A typology of power in implementation: Building on the exploration, preparation, implementation, sustainment (EPIS) framework to advance mental health and HIV health equity

Abstract

Background

Persistent inequities in HIV health are due, in part, to barriers to successful HIV-related mental health intervention implementation with marginalized groups. Implementation Science (IS) has begun to examine how the field can promote health equity. Lacking is a clear method to analyze how power is generated and distributed through practical implementation processes and how this power can dismantle and/or reproduce health inequity through intervention implementation. The aims of this paper are to (1) propose a typology of power generated through implementation processes, (2) apply this power typology to expand on the Exploration, Preparation, Implementation, Sustainment (EPIS) framework to advance HIV and mental health equity and (3) articulate questions to guide the explicit examination and distribution of power throughout implementation.

Methods

This paper draws on the work of an Intermediary Purveyor organization implementing trauma-informed care and harm reduction organizational change with HIV service organizations. The expanded framework was developed through analyzing implementation coaching field notes, grant reporting, and evaluation documents, training feedback, partner evaluation interviews, and existing implementation literature.

Results

The authors identify three types of power working through implementation; (1) discursive power is enacted through defining health-related problems to be targeted by intervention implementation, as well as through health narratives that emerge through implementation; (2) epistemic power influences whose knowledge is valued in decision-making and is recreated through knowledge generation; and (3) material power is created through resource distribution and patterns of access to health resources and acquisition of health benefits provided by the intervention. Decisions across all phases and related to all factors of EPIS influence how these forms of power striate through intervention implementation and ultimately affect health equity outcomes.

Conclusions

The authors conclude with a set of concrete questions for researchers and practitioners to interrogate power throughout the implementation process.

Plain language summary

Over the past few years, Implementation Science researchers have committed increased attention to the ways in which the field can more effectively address health inequity. Lacking is a clear method to analyze how implementation processes themselves generate power that has the potential to contribute to health inequity. In this paper, the authors describe and define three types of power that are created and distributed through intervention implementation; discursive power, epistemic power, and material power. The authors then explain how these forms of power shape factors and phases of implementation, using the well-known EPIS (exploration, preparation, implementation, sustainment) framework. The authors draw from their experience working with and Intermediary Purveyor supporting HIV service organizations implementing trauma-informed care and harm reduction organizational change projects. This paper concludes with a set of critical questions that can be used by researchers and practitioners as a concrete tool to analyze the role of power in intervention implementation processes.

Background

Inequities in HIV and mental health

Despite biomedical advances in the last 15 years, the HIV epidemic in the U.S. has continued to be characterized by profound health inequity (Becasen et al., 2019; Centers for Disease Control & Prevention [CDC], 2020a; CDC, 2020b). Health inequity refers to unjust, avoidable, and systematic differences in health stratified by social location (e.g., race, gender, class, geography) (Arcaya et al., 2015) and is the result of social, economic, and/or cultural exclusion due to inequities in the distribution of power (e.g., marginalization, structural oppression, inequitable policies) (Baah et al., 2019; Sevelius et al., 2020). In the U.S., HIV health inequities are made visible by the disproportionate impact of HIV on marginalized groups, such as Black, Latinx, Transgender and Gender Non-Conforming (Trans/GNC) communities (Becasen et al., 2019; CDC, 2020a; CDC, 2020b).

Mental health among people living with and disproportionately impacted by HIV is one factor impacting HIV health outcomes (Remien et al., 2019). People living with HIV experience depression (25%–36%) and anxiety (19%–23%) at higher rates, compared to people not living with HIV (6.7% and 2.1%, respectively) (Beer et al., 2019; Bing et al., 2001; Do et al., 2014; Remien et al., 2019). For people living with HIV, challenges related to mental health can negatively impact treatment (Kendall et al., 2014; Pence et al., 2018; Yehia et al., 2015). Conversely, engaging in mental health care improves retention in care (Beckerman & Auerbach, 2010; Blashill et al., 2011), and health outcomes (Beckerman & Auerbach, 2010; Blashill et al., 2011; Mkanta et al., 2010; Yehia et al., 2015). Socio-structural factors, such as housing instability, community violence, poverty, social stigma, discrimination, and inadequate access to health care also contribute to increased risk for HIV, trauma, and poor mental health outcomes (Remien et al., 2019).

