Incorporating Stigma Into Implementation Science: Linking Societal Inuence to the Clinical Encounter to Understand Inequities in Healthcare Delivery

Background: Disease-related stigma is an important, but under recognized barrier to the implementation of evidence-based therapies. Existing implementation science frameworks do not adequately specify the mechanisms by which the outer societal context produces disparate implementation outcomes. Our aim in this study was to evaluate how stigma inuences the implementation of evidence-based hepatitis C virus (HCV) care and, in so doing, make the case for incorporating stigma into the Health Equity Implementation Framework. Methods: From 2015-2019, we conducted a concurrent explanatory mixed-methods study among people living with HCV in Philadelphia. We administered the validated 33-item HCV-stigma scale. Semi-structured interviews were conducted among a purposive subsample of survey respondents. Summative HCV-stigma scale scores and descriptive statistics were calculated. Interviews were transcribed and analyzed for common themes. Survey responses were linked to interview data. Results: Surveys were completed by 265 participants; 22 interviews were conducted with a subset of these respondents. Hispanic/Latinx ethnicity was associated with signicantly higher HCV-stigma scale scores (adjusted mean, 81.39 [95% CI, 76.44-86.33] versus 74.28 [95% CI, 70.51-78.05]; P=0.025). In interviews, participants described the impact of stigma on their engagement with HCV care. Stigma associated with their diagnosis contributed to feelings of dirtiness and shame. Participants described the critical importance of interactions with medical providers and the healthcare system as a whole to either facilitating or mitigating the negative impact of enacted stigma on their trust in medicine and willingness to engage with HCV care. Interactions with family and friends, especially surrounding disclosure of one's diagnosis, were an important mediator of the experience of stigma. Hispanic/Latinx participants described greater feelings of isolation and stigma from family and friends as a result of their diagnosis. Finally, participants described the way stigma shaped their and society's perceptions of HCV therapies, including the media portrayal of pharmaceuticals, the high cost of drugs and insurer denials of treatment.

These ndings address noted gaps in previous literature, by investigating the social relations that produce stigma experiences and the way stigma impacts implementation, contributing to disparities in access to evidence-based care.
Background Implementation science has the potential to substantially improve population health by establishing and evaluating strategies to increase uptake of evidence-based interventions to prevent and cure disease (1)(2)(3). However, the population-level effectiveness of implementation science is restricted by frameworks that do not account for the role of healthcare disparities on the equitable delivery of interventions (2,4). Implementation strategies that do not account for health disparities have the potential to widen inequalities and further marginalize vulnerable patients (2,4,5). Integrating theory, methods, and measures that account for health disparities within implementation science frameworks can promote the equitable distribution of evidence-based innovations that improve health.

The Health Equity Implementation Framework
The recently proposed Health Equity Implementation Framework (4) combines conceptual frameworks from implementation science and healthcare disparities research to better integrate health equity determinants into research on the factors that shape the uptake of evidence-based innovations (Fig. 1a).
Central in the framework is the clinical encounter, which is conceptualized as a site of interaction between an innovation and its intended recipients, including the patient, clinician, and others involved in the delivery or receipt of the innovation (e.g. peer educator, family member). A novel contribution of the Health Equity Implementation Framework is its explicit integration of sociopolitical and economic in uences on the context in which innovations are delivered. While existing determinant frameworks in implementation science include some reference to the broader social world in which innovations are enacted (the "outer setting"), the mechanisms by which societal factors produce implementation outcomes are often underspeci ed.
Frameworks in implementation science deliver value to the eld by unifying inconsistent terminology across theories to provide common de nitions of phenomena and foundations for measurement (1,6,7).
To bolster the utility of the Health Equity Implementation Framework in advancing our understanding of how health equity determinants shape implementation outcomes, we suggest explicit incorporation of an additional theoretical construct: stigma. Stigma is de ned as a "deeply discrediting attribute" that differs from socially-de ned "normal" attributes (8). Stigmatized attributes include disease/illness, behaviors, and social identities, including race and ethnicity (8). The consequences of stigma on worsening health outcomes and health disparities are well established (3,9,10). When multiple stigmatized identities intersect, experiences of stigma, social exclusion, and marginalization are intensi ed. Intersectional stigmas reinforce disparities in healthcare by impacting delivery of service and erecting barriers to disease understanding and treatment, potentially altering the course of illness (11)(12)(13)(14).
In this manuscript we present a case study that demonstrates how disease-related stigma shapes the implementation of an evidence-based clinical innovation. We argue that stigma is an important mechanism connecting societal in uences to the clinical encounter in the Health Equity Implementation Framework. By increasing the speci city regarding how social and structural in uences impact recipients and characteristics of the innovation, we can incorporate methods and measures speci c to health disparities to improve our understanding of the way that social factors in uence implementation.
Through the use of an intersectionality framework, we can better understand how the individual microlevel experiences of disease-related stigmas are shaped by macro-level social inequalities, such as racism (39)(40)(41)(42)(43)(44)(45). Intersectionality frameworks depart from previous public health frameworks that independently evaluate systems of privilege and oppression as individual processes, by recognizing that multiple structural systems interlock in ways that promote advantage and disadvantage at social and individual levels. (45,46) Stigma is inherently rooted in systems of social, economic, and political power and control, and intersectional stigmas manifest within matrices of structural inequalities. (47,48) We applied an intersectionality framework to the developed questions and analyses a priori to elucidate how structural marginalization and social stigma impact health experiences and stigma associated with HCV. (48)(49)(50) In this study, we use a mixed-methods approach to assess how intersectional stigmas, including the intersection of racism with HCV-related stigma, impacts the implementation of evidence-based DAA therapies. Based on the ndings of this analysis we propose that the Health Equity Implementation Framework be modi ed to include stigma as a factor that speci es a mechanism by which societal in uence shapes the clinical encounter.

