How power is conceptualized and operationalized in evidence-based intervention implementation: a scoping review

Abstract

Background

The potential for implementation science (IS) to address health inequities is limited by insufficient attention to power. Although the field emphasizes context, it remains unclear how best to examine and intervene on power relations related to the implementation of evidence-based interventions (EBIs). This scoping review aimed to determine: 1) To what extent do research projects studying EBI implementation explicitly examine power? 2) In these studies, how is power conceptualized, defined, operationalized, and understood? 3) What opportunities exist to examine and intervene upon power through IS research?

Methods

Following established procedures, we undertook a six-step process: 1) articulating research question and purpose; 2) identifying relevant studies; 3) selecting studies; 4) extracting data; 5) summarizing data; and 6) reporting results. We characterized studies’ attention to power using Fung’s power framework, which attends to everyday, policy, structural, and ethical power. Based on publications available as of February 2022, we included English-language EBI implementation studies from clinical, community, and public health settings that explicitly attended to power. Data extraction included study context, IS frameworks used, definitions and measures of power, and characterizations of how power influenced implementation processes and outcomes.

Results

Of 3,531 articles screened, 28 papers explicitly discussed power in relation to EBI implementation and 11 presented a formal definition of power. Most studies explored everyday and structural power, with far less attention to policy power and almost none to ethical power. Conceptualizations and operationalizations of power varied widely, and few studies reported grounding in IS frameworks. Explicit strategies to intervene upon power were limited. Most studies focused on short-term integration goals, with limited discussion of how power dynamics shape what counts as evidence, whose interests are served, or opportunities for systems transformation.

Conclusions

Researchers have the opportunity to explicitly integrate power theories and frameworks into conceptual models for IS studies to reshape IS efforts and build this evidence base. This will allow the field to move towards critical, systems-focused perspectives that examine how power shapes and can be used to reshape the evidence base, implementation systems, implementation strategies, and implementation, health, and system outcomes.

Supplementary information

The online version contains supplementary material available at 10.1186/s13012-026-01511-5.

Contributions to the literature

• This scoping review highlights the dearth of literature explicitly addressing power in the context of the implementation of evidence-based interventions • The manuscript applies Fung’s liberatory framework and four levels of power (everyday, policy, structural, and ethical) to identify the extent to which and how power is studied in implementation science • The review highlights opportunities to integrate theories and frameworks of power into implementation science projects to identify how power shapes (and can be used to shape) implementation systems and outcomes

Introduction

There are numerous calls to leverage the force of implementation science (IS) to address health inequities, thereby integrating a systems-focused approach in efforts to translate knowledge into routine practice in public health and healthcare settings. With this orientation, IS efforts must consider a wide range of factors as intervention targets, from the active engagement of key partners to political environments, so that implementing systems can be reshaped and leveraged to promote equity [1, 2]. This is a contrast to the norm to either exclude aspects of social systems that preserve current patterns of inequities or assume that those broader patterns will remain constant and cannot be intervened upon [3]. Recent examinations of the IS field have highlighted the emphasis on improving practice and health outcomes without sufficiently situating implementation efforts within broader historical, political, social, and cultural contexts [4, 5]. As Cooper and colleagues describe, ignoring macrosocial forces not only leads to ineffective solutions but can also unintentionally exacerbate inequities. In contrast, examining such forces supports the ability to understand patterns of how opportunities and hazards are distributed and attend to changing those distributions [6]. Here, we focus on power in the implementation context. While IS theories, models, and frameworks (referred to hereafter as TMFs) are attentive to understanding the context in which research is implemented, the field has yet to engage fully in examining how power dynamics play out in such settings and systems, and related impacts on health outcomes and inequities [2, 7, 8]. Instead, evidence-based interventions (EBIs—including practices, programs, policies, and more) are often used within a hierarchical knowledge economy that emphasizes technical solutions to discrete health problems, privileging economic gain and the maintenance of the status quo [9–11]. Our interest in this topic hinges on the premise that shifts in power (e.g., related to authority and access to resources) could lead to more equitable implementation, services, and health outcomes. One can imagine a situation in which authority and decision-making sit with a small group of (often external) actors, who have the ability to decide which sites receive an EBI, what criteria are used to deem interventions as ‘evidence-based,’ how much authority implementers and other key constituents have over EBI implementation and adaptation, how resources are distributed, and ultimately which implementation, service, and health outcomes are prioritized. However, using a power analysis reorients researchers away from assumptions that these patterns must remain constant, instead prompting attention to how implementation conditions were shaped, whose interests are served, and what opportunities exist to create more equity-promoting systems.