HIV-related mental health interventions have had a modest positive impact on mental health (Van Luenen et al., 2018), largely through integrating mental health care into HIV-related services (Chua et al., 2017; Farber et al., 2012) and providing mental health interventions specifically developed for people living with and at disproportionate risk for HIV (Heckman & Carlson, 2007; Hergenrather et al., 2013; Kempf et al., 2015). Despite this, inequities persist. This is due, in part, to barriers to effective implementation of HIV-related mental health interventions such as higher rates of attrition for participants of color compared to white participants, mental health and HIV stigma, and lack of intervention flexibility to meet community need (Chua et al., 2017; Edelman et al., 2016; Farber et al., 2012; Heckman & Carlson, 2007). HIV-related mental health interventions may also face implementation barriers found for HIV interventions generally, including challenges with client buy-in (Palinkas et al., 2015), untrustworthiness of providers (Beach et al., 2018), provider bias, and/or skills deficit (Esponda et al., 2020; Maulsby et al., 2017; Piper et al., 2020; Pleuhs et al., 2020; Wood et al., 2018), and mismatch between intervention and community (Beach et al., 2018; Pinto et al., 2018a). These implementation barriers disproportionately and negatively impact marginalized groups.

In fact, Freeman et al. (2017) found structural racism to be the overarching context that informed patient's barriers to engagement in HIV care. Structural racism refers to “macro-level systems that create, sustain, and reinforce inequities among racial and ethnic groups,” (p. 3) and includes unequal access to high-quality HIV care for communities of color, as well as systemic factors related to disproportionate HIV risk, such as housing instability (Aidala et al., 2016). Importantly, individuals who consider themselves to be nondiscriminatory can work within and perpetuate structurally racist systems (Freeman et al., 2017). For Black and Latinx people living with HIV, the historic context of maltreatment in medical research as well as present experiences of structural racism within implementation contexts, such as dehumanization within institutional settings, negatively impact engagement in care (Freeman et al., 2017). Achieving equitable health outcomes through HIV-related mental health interventions, therefore, necessitates thoughtful, equity-centered implementation.

Implementation science and health equity

Implementation Science (IS) offers tools to systematically investigate multilevel (individual, organizational, community, and macro) factors influencing the successful implementation of new practices (Bauer et al., 2015; Nilsen, 2015; Proctor et al., 2009; Rankin et al., 2016; Rapport et al., 2018)). IS concepts and strategies are organized and presented through a plethora of frameworks and conceptual models (Birken et al., 2017; Nilsen, 2015; Nilsen & Bernhardsson, 2019; Tabak et al., 2012).

Recently, scholars have begun to examine how the field of IS can more intentionally address health inequity (Baumann & Cabassa, 2020; Chinman et al., 2017; McNulty et al., 2019; Shelton et al., 2021a). A nuanced and multilevel analysis of varying contexts influencing intervention implementation is central to IS and provides a structure to examine specific mechanisms through which health inequity may be recreated (Baumann & Cabassa, 2020). Additionally, adaptation in implementation can improve fit between interventions and communities experiencing health inequity (Rathod et al., 2018). Tracking feedback from diverse implementation contexts can inform ongoing intervention evolution based on the needs and preferences of marginalized groups (Aarons et al., 2012; Chambers & Norton, 2016). Finally, recent research has identified strategies to more effectively center low-resourced communities in implementation studies (Yapa & Bärnighausen, 2018) and advance participatory implementation processes (Zimmerman et al., 2016) and research methodologies (Ramanadhan et al., 2018). Working with this momentum, researchers in the field have made calls to further deepen the relationships between IS and health equity by treating systemic issues, such as structural racism, as foundational to our conceptualization of implementation contexts, process, and measurement (Shelton et al., 2021a).

There are theoretical wells from which to draw to accomplish these goals. Snell-Rood et al. (2021) summarize important theoretical concepts that can inform the efforts of implementation scientists to analyze core dynamics of power that impact equity in implementation. Ford & Airhihenbuwa's “Public Health Critical Race” praxis (2010) also provides theoretical guidance for applying a critical race consciousness lens to public health research. Implementation scientists have begun to apply these and other equity lenses, evidenced by a trend of recent studies in which researchers use critical theories of power to layer on or augment existing implementation frameworks to deepen their analysis with regard to inequity (see Allen et al., 2021; Etherington et al., 2020; Woodward et al., 2019; Yousefi Nooraie et al., 2020).

This paper builds on this line of work by examining not only how implementation processes respond to and are influenced by power, but how power is generated and distributed by implementation processes themselves and how these forms of power can dismantle and/or reproduce health inequity through intervention implementation. Though relevant to a range of health problems, such an analysis is critical for health fields characterized by inequity, such as HIV and mental health. The aims of this paper are therefore to 1) offer a typology of forms of power that are generated through implementation, 2) apply this power typology to expand on the Exploration, Preparation, Implementation, Sustainment (EPIS) framework and 3) articulate questions to guide the explicit examination and distribution of power throughout implementation.