Study Design & Setting
We conducted a concurrent explanatory mixed-methods study among patients with a history of HCV infection who presented for care at outpatient infectious diseases clinics across Philadelphia,
The study team was led by MES, a female PhD student in Epidemiology with a master's degree in anthropology, who collected all data as part of her dissertation research. She was supervised by JES & FKB, female social scientists with expertise in mixed-methodology and VLR, an infectious diseases physician with expertise in epidemiology. MES attended each clinic during the times when providers who treat mostly people living with HCV were providing care and approached potential participants for enrollment at the end of their outpatient appointment. Participants were recruited in person by the principal investigator, whom they did not know prior to study commencement. The goal of the study, to understand how stigma impacts the experiences of people living with HCV, was made clear to participants at the time of recruitment.
For survey data, 248 participants were deemed su cient to achieve 80% power to detect associations between HCV-related stigma and hypothesized risk factors. For linked semi-structured interviews, we recruited a subsample of participants who completed the survey using purposive sampling aimed at maximizing variation with regards to race, sex, and gender. Interview recruitment continued until thematic saturation was achieved, whereby further data collection yielded no new or relevant information (55,56).

Data Collection
Surveys were administered on laptop computers equipped with headphones using audio-computer assisted self-interview (ACASI) software, a user-friendly digital data collection interface, using audiorecorded instructions, questions, and answers to guide participants through a survey (57). Survey data included self-reported patient demographics (age, sex/gender, race, ethnicity, education), history of injection drug use, sexual orientation, HIV infection status, years since HCV diagnosis, and stage of HCV management (i.e., spontaneously cleared; diagnosed, untreated; previously treated, not cured; currently being treated; treated, cured). All participants completed the HCV Stigma Scale (HCV-SS), a previously-validated (construct validity, 85%) and reliable (Cronbach α, 0.96) self-administered 33-item questionnaire (58) adapted from the Berger HIV Stigma Scale (59). Questions address domains of personalized stigma, negative self-image, disclosure of infection, and public attitudes, with responses recorded on a four-point Likert scale (1 [strongly disagree] to 4 [strongly agree]). The total HCV-SS score is summed from item scores (range, 33-132), with higher scores indicating greater experiences of stigma.
Qualitative data were collected via an open-ended free-response question within the survey instrument and in-depth semi-structured interviews among a subset of participants. To contextualize situations within which perceptions of HCV-related stigma arise and in uence engagement with healthcare, our interview guide contained four thematic domains: 1) individual perceptions of HCV infection; 2) medical knowledge about HCV infection; 3) social support and interpersonal interactions; and, 4) healthcare interactions. Sub-questions and probes encouraged respondents to re ect on actual interactions or situations. All interview questions were designed to be open-ended and non-leading; participants were encouraged to share stories and experiences that were most important to them.
All interviews were audio-recorded, with the permission of the respondent. The Interviewer took notes by hand and wrote memos following each interview. Audio-recordings were transcribed by a third-party Health Insurance Portability and Accountability Act of 1996 (HIPAA)-certi ed transcription service.
Interview transcripts and survey free-response data were uploaded to NVivo 11 qualitative data analysis software (QSR International, Melbourne, Australia, 2015) and linked to survey data. Data were analyzed using a combination of inductive theme generation, (60, 61) the application of an intersectionality framework to coded data (62,63) and, for the purposes of the analysis presented in this manuscript, organization of our ndings using the Health Equity Implementation Framework. All codes were evaluated for meta-themes, thematic variation, and relationships among codes (MES coded all qualitative data, JES & FKB reviewed codebooks, themes, and analyses) (64).

