review of demographic and clinical outcomes over the first 6 months after diagnosis. Participant demographics included age, race/ethnicity, gender identity, sexual orientation, language, and health insurance. HIV risk factors (heterosexual, MSM, IDU, blood transfusion, etc.) and clinical outcomes were recorded, including number of comorbidities, dates and locations of HIV diagnosis and iART offer/initiation, and date of initial viral suppression, if ever. Calculations were derived from the date of events to establish time (in days) to linkage to care, offer of iART, initiation of iART, and viral suppression. Retention in care, or visit adherence, was gathered 6 months post-iART (number of completed HIV medical care visits over number of scheduled visits) [21].

Data Analysis
Data from the demographic survey and EMR were analyzed using descriptive statistics. HIVSS and MMI were analyzed  using bar graphs and box plots. Fisher's exact test was used to examine differences between categories of time to ART initiation. Mann Whitney U test was used to analyze difference between median and ANOVA was used to analyze difference between means. In order to limit misleading results due to a small cohort, continuous variables were assessed for skewness to determine whether the median (range) or mean (standard deviation) was more appropriate to report. If the results were skewed, median results (range) were reported. For variables following a normal distribution, mean results (standard deviation) were reported. For the HIVSS and MMI scores, we used mean (standard deviation) because this has been the traditional measure reported in the literature. We also included a sensitivity analysis breaking down the iART groups as following: 0-7 days, 8-30 days, > 30 days.

Qualitative Analysis
Using an inductive approach, OM and CL developed an initial codebook using a priori codes derived from the interview guide and the conceptual framework. Emergent codes were also identified. Each researcher independently coded a set of transcripts and inconsistencies were reconciled among the larger research team. We then calculated intercoder agreement (pooled Kappa of 0.91). We used Dedoose to analyze data, looking for code co-occurrence pertinent to our research question. Each quote is contextualized by the iART group, and the race and ethnicity answered by participants in the demographic survey. Other demographics were not included with each quote to protect the identity of the participants.

Data Triangulation and Integration
Qualitative data were triangulated with survey and EMR data to enhance quality. For example, if a participant's race was marked as "Other" in chart review, their demographic survey and in-depth interview responses were used to clarify their race. Quantitative (HIVSS, MMI scores, and visit adherence) and qualitative data (themes from in-depth discussion) were then combined to determine convergence and divergence of the data in an integrated table.

Quantitative Results
Of the 30 enrolled participants, 20 conducted the study virtually, and 10 conducted the study in-person. Four interviews were conducted in Spanish and 26 in English. Table 1 shows the overall demographic characteristics and of each group by time to ART initiation. The overall cohort (N = 30) had a mean (SD) age of 35.4 (11.7), was mostly Black (33%) or Hispanic (60%), English-speaking (73%), gay (53%), male (80%), had at least some college (73%), making less than $20,000 (43%), stably housed (93%), and on Medicaid (60%). The 4-30 day group reported a higher income. Table 2 shows clinical characteristics, HIVSS and MMI scores, and HIV outcomes. The overall cohort identified MSM sexual contact as their major HIV risk factor (63%), had less than 2 comorbidities (80%), and about a quarter identified either a substance use (23%) or mental health diagnosis (27%). About 20% had a CD4 less than 200 at presentation. The > 30-day group had more individual participants reporting 3 or more comorbidities.
The median time to linkage to care at our clinic was 9 days, and the median time to viral suppression was 49 days. There was a significant association between time to ART initiation and time to linkage care, and time to ART initiation and time to VLS.
The mean visit adherence over 6 months post-diagnosis was 0.86 among the 0-3 day group, 0.91 among the 4-30 day a Preference was given to patient survey responses over chart review when determining race and ethnicity. All responses of "Other" in which "Latino," Hispanic," or some variant were enteredwere maintained as "Other" as per NIH guidelines (patients also responded "Puerto Rican" "Spanish" and "Dominican"; one patient said "mixed" and one said "biracial") b PWH who are uninsured or underinsured in New York state are eligible for immediate coverage via ADAP (AIDS Drug Assistance Program). Patients who present to our site without insurance are enrolled immediately, so no patient in this cohort is listed as uninsured group, and 0.85 among the > 30 day group, and there was no statistically significant difference between them. Table 2 shows no statistically significant difference between the mean HIVSS and MMI scores between the groups. Figure 2 shows boxplots of each group's HIVSS scores at 6 months. The scale items and distribution of each subscale are outlined in Supplement Table S1. The 4-30 day group had the lowest mean HIVSS scores compared to the other groups.
Breaking down the HIVSS further into its four subscales corresponding to different types of HIV-related stigma (Fig. 3), the 0-3 day group had the highest negative selfimage (e.g., internalized stigma). The > 30-day group had the most personalized stigma (e.g., anticipated, perceived, or enacted stigma) but the least negative self-image or disclosure stigma. Figure 4 shows boxplots of the MMI scores for each group at 6 months. The scale items are outlined in Supplement    Table S2. The 0-3 day group had the lowest mean MMI score, which increased as time to ART initiation increased.
The results of the sensitivity analysis summarized in Supplement Tables S3 and S4 also showed significant association between time to ART initiation and time to linkage care, and time to ART initiation and time to VLS. The highest mean HIVSS was in the 8-30 day group, the highest mean MMI score was in the > 30 day group, and the highest visit adherence was in the 0-7 day group.