Expanded framework development context and process

The authors are the Director of Evaluation, Center Director and staff of an equity-focused Intermediary Purveyor and Funding Organization (IPO) working with HIV service organizations to build capacity around mental health, trauma-informed care, and harm reduction through funding, training, and implementation support. IPOs provide organizations with a range of resources including implementation planning and facilitating implementation strategies (Proctor et al., 2019). This IPO's work is guided by an advocacy group of people living with HIV and its partnerships with consultants from communities disproportionately impacted by HIV. The following power typology and expanded framework were derived from information collected over three years of partnering with HIV service organizations to implement trauma-informed care and harm reduction organizational change projects, including implementation coaching field notes, grant reporting and evaluation documents, training feedback, and partner evaluation interviews. The power typology emerged through ongoing and iterative analysis between the authors regarding the movement of power within the implementation.

Typology of implementation power

This paper argues there are three distinct yet interrelated forms of power that are generated through intervention implementation; discursive power, epistemic power, and material power. Discursive power refers to ways in which dominant perspectives ascend to their position as reality and shape the actions and capabilities of others (Foucault, 1972). This is the power to create and promulgate health narratives. Discursive power is enacted in implementation through defining the health-related problem to be targeted by intervention implementation. Intervention implementation processes also generate narratives that shape our understanding of health dynamics. Narratives of resistant or hard to reach populations are an example of one such implementation narrative that may contribute to health inequity (Martinez et al., 2014). Epistemic power is the ability to assert one's knowledge as integral to a collective set of beliefs or truths (Geuskens, 2018). Epistemic power is generated through decisions regarding whose knowledge is valued in decision-making. Epistemic power is also enacted in the selection of evidence used to inform implementation decisions and who deems that evidence as valid in implementation contexts. Epistemic power is created through knowledge production and dissemination throughout implementation (i.e., what knowledge is collected and valued to inform future action). Finally, intervention implementation inevitably involves resources, the control over and disbursal of which will create microsystems of material power in organizations, communities, and regions. Material power is also enacted through who successfully accesses the health resources and acquires the health benefits provided by the intervention. Decisions made throughout implementation will influence how these forms of power striate through intervention implementation and ultimately affect health equity outcomes.

Interrelated HIV health and mental health inequity is driven by the ways in which power operates through gender, class, race, and sexuality often referred to as intersectionality (Bowleg et al., 2013). In the U.S., HIV health and mental health inequity is characterized by White Supremacy, structural racism against Black and Latinx communities and intersecting oppression related to sexual orientation and gender identity/transphobia (Becasen et al., 2019; CDC, 2020a; CDC, 2020b). To keep this expanded framework broad enough to analyze different intersecting power dynamics influencing intervention implementation in different contexts, the authors do not center on any specific set of targeted identities or domains of sociopolitical marginalization.

The Exploration, Preparation, Implementation, Sustainment (EPIS) framework

The EPIS framework is a conceptual model which describes phases of intervention implementation (exploration, preparation, implementation, and sustainment) and articulates factors shaping implementation at every phase (Aarons et al., 2011; Moullin et al., 2019). EPIS is both a determinant and process framework which also addresses implementation mechanisms. Inner context factors refer to the internal context within which implementation takes place, such as organizational structure, culture, leadership, and staff preparedness. Outer context factors refer to the service environment including the policy landscape, networks of other organizations, and statutory obligations. Bridging factors span inner and outer contexts and refer to a range of partnerships and connections that facilitate implementation, including academic-community partnerships, private–public partnerships, and implementation consultants. Innovation factors refer to specific characteristics of the intervention/innovation itself that may impact implementation and the intervention's fit with the organization. A recent systematic review demonstrated that the EPIS framework has been used across various settings internationally to address a diverse set of issues (Moullin et al., 2019). The EPIS framework is particularly suited to supporting an analysis of power due to its focus on interconnections between phases and factors. Table 1 presents the power typology as well as several overarching questions pertaining to EPIS.

Table 1.

Typology of implementation power.

Power type	Exploration	Preparation	Implementation	Sustainment
Discursive Power: Ways in which dominant perspectives ascend to their position as reality and shape the actions and capabilities of others –Defining health problem to be addressed –Development of health narratives	How is the health problem to be addressed defined?	What are organizational narratives related to the health problem? How might unconscious bias or stigma, e.g., impact implementation?	What narratives are being generated through intervention implementation? How and with whom are they being shared?	What equity-related insights resulted from implementation? How might they contribute to broader organizational narrative shifts?
Epistemic Power: The ability to assert one's knowledge as integral to a collective set of beliefs or truths –Whose knowledge is valued throughout the implementation process –What data is collected and how it is used	Whose knowledge is valued in problem conceptualization and intervention/innovation selection? Why?	Was the intervention developed with community knowledge? What knowledge can be brought in to adapt the intervention to meet community need?	Do data structures exist to elicit candid community and staff feedback regarding intervention experience? What data is collected and whose goals does it serve?	What intervention-specific knowledge structures can be sustained and potentially expanded (to spur epistemic power shift in organization)?
Material Power: Control over and distribution of critical resources –Distribution of resources related to intervention implementation –Access to/distribution of intervention benefits	What organizations, communities and individuals are being invested in and why?	How will structural inequity be addressed to maximize access to intervention benefits?	Who is benefitting from intervention implementation? Who is not? Why?	On whom is the intervention dependent for funding? How does this impact ongoing implementation?