Statistics
Baseline descriptive statistics were calculated as counts and proportions. Some response categories were collapsed in reported data to limit small cell counts. Difference in characteristics between participants with and without interview and/or free-response data were assessed using chi-square or Fisher's exact tests, as indicated, for categorical variables and t tests for continuous variables. Mean HCV stigma scale scores were calculated overall and by race/ethnicity. Due to the low number of HCV-monoinfected participants participating in interviews, we did not stratify analyses by HIV status. All quantitative analyses were conducting using Stata 15.0 (Stata Corporation, College Station, TX).

Participant Characteristics
We approached 288 patients for the survey study, of whom 270 (96.4%) agreed to participate, and 265/270 (98.1%) completed the entire survey. Qualitative data (interviews and/or free-response survey item) were collected for 96 participants (36.2%), of whom, 84 (31.7%) provided survey free-response data and 22 (8.3%) participated in in-depth interviews. Each interview lasted approximately 45 to 60 minutes.
There was no difference in baseline characteristics between those who chose to contribute qualitative data (interview and/or free-response) and those with quantitative data only. However, compared to those who were not interviewed, interviewees were more likely to be coinfected with HIV (P = 0.002) and diagnosed within the previous ve years (P = 0.016). The likelihood of contributing qualitative data (P = 0.896) or interview participation (P = 0.483) did not differ by HCV-SS score (Table 1).

Experiences of HCV-related Stigma
Qualitative analyses revealed the in uence of stigma within the clinical encounter, as well as downstream societal experiences. Subtheme analyses identi ed how stigma within the clinical encounter may impact engagement with care and uptake of DAA therapies. Table 3 summarizes these subthemes and presents exemplar quotes.

Patient Factors
Many patients described feelings of internalized stigma (inward negative feelings or self-image), with salient subthemes including feeling unclean and shame. Of note, all participants who described feeling unclean reported a history of injection drug use, describing how their internalized feelings of dirtiness related to HCV intersected with feelings of dirtiness associated with use of injection drugs. Some patients elaborated a fear that the sense of dirtiness due to HCV infection would persist following cure.
Many patients described feelings of guilt and shame regarding their HCV diagnosis; some further contextualized this within shame of infection through injection drug, describing HCV as a symbol of their drug use history, in addition to a disease. Multiple participants expressed concerns that disclosing their HCV status would reveal their history of drug use. Alternatively, patients who explicitly described not feeling ashamed also contextualized this within discussion of transmission, with some directly stating that they did not contract HCV through drug use or the mode through which they were infected (e.g. blood transfusion).

Provider Factors
Patients widely described supportive healthcare environments as an important facilitator of engagement with care and treatment. Multiple participants described their HCV-providers as sources of emotional support who eased experiences of HCV-related stigma.
While participants overwhelmingly reported positive experiences with their HCV-care providers, the juxtaposition of negative experiences with HCV stigma in other healthcare settings underpinned the importance of a supportive treatment environment. Negative healthcare experiences were reported within clinical encounters where patients were seeking HCV treatment, as well as in other non-specialty healthcare settings, including the Emergency Department. Healthcare stigma ranged from perceptions of differential treatment to direct refusal of care. For many patients, experiences of HCV stigma in healthcare posed signi cant barriers to linkage with specialty care.
Multiple participants who identi ed as Black/African American recounted experiences of racism in healthcare, including how provider-level racial biases and structural racism in medicine impacted engagement in HCV care. Moreover, Black/African American patients more frequently discussed the importance of trust in providers, and how mistrust of healthcare is an important barrier to care. Alternatively, Hispanic/Latinx participants who were born outside of the US speci cally described their healthcare providers as their sole source of social support.