"My Beauty wasn't Going to be Affected": iART as Stigma Prevention
Participants discussed iART as a means of preventing different stigmatized outcomes they feared after receiving an HIV diagnosis, including physically appearing sick, having detectable viral load, experiencing discrimination or enacted stigma, and progressing to AIDS and death. A gay Hispanic male described how his fears after diagnosis resulted from the historical and intersectional stigma of HIV/AIDS with both racism and homophobia at play, and how iART was seen as an opportunity to continue living his life without major change.
"When I was diagnosed, I was just really eager to take medication. The first doctor at CityMD was like 'It's not a death sentence.' When I was diagnosed, the warnings from the '80s and '90s all came back to me. I get they were trying to warn people. Still, it created a huge stigma…during that time when people, even scientists, and doctors, were just bashing the gay community or pinning it on gay people like us, and the minority groups as well…that's where the stigma comes from, people are scared or nervous about the disease. After I took the medication, I guess I'm somewhat the same person…As long as you take your medication, you'll be all right." (Participant 5, 4-30 day ART group, Hispanic Male, Other "Hispanic") Immediately following diagnosis, participants voiced their concerns about looking physically ill. Being offered iART provided a way to cope with the possibility of looking sick or physically embodying the disease. A gay Dominican male described how HIV was a threat to his physical appearance and how iART helped combat that process. Many participants described the internalized stigma resulting from the transition from being a "healthy" person to being a "sick" person. However, iART reframed this dichotomous paradigm for some participants. Participant 25 described how iART helped them realize that because they now were living with HIV didn't mean they were necessarily "sick".
"It helps you live beyond the stigma because you know that when you start taking it, you are taking something to better your situation and correct whatever is going on so that you are not going to get sick. Having HIV doesn't mean you're sick." (Participant 25, 0-3 day ART group, Non-Hispanic Male, Black/African American)