EPIS: Integration of an analysis of power

In this section we present (a) a description of how discursive, epistemic and material power influence each phase of the EPIS framework; (b) an articulation of how power manifests in each domain of the EPIS phase (inner/outer/bridging/innovation) (Moullin et al., 2019); and (c) examples of analyzing and addressing power in implementation.

Exploration

In the exploration phase, stakeholders convene to identify a health need that will be addressed by an intervention/innovation. During this exploratory stage, stakeholders prioritize the issue to be addressed, frame the defining features of the issue, and explore interventions available to them (Aarons et al., 2011).

Role of power in exploration

Power is enacted through these early decisions. Though structural explanations have garnered increasing attention due to clear and durable HIV and mental health inequity (Gupta et al., 2008; Panagiotoglou et al., 2018), dominant narratives of HIV and mental health still privilege individual-level behavioral perspectives. These narratives are a currency of discursive power that shape and are shaped by how implementers define the problem to be addressed and actions to be taken in response. Material power is also associated with decisions regarding the early allocation of resources, for example, who is paid through the project and what communities/organizations share in project resources. Finally, epistemic power is enacted through defining and identifying evidence guiding implementation, as well as through whose knowledge shapes the boundaries of intervention options and who decides which actions are pursued (Shelton et al., 2021b).

Inner context

One key factor of the inner context is project leadership. By paying attention to who is at the table, implementers shape how issues are prioritized and defined. Involving prospective intervention recipients into this process can impact the extent to which an intervention addresses health inequity (Rhodes et al., 2020). Participatory implementation processes have not been widely studied, but the role of contributions from a range of stakeholders—including organizational staff, leadership, and community members who will participate in or benefit from the intervention—has begun to be considered (Ramanadhan et al., 2018).

Too often participatory approaches are introduced after core decisions have been made, limiting the effectiveness of community participation on health equity outcomes (Wallerstein et al., 2019). In implementing trauma-informed organizational change projects the authors have found that when community participants inform implementation processes—starting at the exploration phase—projects focus more intentionally on multilevel and collective trauma, including trauma resulting from racism, anti-Blackness, and gender-based structural violence. This instigates a shift in the discourse of trauma and opens new opportunities to change organizational culture to better meet the needs of marginalized groups, for example, by focusing on holistic and collective healing rather than mental health more narrowly defined.

Outer context

Outer context factors interact in a complex web of sociopolitical determinants operating at multiple levels (federal, state, local) and through multiple mechanisms (e.g., policy, funding) (Moullin et al., 2019). Prevailing discourses, material resource limitations, and dominance of certain epistemic perspectives will circumscribe intervention options available and impact the success of intervention implementation. Outer factors such as policy and funding govern how organizations identify the root issues to intervene on, define those issues, and identify the types of interventions that are possible to implement. Epistemic power, which is primarily seated in academic institutions, research centers, and scientific funding organizations determines what counts as evidence informing evidence-based interventions (i.e., the type of interventions that are being tested, how and who they are tested on).

The IPO has found that most evidence-based interventions produced by this system and available for implementation by HIV service organizations related to mental health are individualistic/behavioral in nature, though this is beginning to change (Pinto et al., 2018a). Many social determinants of health scholars point to the importance of upstream interventions to address structural barriers obstructing health equity (Agurs-Collins et al., 2019). The developing research on multilevel interventions highlights feasible mezzo and macro interventions which can be a stand-alone project and/or complement individual-level interventions (Agurs-Collins et al., 2019; Berkley-Patton et al., 2019; Kegeles et al., 2015; Pinto et al., 2018a). Multilevel interventions may address client advocacy, which is an outer context factor. Client advocacy refers to mechanisms that create systems change based on client needs and priorities (Moullin et al., 2019) in ways that may redirect material and discursive power. For example, the IPO partnered with an organization that embedded ongoing state-level policy advocacy into their syringe service programming by using implementation and outcome data to argue for harm reduction friendly policies. This ultimately expanded the potential of implementing harm reduction services both for the organization and regionally and introduced new material resources (i.e., funding) to support vulnerable communities.