Power can be understood as a dynamic force that allows actors to achieve their goals and bring about political, social, and/or economic change [12]. Power can be controlling/dominating or liberatory/transformative [13]. While an exhaustive review of conceptualizations of power is beyond the scope of this review, we present a core set of ideas and definitions that are particularly relevant and actionable. First, although many discussions of power emphasize the notion of power over others [14], a liberatory frame emphasizes power with and power to, as well as opportunities to use power to push against the status quo [9, 13, 15, 16]. Second, in the context of IS, it is helpful to consider spaces in which power can be deployed, including closed spaces (in which decision-makers wield power without inclusion of those affected), invited spaces (in which a subset of actors are invited to participate in decision-making), and claimed or created spaces (those claimed by those with less power in a given system or those who have mobilized to address an issue) [17]. This relates to questions of who constitutes the dominant group wielding power in a given implementation context and who is marginalized. Finally, participatory and engaged research approaches (e.g., community-engaged research, community-based participatory research), which are inherently rooted in power-sharing, are central to IS efforts to address inequities [1, 2, 18]. Diverse conceptualizations of power offer useful ways to understand power from different analytic angles. For example, Lukes’s work is critical for identifying visible and hidden forms of domination [19]. Stanton and colleagues describe the types of power implementation efforts can generate, including discursive power (e.g., directing what implementation targets are focused on and the narratives around these targets), epistemic power (e.g., whose knowledge and what types of evidence are valued or discounted in decision-making processes), and material power (e.g., access to and control over resources that support or constrain implementation) [7].

In this review, we drew primarily on Fung’s framework not only because it distinguishes important levels at which power operates, but also because it explicitly attends to identifying leverage points for change [20]. The four levels are organized from the individual (the subject of liberation) outward. The first level is everyday power, which centers on common interactions that challenge individuals’ interests. The second level covers laws and policies from governments and other institutions that influence individuals’ pursuit and advancement of their interests. The third level is structural power, which includes the parameters and the environment that shape organizations’ ability to advocate for individuals at the first level through laws and policies at the second level. The fourth level is ethical power, meaning the ideals, values, narratives, and societal norms that shape individuals, institutions, and their perceptions [20]. Fung’s framework focuses not just on analyzing domination but on enabling liberation, aligning with our goal of identifying leverage points for change within complex systems. In other words, we can consider not only how existing forms and levels of power impact implementation of research, but also how implementation strategies and processes can garner and distribute power in implementing systems, with consequences for promoting or inhibiting health equity [7]. Notably, the four levels are not intended to be addressed linearly. Change can arise at any level or order and would ideally co-occur and be aligned across levels. The most transformative change – at the fourth level, ethical power – is considered the most challenging to achieve. As seen in Fig. 1, Fung’s description of levels of power can serve as prompts to interrogate IS research systems.

Fig. 1.

Application of Fung’s four levels of power [13] to EBI implementation

Advancing health equity requires implementation scientists to examine and engage with power intentionally—recognizing how implementation research can either perpetuate inequitable power structures or serve as a vehicle for redistributing power in ways that promote equity [7]. A recent review of the integrated knowledge translation literature, which emphasizes engagement with knowledge users, identifies gaps in how power is defined, theorized, and operationalized. Similar challenges are noted across related fields such as patient and public involvement, where conceptual ambiguity and inconsistent application hinder meaningful assessments of power dynamics. These findings highlight a need within IS research to more rigorously incorporate the study of power, including more precise definitions, deeper theoretical grounding, and refined methodological approaches, to understand and address equity implications of implementation processes fully [21, 22].