Interpersonal Relationships
Support from friends and family emerged as an important facilitator of engagement with care and motivator of treatment. Perceived HCV-related stigma reduced participants' willingness to disclose their HCV status and seek out social and emotional support. Almost all participants who described positive disclosure experiences had HCV-SS scores below the mean.
White participants more often endorsed that they had not been treated differently by friends or family following their HCV diagnosis, while Hispanic/Latinx and black/African American participants more frequently described experiences of shaming, social distancing, and negative disclosure experiences. Moreover, Hispanic/Latinx participants frequently described familial experiences of HCV-related stigma intersecting with stigmas towards injection drug use. HCV-related stigma from family regularly led to internalized feelings of shame and decreased self-worth. Furthermore, the extent of social support that participants received outside of healthcare impacted both their willingness to seek care, as well as their reliance on healthcare as a source of social support.
Characteristics of DAA Therapies Across analyses, participants described the costs of DAAs and insurer restrictions on treatment as a barrier to care that illustrated societal stigmas towards HCV. Moreover, participants frequently discussed how the intersection of HCV and injection drug use stigma impacts treatment delivery. For many participants, the high cost of DAA therapies fosters distrust in the healthcare system as a business that differentially cares about the health of its consumers. Insurer-denial of therapies reinforces this distrust and the stigma experienced by patients, and itself is a barrier to continued engagement with healthcare.
Black/African American participants more frequently discussed the power of the media to facilitate HCV care and reduce stigma. Speci cally, participants explained how advertisements for DAA therapies raise awareness of HCV and the curability of the disease. Alternatively, representations that normalized HCV, such as diverse portrayals of patients, dispelled some of the myths and stigmas that intersect on HCV.

Discussion
Using a mixed-methods approach, we evaluated the in uence of HCV-related stigma in the DAA treatment era. We found that most patients experience some degree of stigma related to their HCV diagnosis. While the likelihood of experiencing HCV-related stigma did not vary by race/ethnicity, Hispanic/Latinx participants had signi cantly higher HCV-SS scores than their White counterparts. In our qualitative analyses, we found that among Hispanic/Latinx patients, social rejection and isolation due to HCVrelated stigma may be heightened, ultimately reinforcing internalized stigmas. Our ndings demonstrated the important impact of HCV-related stigma within the clinical context at the level of self, the provider, friends and family, and perceptions of DAA therapies. HCV stigma within the clinical setting was an important barrier to care and access to DAA therapies. Additionally, participants described experiences of differential treatment due to both HCV status and race. Speci cally, we found that Black/African American participants more frequently described experiences of racism in healthcare.
Taken together, our ndings suggest race and racism intersect upon patients' experiences of HCV-related stigma, shaping experiences with HCV and access to evidence-based curative therapies. Our nding that many patients with HCV experienced stigma in healthcare is consistent with pre-DAA era research, which reported that patients with HCV often experience both overt and subtle refusal of care across clinical settings (65)(66)(67)(68). Further, Black/ African American patients more frequently described discrimination and mistrust of providers, illustrating how racism in healthcare may intersect with HCV-related stigma. Treloar et al. (2013) identi ed patients' mistrust in providers as essential barriers at multiple stages of HCV care, including engagement with healthcare, uptake and adherence to treatment, and behavioral changes (69). Similar to previous research (70), we found that among patients with HCV who are highly stigmatized in healthcare, compassionate treatment by a provider is seen as exceptional, rather than expected. Alternatively, previous studies among patients with HCV found that building trust with physicians can help counter other experiences of stigma and promote positive behavioral change (69,71). We also found that positive experiences in healthcare lessen the burden of HCV-related stigma among patients, and patients who have been shunned by friends and family increasingly relied on healthcare as an important source of social support.
We found that stigma associated with DAA therapies, speci cally insurer and economic restrictions on access to treatment, reinforced perceptions of HCV-related stigma in healthcare. Societally, the high cost of DAAs has been associated with insurer treatment restrictions and reimbursement denial, effectively rationing curative therapies (28,29,72,73). Criteria for reimbursement vary across insurers and statebased Medicaid programs, often requiring evidence of advanced liver brosis and/or abstinence from alcohol or illicit drugs (28,74,75), which is inconsistent with current treatment guidelines (18,27).
Participants perceived restrictions of DAA therapy as symbolic of HCV stigma in society and healthcare. Our study adds to a growing body of literature that demonstrates the multiple levels of in uence at which HCV stigma and perceptions of differential treatment can further marginalize already vulnerable patient populations and be a signi cant barrier in the implementation of evidence-based treatment (66).
Stigmas are socially-produced and maintained through differential access to power and social inclusion (76). The downstream in uences of stigma on recipients and the innovation can produce disparate outcomes; when disease-related stigmas intersect with other experiences of stigma, such as racism or injection drug use stigma, the effect may become compounded. Woodward et al (4) argue that attention to the clinical encounter is particularly important in health equity frameworks, as the unique interactions within the clinical setting can have important implications for whether an evidence-based intervention is delivered. They argue that implementation science should measure societal in uences that have downstream impacts on the delivery of an intervention, particularly as they impact disparities in healthcare. Our ndings add to this by demonstrating the important role of stigma as a vehicle through which social factors impact the clinical encounter. Experiences of intersectional stigma in healthcare map on to the levels of the Health Equity Implementation Framework in a way that allows us to evaluate how societal in uences shape the clinical encounter, and ultimately, implementation success (Fig. 1b).
Frameworks are useful because they encourage the shared use of rigorous methods utilizing validated measures to advance implementation science research (77). Recognizing that stigma may be operationalized differently across levels of the model is necessary. While stigma may arise from the outer societal in uence, it is not relegated to this layer, and may penetrate deeper to directly incorporate into a patient's sense of self and impact interactions with others. Therefore, measures that assess stigma at the level of the self, provider, social circle, and innovation will allow us to measure stigma as a barrier to implementation in order to better understand how inequalities may shape delivery of evidence-based innovations.
A central aim of implementation science, is developing generalizable knowledge that can be widely applied beyond individually studied systems (1). We believe that the incorporation of intersectional stigma into the Health Equity Implementation Framework (4) has utility beyond the study of HCV. Intersectional stigmas reinforce disparities in healthcare through the pro ling of patients with substance abuse, emotional problems, and lower education as "di cult" and subsequent treatment of these patients as vectors of disease, criminals, or undeserving of health and social participation (43,(78)(79)(80). The stigmatizing label of a disease and associated assumptions about patients can dissuade medical engagement, impact delivery of service, and alter the course of illness (11)(12)(13)(14); investigation of intersectional stigma provides insights into why patients with the same disease may experience stigma differently (42). Research that contributes to our understanding of the role that social factors, such as disease-related stigma, play in the uptake of evidence-based practices can be applied beyond individual systems of study. This study has some limitations. As with all qualitative research, the amount and depth of information provided by participants depended on interactions with the interviewer, interview environment, and the subject's motivation for participating in the interview. Participant recall of past experiences may not re ect how they felt at that time, but rather be informed by their current perceptions associated with their illness identity. Finally, we only assessed stigma at the level of the patient. More research, including development of validated measures, is needed to better understand bidirectional stigmas between participants.