Bridging factors

The exploration phase provides an opportunity to build partnerships for successful implementation. Common partnerships involved in intervention implementation include academic partners (Uvhagen et al., 2018), IPOs, and other consultants. Implementers can vet potential partners based on their relationship with health equity. For example, the authors assess consultant partners to understand their perspectives on structural determinants of health and analysis of power. Additionally, many of the IPO's consultants are reflective of communities disproportionately impacted by HIV and/or have lived experiences relevant to trauma-informed care and harm reduction approaches. Organizational staff who receive coaching and training from these partners regularly report that consultant's combination of lived and professional expertise heightens organizational commitment to equitable implementation. Directing significant material resources to people with lived experience is also a transfer of material power to the community and an act of economic justice (Stanton & Ghose, 2017).

Innovation factors

A key outcome of the Exploration phase is the identification of the intervention that will be implemented. Those making this selection can consider whether a given intervention was developed in collaboration with community (O’Mara-Eves et al., 2015), as well as whether the intervention itself integrates an analysis of systemic power into its theory of change (e.g., Maiorana et al., 2020; Stein et al., 2015). Both criteria may lead to intervention selections aligned with an equity-centered approach.

Preparation

The preparation stage is when those involved with implementation develop strategies to integrate an intervention into the organization's systems, structures, and culture. It is the time to assess and plan for implementation barriers and facilitators, develop intervention policies and procedures, and train staff, among other preparatory activities (Aarons et al., 2011). Throughout the preparation, continued attention must be given to the ways in which outer and bridging factors may support or impede equity in inner context implementation strategies. Preparation is also when the organization will develop structures of data collection, analysis, and use (Aarons et al., 2011) and systems-level preparation to enhance scale-up and sustainment (e.g., Hurlburt et al., 2014).

Role of power in preparation

Evidence-based mental health interventions will only reduce HIV health inequities if individuals who are disproportionately impacted by HIV access and benefit from these interventions. Inattention to structural barriers to care, therefore, drives health inequity through misallocation of material power, because people from marginalized groups will be obstructed from receiving intervention benefits. Determinants of intervention accessibility to marginalized groups include a range of inner factors, such as staff expertise in working with communities impacted by health inequity, and outer factors, such as structural barriers to care. Additionally, epistemic power is enacted through decisions around implementation and outcome data collection. These decisions regarding data will ultimately inform the project narrative, which will generate discursive power.

Inner context

During the preparation phase, it is important to gauge the ability of staff and leadership to equitably serve communities that will receive the intervention. Healthcare engagement has been found to be impeded by medical/provider mistrust (Brincks et al., 2019; Shelton et al., 2021b), stigma (Geter et al., 2018), discrimination (Galvan et al., 2017), language injustice (Thonon et al., 2021), institutional betrayal (Selwyn et al., 2021) and culturally dissonant or hostile environments (Bockting et al., 2020; Lacombe-Duncan et al., 2020). Organizations must be prepared to engage with marginalized clients in a way that recognizes and addresses these issues. Metzl and Hansen’s (2014) concept of structural competency refers to a healthcare provider's ability to recognize social and economic forces impacting individual health symptomatology, integrate this knowledge into clinical recommendations and develop their efficacy in promoting structural change. Preparation for intervention implementation with marginalized groups can include training on structural competency to develop staff capacity to address power.

The authors have found that preparing staff to discuss power is essential to the successful implementation of trauma-informed organizational change. The IPO's approach to trauma-informed HIV care emphasizes the role of racism in trauma experienced by communities disproportionately impacted by HIV and challenges organizations to address structural racism as a core component of trauma-informed organizational change. This approach is largely welcomed by staff in our training and consultations. The authors also encounter staff who do not feel prepared to apply this analysis to their own work and few who perceive this as an inappropriate intrusion of politics into their work. Intentional staff training can prepare organizations to facilitate the difficult conversations often required in equity-centered work. Explicit discussion of power and equity can also be built into ongoing implementation processes such as clinical supervision, coaching, and fidelity assessment.

Evaluation planning, if it is to be included, also occurs during this phase. Evaluations of implementation efforts pose considerable challenges due to the complexity of implementation processes in dynamic organizational settings (Murdoch, 2016). However, implementation evaluation can open up the black box connecting intervention design and health outcomes through identifying and assessing the presence, timing, dosage, and impact of implementation strategies (Huynh et al., 2018) and assessing the influence of outer context factors on outcomes (Mendel et al., 2008). Huynh et al.’s (2018) model of strategy mapping for complex interventions, for example, provides a process for mapping the roll-out of implementation strategies and tracking their impact on intervention outcomes. This approach provides an opportunity to understand the relationships between specific strategies and outcomes for communities experiencing health inequity, enhancing the specificity with which implementers can analyze the role of power in influencing the implementation and its impact on health.