The extent to which the IS evidence base examines and addresses the role of power in implementation processes and outcomes, as well as ways to intervene upon inequitable power structures, is unclear. It is also unclear to what extent and how a liberatory perspective, which aims to shift systems towards equity and justice, has been taken. In this review, we treat power as the resources and capacity held by key actors to shape implementation processes and systems (e.g., norms, decisions, relationships, and the knowledge base) in ways that impact the effects of implementation activities. This scoping review focuses on answering three key research questions: 1) To what extent do research projects focused on studying the implementation of EBIs explicitly examine power as an influencing factor? 2) In the studies that examine power, how does the study team conceptualize, define, operationalize, and understand power? 3) What opportunities exist to examine and intervene upon power through IS research? Here, we focus on scientific products that explicitly conceptualize and operationalize power, given the need to focus the field’s attention on opportunities to consider and reshape the distribution and enactment of power in implementation systems and processes.

Methods

Study design

We followed guidance from Arksey & O’Malley [23] and Levac and colleagues [24] to develop the protocol for this review. Protocol details are provided elsewhere (manuscript under review), and the study was prospectively registered via the Open Science Framework (http://osf.io/uy3ez). We followed a six-step process: 1) articulation of the research question and purpose; 2) identification of relevant studies; 3) selection of relevant studies; 4) data extraction; 5) data summarization; and 6) reporting results. We reported results using the PRISMA scoping review (PRISMA-ScR) protocol [25].

Search strategy

To guide the search strategy for this review, we define implementation science as the scientific study of methods to promote the systematic uptake of research findings and other EBIs into routine practice and, hence, to improve the quality and impact of health-focused interventions [26].

Following multiple rounds of piloting and discussion with the review team, a health sciences librarian drafted and conducted a search of PubMed, Embase, APA PsychINFO, MEDLINE, and CINAHL for relevant peer-reviewed literature. This was a complete search strategy, including keywords and relevant subject headings search terms, in combination with the Boolean term “OR” (see Table 1).

Table 1.

Pilot search in MEDLINE

Date	Keywords	Search results
February 23, 2022	Evidence-Based Practice/or Implementation Science/or (evidence based or evidence informed or knowledge translation or implementation science or practice-based evidence).ti,ab. Power, Psychological/or Empowerment/ empower*.ti,ab. (power adj2 (dynamic* or differen* or imbalance* or structur* or relationship* or sharing or share*)).ti,ab. 2 or 3 or 4 hierarchy.ti,ab. 5 or 6 1 and 7	2,000 studies

All articles yielded through this search were uploaded to Covidence systematic review management software [27] for study and process management. Duplicate entries were removed automatically as a first step and then manually by team members as duplicates were identified.

Inclusion and exclusion criteria

We included any human subject studies conducted in clinical, community, or public health settings, including, but not limited to, community organizations, coalitions, nonprofits, schools, clinics, hospitals, and care facilities. We only included studies with a substantive focus on implementation science, as defined by the NIH as “the use of strategies to adopt and integrate evidence-based health interventions and change practice patterns within specific settings” (e.g., studies that describe planning, implementing, sustaining, or scaling up an EBI). [26] A study was included if it described a specific EBI and cited published effectiveness data for that intervention. We did not seek to examine the quality of the underlying evidence for each EBI or the original audience for the EBI as we deemed it out of scope for this review. Additionally, studies had to explicitly examine power or attempt to intervene upon power through their IS efforts in public health, clinical, or community settings. Both observational and experimental studies in the peer-reviewed literature were included, and study inclusion was not limited by year of publication or country of origin. Only studies written in English were included in the full-text analysis.

Studies that did not substantially focus on health or IS, were purely conceptual, or did not address power explicitly, were excluded. Studies outside of clinical, community, and public health settings were also excluded. Finally, studies focusing solely on dissemination, defined by the NIH as “the targeted distribution of information and intervention materials to a specific public health or clinical practice audience with the intent to spread knowledge and the associated evidence-based interventions [26] were also excluded as they were beyond the scope of this project.