Conclusions
In conclusion, we found that disease-related stigma among patients with HCV intersects with other signi cant health disparities factors in ways that can hamper the equitable uptake of curative therapy. By incorporating intersectional stigma into the Health Equity Implementation Framework, future implementation studies may be able to better elucidate social in uences that promote health disparities and impact delivery of evidence-based innovations.

Declarations
Ethics approval and consent to participate This study was approved by the Institutional Review Boards of the University of Pennsylvania (Protocol, 829550) and Philadelphia FIGHT Consent for publication All participants provided verbal informed consent before participating in the study, which included consent to publish anonymous quotes from individual participants.

Availability of data and materials
The data are not publicly available because (1) they contain sensitive, identi able protected health information and (2) research participants did not consent to public release of their data. De-identi ed data may be made available to quali ed researchers by request of the corresponding author, the University of Pennsylvania O ce of Research Services (ORSMTA@pobox.upenn.edu), and the University of Pennsylvania Institutional Review Board (irb@pobox.upenn.edu). Requests will be considered on an case-by-case basis. If approved, researchers must submit a Material Transfer Agreement to the corresponding author and University of Pennsylvania O ce of Research Services. Additionally, research proposals must have an institutional review board-approved research protocol from their home institution.

Competing interests
The authors declare that they have no competing interests.  Abbreviations: HCV=Hepatitis C Virus, SD=Standard deviation; HIV=Human Immunodeficiency Virus P-values <0.05 are bolded *Qualitative data includes participants with interview and/or free-response data † P-value for characteristics of participants with any qualitative data (interview and/or free response), compared to study participants with no qualitative data ‡ P-value for characteristics of interview participants compared to study participants who were not interviewed Table 1 Legend: Characteristics are presented overall, for interview participants, and for participants with any qualitative data (interview and/or free response). P-values compare characteristics of study pool to interview participants and participants with any qualitative data.   "I feel like they didn't want to pour a lot of money into you either, do a lot of tests and something like that. Because, for what if you're not taking care of yourself and you just going to go out and die? That's like a waste of time for them and a waste of money and resources. That's how I feel. That's exactly how I feel like they look at you." (Black, non-Hispanic; HCV-SS Score, 90) Figure 1 In uence of Stigma within The Health Equity Implementation Framework. Fig 1a is The Health Equity Implementation Framework developed by Woodward et. al (2019). Within this framework, the clinical encounter is an interaction between recipients (patients, providers, and friends and family of who provide social support to the patient) and the innovation itself. Fig 1b revises The Health Equity Implementation