Outer context

Project funders influence the movement of material power. Funders often have restrictive timelines for project preparation due to political pressures, health system, or fiscal requirements. Urgency to launch interventions or achieve community reach benchmarks is misaligned with community-based approaches and attentiveness to power. Urgency can result in the exclusion of communities and obstruct the establishment of functional and empowered community boards. Funding must include support for preparation phases to generate community-centered responses to health problems and equity-centered systems.

Additionally, an outer context analysis of the policy landscape, social environment, and built environment is essential to informing intervention selection. For multilevel interventions, this analysis will inform upstream actions related to policy, social norms, and community mobilization. This analysis is also relevant to preparation because it illuminates structural issues external to the organization that must be addressed to facilitate community access to intervention. Common issues related to lack of material power such as transportation, inability to meet during business hours, childcare, and cost will require a strategic implementation response.

Bridging factors

Sustainable intervention implementation often requires system-wide change and coordination across inner and outer contexts (Hurlburt et al., 2014). Taking a system-level view in the preparation phase may reveal opportunities to integrate effective bridging strategies to promote equity. The authors worked with organizations implementing trauma-informed organizational change who built regional trauma-informed care training and train-the-trainer models into their implementation plans. This represented a reallocation of material power, in which their project funds were used to build the capacity of others. This bridging approach was deemed worth the cost to enhance the consistency of trauma-informed care across service systems. These bridging strategies can counter health inequities that are reproduced when competition pressures and material power scarcity narratives result in service gaps, poor service, and redundancies (Pinto et al., 2017).

Innovation factors

During the preparation phase, it is important to assess intervention adaptation, or the process of modifying an evidence-based intervention to better fit participant's experiences, needs, and preferences based on culture and/or locality (Stirman et al., 2019), while maintaining theoretical congruence to the intervention's core components (Castro & Yasui, 2017). Adaptation should take place with meaningful involvement from intervention recipient communities. Well-planned adaptation may contradict stigmatizing implementation discourses. For example, Martinez et al. (2014) adapted an HIV intervention to include social media-based recruitment strategies and found that Latino gay couples were not a hard to reach group—as they have often been labeled—if culturally appropriate outreach strategies are employed.

Implementation

The intervention is launched in the implementation phase. Implementation is a time of ongoing monitoring, learning from doing, and course correcting as implementation stakeholders see how the unique interplay of inner and outer contexts impacts implementation and intervention outcomes.

Role of power in implementation

Ongoing assessment is needed throughout the implementation phase to recognize and address unexpected ways power impacts intervention implementation. It is a time to assess whether attention to power in the exploration and preparation phases informed an implementation strategy that facilitates community access to intervention health benefits. Shifting landscapes of epistemic, discursive, and material power internal and external to an organization will impact ongoing implementation and health equity. Without mechanisms to analyze these issues, organizations run the risk of providing interventions on autopilot while equity gaps grow.

Inner context

Stable internal processes and protocols are needed to gather and interpret information related to implementation. Though plans for implementation tracking—and the epistemic power that process generates (what is being tracked and why)—are developed in the preparation phase, it is the implementation phase during which organizations will see how the implementation is impacting clients and staff. Individuals involved in implementation can use data to monitor for barriers to access and course correct accordingly. Drawing on an example from trauma-informed care implementation, staff can be monitored for symptoms related to stress and/or vicarious trauma throughout implementation (Branson, 2019). Such stress may lead to symptoms of burnout which may compromise intervention quality as well as staff physical and mental health and can be addressed at the organizational or systemic level. Implementation science research has identified qualitative methods, such as periodic reflections (Finley et al., 2018) for collecting on-the-ground data to inform action.

Outer context

Outer context factors will not calcify once an organization has made an implementation plan. Events external to the organization will shift in ways that impact implementation. The policy context, in particular, can profoundly impact and be impacted by changes in health discourses related to the issue being addressed, evidence being presented and accepted, and availability of resources to address the issue (Bruns et al., 2019; Mendel et al., 2008; Powell et al., 2016). Stakeholders must track how these power dynamics change and assess the ways they may impact sustained implementation and client outcomes. Shifting social consciousness may also open windows of opportunity for health equity work. Current social uprisings in response to state violence against Black communities have, for many, (re)ignited conversations on structural racism. This moment of heightened awareness can be translated to meaningful organizational change related to HIV, racism, trauma, mental health, and White Supremacy and may impact intervention implementation. The IPO has worked with this moment, for example, to deepen conversations related to racism (specifically anti-Blackness), HIV, and trauma with its organizational partners. IPO partners have identified opportunities for community and staff to engage in direct action. The IPO has also recognized how recent events have activated trauma responses among some staff and clients and has provided resources and guidance to support organizational efforts to promote individual and collective healing. Integrating these outer context factors into intervention implementation will depend on the implementation staff's comfort and competence addressing power and their willingness to expand what is encompassed by the intervention. The decision to address structural racism, trauma, and HIV, or not, is in and of itself an intervention of discursive power.