"I was Looked Upon Differently": iART Exacerbating Anticipated or Perceived Stigma
Many participants stressed that while iART was helpful with the internalized stigma, the offer of medication was different from actually picking up and initiating the medications, which brought new stresses and feelings of stigma. The pharmacy was a source of anticipated or perceived stigma among participants and many recalled their struggles with their first visit. While discussing how starting immediately may have brought up anticipated or perceived stigma, participants disaggregated the immediate offer (the 'i') from the medication itself (the 'ART'). While many had a positive perception of iART being offered right away, they had a negative perception of the physical ART pill, which was a source of stigma. Participant 7 described how the medications were a constant reminder of his HIV: "At first when I would take the medication I would feel dirty. I would tell myself "You're contaminated." In the beginning taking the medication every morning was the most difficult thing, knowing I had the virus, it was harder to have the medication in hand" (Participant 7, 4-30 day ART group, Hispanic Male, Other "Latino") Another described feeling that their entire identity was reduced to having HIV due to having to take medication every day. For participant 9, this initial stigma abated over time.
"I felt like I was the disease in a sense because I was having to take this obligatory one pill a day. I thought, 'Oh my God, I am the disease.' I take this pill everyday. It's such a burden. But now it's like anyone else who has to take medications." (Participant 9, 4-30 day ART group, Non-Hispanic Male, White) While iART alleviated internalized feelings of stigma, it did not remove the stigma associated with taking the medication and the anticipation of a negative reaction from anyone who would see them. Participant 4 discussed the various places where he is preoccupied with hiding his medication, so it is not incidentally found. However, iART often reminded participants of nefarious profit motives of the healthcare system, which included drug stores, pharmaceutical companies, large hospitals, and other entities they saw as profiting. In this context, the iART emphasis on immediate medication created concern that treatment recommendations were driven by motivations other than patient health. Participant 4 described their experience choosing to start medication while not trusting the health system.
"I was reading about these companies that make a lot of money from making HIV treatment for people like Participants also identified a shared-decision making approach as an essential component of building trust through iART. Participant 12 identified shared decision-making as key to building mutual respect through the first major decision for provider and patient.
"I would recommend that when seeing patients like myself, especially in a clinic where a lot of the doctors are white males and treating a lot of patients that look like myself in the hospital. I would say what my doctor did, when he laid out the different options in front of me, that was very comforting, I liked that a lot…Even though he's the one with the power, it feels more so like a dialogue and how we can both move forward." (Participant 12, > 30 day ART group, Hispanic Male, Other "Mixed") Participants emphasized that trust is a process rather than an immediate feeling. Some discussed coming in actively distrusting the healthcare system but then seeing how interpersonal relationships made an impact longitudinally. Participant 7 discussed how this process evolved through iART.
[iART] is very very effective. I didn't think the group of doctors would be this prepared, it was practically immediate. I was diagnosed between 10 and 11pm on Thursday or Friday, and started treatment on Satur-day… In the beginning, [my trust] was very low. You go in there thinking they don't care what I am going through. As you come back every month, every three months, that changes. They let you know that they are here for you." (Participant 7, 4-30 day ART group, Hispanic Male, Other "Latino")

"I Wanted to see my Progress": iART as Motivation for Care Engagement
For most participants, starting ART immediately served as motivation to stay engaged in HIV care. Participants defined their engagement in different ways, some said it was about physically coming to clinic and taking their daily However, physically coming to appointments exposed many participants to stigma, particularly in the waiting room. Participant 2 discussed struggling with the lack of privacy involved in coming to the clinic.  Table 3 shows integrated findings from quantitative and qualitative results. Identified themes that were present in the majority of cases in each group were included. The 0-3 day group had the highest mean HIVSS score, the lowest MMI score, and a lower visit adherence than the overall mean. iART as stigma prevention was the theme unique to this group. This group emphasized that iART mitigated community and anticipated forms of stigma, and had the lowest personalized stigma subscale score, yet this group still had the highest negative self-image subscale score.

Integrated Quantitative and Qualitative Results
The 4-30 day group had the lowest mean HIVSS score and the highest visit adherence. Alleviation of internalized stigma was the theme unique to this group. This group emphasized that iART actively reduced internal feelings of shame, and this group had a lower negative self-image subscale score than the 0-3 day group, meaning iART potentially reduced their stigma scores 6 months post-diagnosis.
The > 30-day group had a higher mean HIVSS score, the highest MMI score, and a lower visit adherence than the overall mean. Exacerbation of perceived or anticipated was the theme unique to this group. This group emphasized that iART perhaps worsened anticipation or perception of stigmatizing experiences, meaning that iART potentially increased their stigma scores 6 months post-diagnosis. The > 30 day group had the highest personalized stigma subscale score (e.g., anticipated, perceived, or enacted stigma), which is consistent with the qualitative data.
The themes of Foundation of Provider Trust and iART as care motivation were equally present in all three groups.