Bridging

Structures must exist to channel information about outer context changes to implementers. Consistent, relevant community insight will not be achieved through one-off focus groups or rubber stamp community advisory boards, but through meaningful epistemic power-sharing. In addition to changes at the community level, implementers must have access to knowledge regarding changes at the funding and policy levels to understand whether there will be ongoing resources to achieve intervention sustainment. This further points to the need for noncompetitive interorganizational collaboration.

Additionally, during implementation, the ongoing match with bridging partners must be assessed. Necessary changes to implementation-related staffing can be made with transparency and integrity. Ongoing coaching should integrate discussion of health equity, access, and power to support ongoing critical reflection and analysis.

Innovation

During implementation, the fit between intervention adaptations and community need and preference will be tested. Intervention fidelity is increasingly seen as a balance between core components and community context-specific alterations (Cabassa & Baumann, 2013). Finding and documenting this balance informs the ongoing evolution of evidence-based practice, but only if adaptations are thoughtfully measured in a way that informs tailoring processes and their meaning and relevance for stakeholders (Aarons et al., 2012; Chambers & Norton, 2016; Stirman et al., 2019). Additionally, there may be an element of the implementation structure related to power (such as systematic checks of provider bias or community feedback mechanisms) that may be integrated into the standard evidence-based practice package if proven to improve equity outcomes, further equalizing the epistemic power between academic intervention developers and communities participating in interventions.

Sustainment

During the sustainment phase, the intervention becomes a fully integrated element of the organization. This includes the identification of stable funding and ongoing quality assurance. If not considered before, sustainment is a time to consider scale-up of successful interventions (Milat et al., 2013) or system-wide change.

Role of power in sustainment

Intervention implementation which is unresponsive to dynamic changes on the ground will become increasingly ineffective with community members. Conversely, intervention/implementation context fit and effectiveness can be strengthened throughout sustainment by promoting a culture of organizational learning and dynamic sustainability (Chambers et al., 2013). Long-term structures of epistemic power sharing will provide mechanisms to communicate community knowledge and facilitate nimble response to community change. Organizations also often share outcomes of intervention implementation when seeking further funding or as a part of their social marketing efforts. The narratives about HIV, mental health, and community that are created during the sustainment phase are vehicles of discursive power and must include voices of those most impacted. Finally, organizations’ ability to navigate the funding landscape and/or advocate for funding will largely determine the extent to which they have the material power to sustain intervention implementation.

Inner context

As intervention implementation stabilizes, there may be opportunities to consider how health equity and power sharing lessons learned from implementation may be relevant to larger organizational structures. Intervention-specific analysis of power and small-scale structural change can be a gateway to sustained organizational change. For example, the IPO supports organizations in applying Meaningful Involvement of People Living with HIV/AIDS (MIPA) (AIDS United & U.S. People Living with HIV Caucus, 2019) principles into intervention planning and implementation. Several of these organizations have used this experience as a launching pad to examine larger organizational structures that may be strengthened by including people with lived experience in leadership roles for a more equitable distribution of epistemic, discursive, and material power across the organization. Using a structural lens to drive health equity-focused implementation may lead to organizational changes that are sustainable beyond the intervention itself, resulting in a profound inner context shift that will impact the implementation of future interventions.

Outer context

Funding is a core outer-context factor related to sustainability and material power. Dependence on unpredictable funding makes it unclear whether programs will be sustained long enough to integrate adaptive learning beyond single grant cycles and research projects. These obstacles to long-term planning create an unequal material power dynamic because organizations—and by extension communities—are unable to plan for a dynamic future. It therefore becomes difficult for community health programming to evolve according to community need. A key to sustaining health equity–focused intervention implementation will be for systems-level leaders to understand the funding needs of equity-focused implementation and champion sustainment (Aarons et al., 2016). In the absence of significant changes to funding structures, organizational networks may be able to leverage partnerships to account for resource limitations and sustain interventions through creative redistribution of material power (Pinto et al., 2018b).

Bridging factors

Organizations can become regional leaders in applying a health-equity implementation framework. The IPO has worked with many organizations that have implemented HIV-related mental health interventions with a structural lens and go on to organize regional implementation strategies and train other organizations. Scarcity of funding often drives a sense of competition and keeps community-based organizations from being transparent with one another about their inner workings. Transparency about the challenges and benefits of addressing power in implementation can therefore be a discursive intervention that impacts HIV health and mental health narratives.

Innovation

Organizations must recognize when a chosen intervention is not a good fit or adaptation efforts have not successfully met community needs. Emerging research on de-implementation looks at decisions to cease intervention implementation and processes to ethically extract (Pinto et al., 2018a, 2018b; Prasad & Ioannidis, 2014). When emerging scientific evidence necessitates changes to implementation or de-implementation, implementers will rely on community trust to support the changes and not further dynamics of medical mistrust (Shelton et al., 2021b). Power sharing strategies discussed throughout, including involving community members who will be affected by the change, may assist in this process.

Conclusions

If power is not addressed in implementation, HIV-related mental health interventions will fail for marginalized groups, preventing access to potentially lifesaving care. This expanded framework offers a shared conceptual language for implementation stakeholders to identify and address the ways in which discursive, epistemic and material power are generated by and influence implementation processes. We recommend IS scholars and practitioners explicitly assess the influence of power on implementation processes and the forms of power generated through implementation. We offer in Table 2 critical questions to guide them. Attention to these issues can result in meaningful changes to implementation planning and execution which enhance the health equity potential of evidence-based interventions in community settings.

Table 2.

Critical questions to assess and address power in implementation.

EPIS phase	Critical questions
Exploration
Inner Context	Who holds decision-making power/leadership for the initiative? Are individuals who are impacted by the health issue of focus meaningfully involved? Are these individuals meaningfully compensated for their labor?
Outer Context	What factors outside the organization will impact innovation/intervention selection? How does the policy environment influence the identified issue? The intervention/innovation options? How might the built environment, sociopolitical and cultural environments impact innovation/intervention selection? ▪ e.g., Stigma; Racism What role do funders play in the social construction of the issue at hand? Is a multilevel intervention feasible? (Is there a need/role for advocacy and/or structural change in your intervention plan?)
Bridging Factor	What is the perspective of those you are partnering with (academic institutions, IPOs, Trainers and Consultants)? Do they integrate a multi-level approach or perspective? Do they have experience working in partnership with the community? Are their goals aligned with yours and the community’s? What other organizations in the region may be doing similar or intersecting work? How will your organizations’ intervention decisions impact this larger landscape?
Innovation	What is the range of interventions/innovations/actions available to address the selected issue? Have any of them been developed in similar contexts, with similar populations as yours? Do any of them intentionally integrate relevant issues of power related to the selected issue? (e.g., racism, gender-based violence, homophobia, transphobia, unconscious bias)
Preparation
Inner Context	What dynamics exist within your organization that may impact the community's ability to access intervention benefits? Do community trust and feel welcomed by your organization and staff? Is the community supported in your space? Physically through building accessibility? Emotionally and culturally? Linguistically? What training might staff need to support each other and the community for an intervention implementation that successfully benefits marginalized groups and individuals? What mechanisms can be developed to facilitate ongoing support and accountability for staff as implementation proceeds? Do these include mechanisms to collect and integrate community feedback?
Outer Context	How will your organization address structural barriers external to your organization that may impede access to your intervention and/or obstruct community members from actualizing full intended health benefits of intervention? Are you developing structures to collect the information necessary to identify emerging inequities in access and outcomes and their potential influencing factors?
Bridging Factors	Have you assessed your network of referrals for alignment with the community's values, priorities and preferences? Are there opportunities to develop networks to explore system-wide change?
Innovation Factors	Does your intervention of choice require adaptation? Does your adaptation plan meaningfully integrate community knowledge?
Implementation
Inner Context	Do you have a plan to support staff in working with marginalized groups and addressing power? Do you have mechanisms to collect and integrate ongoing feedback from community members participating in the intervention? Have you established a culture of trust and transparency to receive uncensored feedback from the community and staff without fear of reprisal?
Outer Context	What evolving dynamics at the organizational, neighborhood, regional and national level may impact ongoing implementation? What new barriers may emerge? What policy windows may expand possibilities? Do you have necessary information network channels to access this information as it emerges?
Bridging Factor	Do you continually reassess your partnerships, their utility and what additional partners may be needed?
Innovation	How might your organization's experience with implementing the intervention impact ongoing intervention development? What is your organization and community learning from the implementation process that may enhance health equity potential of the intervention if shared with the broader field?
Sustainment
Inner Context	What lessons has your organization learned about pursuing health equity through intervention implementation that may have implications for structure and culture of the larger organization?
Outer Context	What are the larger sociopolitical factors impacting the sustainability of the intervention? What funding sources may support your ongoing, dynamic engagement with intervention implementation? Have new potentialities for multilevel interventions emerged since project inception?
Bridging Factor	What can your role be in the region regarding health equity? Can you become a bridging factor for other organizations seeking more equitable intervention outcomes?
Innovation	Do you keep vigilant in measuring health equity outcomes? Do you have a plan to assess the need for de-implementation and an ethical de-implementation plan? Are you prepared to grow with the developing intervention evidence?