Discussion
This is the first mixed-methods study to explore the impact of iART on the important psychosocial constructs of stigma and medical trust/mistrust, as well as longitudinal HIV care engagement. This builds on recent studies examining barriers to and facilitators of rapid ART to further explore the impact of iART on newly diagnosed PWH.
This study demonstrates the importance of a mixed-methods approach. Electronic medical chart review is often inadequate in capturing important demographic variables when considering HIV care inequities and capturing the complex process from diagnosis to linkage to ongoing care. Our approach allowed us to triangulate our data to have the most accurate understanding of each participant's case. Furthermore, we contextualized our quantitative measures of levels of stigma and mistrust with qualitative data that illuminated the longitudinal psychosocial processes that occurred after ART initiation.
This study also begins to characterize HIV care trajectories based on timing of ART initiation. It illuminates the consequential role of HIV-related stigma for newly diagnosed PWH. For the group that started ART same-day or within 72 h, iART appeared to mitigate the development of anticipated or enacted stigma, and community stigma. Significant advancements in the effectiveness of ART empowered participants to feel they could avoid future stigmatizing experiences if they started ART immediately. Because this did not necessarily lead to greater mean visit adherence than other groups, other mediators or systemic factors were likely at play when it came to long-term HIV care engagement for this group.
For the 4-30 day group, iART helped alleviate internalized HIV stigma. Participants coming in feeling shame about their diagnosis were able to process their emotions through starting iART, providing an opportunity for coping with and reducing the internalized HIV stigma built up over time. This group also had the highest visit adherence, possibly suggesting that the active alleviation of these feelings made this group more comfortable keeping their appointments. Future iART implementation may include a stigma intervention along with other social determinants of health that are part of the iART package (e.g. housing, insurance).
Participants who took the longest to initiate ART (> 30 days) specifically highlighted the role of anticipated, perceived, or enacted stigma. The ART in iART was the first manifestation or reminder of one's HIV status in a vulnerable time after diagnosis, which led participants to hide medications or highlight difficult experiences picking up ART at the pharmacy. This group also had the highest level of mistrust and a lower visit adherence. This provides a ripe area for future study and interventions, such as enhanced peer support with the pharmacy or more pharmacist education regarding iART and stigma reduction.
Within every group, participants emphasized that iART was an opportunity to build trust and engage with their provider. The concept of shared decision-making, mainly when participants felt they did not choose whether or not to initiate ART, was particularly stressed. This could be a critical concept in educating providers and clinics on how to implement iART successfully. Furthermore, scores from the MMI, which increased with time to ART initiation, may point to an opportunity to engage with feelings of mistrust with the health system and medical establishment at diagnosis.

Limitations
There were a few limitations to this study. First, all participants were retained in care. Given that we recruited in the clinical setting, it was challenging to recruit patients who had fallen out of care. This likely introduced selection bias to our results, missing the crucial perspectives of those lost to follow-up or not consistently in care. In addition, our study was retrospective rather than prospective. Without a pre-iART baseline, we cannot make any definitive conclusions about the effect of iART on HIVSS and MMI scores, and vice versa. Ideally, we would also want to know how scores from the MMI and HIVSS surveys at the time of diagnosis/iART change over time. We utilized the qualitative data, in which participants describe the process from pre-ART initiation to 6 months post-ART initiation, to infer how these scores may have changed, however this is a pilot using cross-sectional data to generate hypotheses and establish areas of future inquiry. Lastly, this study only covered 6 months visit adherence, and likely a longer period of observation would yield more pertinent data on HIV care engagement over time.

Future Directions
Future directions include a focus on PWH who are out of care, as they are expected to receive the same iART intervention around the country. Participants who initiated iART > 30 days may share features with those out of care and merit further exploration. Implementation science will be critical to identify other important barriers or facilitators to iART given that iART will be carried out in emergency rooms, sexual health clinics, community organizations, etc. Successful iART implementation will encompass entire hospital systems and community networks. As mentioned above, future studies may use a behavioral science approach to measure longitudinal changes in HIVSS, MMI and other psychometric surveys before HIV diagnosis, at diagnosis, and after diagnosis to better understand associations with iART and HIV care engagement. Lastly, it will be essential to leverage technology and virtual care to optimize iART implementation with the repeated disruptions during the COVID-19 pandemic.

Conclusion
This mixed-methods pilot study is the first to provide crucial context to the psychosocial costs and benefits of iART. In terms of care trajectories, iART likely does not overcome established systemic forces that lead to patient disengagement in care. Still, in our study, this population laid out the benefits of iART when done well. Universal approaches to iART may be coupled with an equitable implementation strategy that addresses the deleterious impact of stigma, incorporates concerns around mistrust, and other psychosocial factors that resonate with the communities most impacted by HIV. Identifying patients who need focused support (e.g., support at the pharmacy picking up medication for the first time) and training providers to practice iART successfully (e.g., highlighting shared decision making) will be paramount to optimal iART implementation.