Screening process

Using Covidence, a team of trained reviewers familiar with both conducting systematic reviews and implementation research independently screened the titles and abstracts of all identified studies. Reviewers were trained as a group and met regularly to discuss any questions, areas of ambiguity, or suggestions that emerged throughout the review process. Two reviewers screened each study title and abstract using the screening guidance provided in the list of inclusion and exclusion criteria outlined above, which was visible to all reviewers through the Covidence interface. Any discrepancies in reviewer screening decisions were resolved by a third reviewer and/or team discussion until consensus was reached. Studies without a readily available abstract were included for further screening during full-text review.

Studies not excluded following the thorough title and abstract review were screened further via full-text review using the same process outlined above. That is, two reviewers reviewed the full text of each study, applying the inclusion and exclusion criteria described above, before determining whether it should be included in data extraction. When discrepancies arose in reviewer decisions, a third reviewer determined study inclusion by applying the inclusion and exclusion criteria and/or consulting with the reviewer team to reach consensus.

Data extraction

Using the data extraction template offered in the Covidence interface, the team developed a framework to guide data charting from included articles. The data extraction framework (summarized in Table 2) included general information about the study, key study details, and whether/how power was examined. For consideration of power, we focused on the definitions of power offered, focal actors, how power was discussed in relation to the EBI or implementation strategies, suggestions for intervening on power, measures of power utilized, key findings related to power, and levels of power addressed (per Fung’s framework). Initially, Fung’s framework supported the conceptualization of power and broad study design. During the process of reviewing included studies, the team found that the liberatory orientation of Fung’s framework aligned with the study’s philosophical and conceptual orientations, and had actionable implications for approaching change in the field of IS. In these ways, Fung’s framework was used both as a conceptual guide and an analytic tool.

Table 2.

Data extraction framework

Content	Extraction framework component
General information	Title Author name Publication year Country or countries in which data were collected Implementation setting: School (includes daycare/preschool through university); Clinic or hospital (e.g., healthcare); Public health departments; Community-based organization (nonprofits that are mission-driven & serve a catchment area, including faith-based organizations); Other (specified) Communities whose health needs are addressed in the study
Characteristics of included studies	Key outcomes of interest Phase of implementation Named implementation science frameworks, theories, or models Study methods Study design type Time frame of data collection Prospective, retrospective, or cross-sectional design Inclusion of community members or other stakeholders in the study research team
Consideration of power	Definition of power presented Actors discussed in the definition of power Discussion of power in relation to implementation or intervention strategies Suggestions for intervening upon power Measures of power included in the study Key findings or patterns related to power identified in the study Level(s) of power addressed, per Fung’s framework

A team of reviewers assessed each included study to identify and extract the data points previously outlined. Two independent reviewers extracted data from each study, and a third reviewer addressed any discrepancies in extracted data using the consensus-based approach described previously.

Data analysis and synthesis

Extracted data were analyzed using both descriptive statistics (for count data, e.g., number of studies explicitly defining power) and a framework analysis approach [28] to characterize the range and detail of content pertinent to the research question.

Positionality

The review team included individuals with a broad range of multidisciplinary research expertise, including equity-focused implementation science from an academic perspective, community-based participatory research, and the understanding of systems change and transformation required by those approaches. Additionally, the team had expertise in health services research, behavioral science, social welfare, medicine, and public health and many individuals have expertise partnering with practice- and community-based implementers. Team members brought rich perspectives based on varied social identities, including those of first-generation scholars, immigrants and children of immigrants, Asian-American and Chicanx heritage, rural and urban residency, diverse gender identities, and socioeconomic diversity. In addition to bringing critical and transformative perspectives, the group structured decision-making and reflexive processes in a way that was attuned to the varied expertise (rather than position or role) individuals brought to the work.

Results

Search results

Our final search yielded 4,474 articles; 3,531 remained after duplicate articles were removed. Of these 3,531 articles, 3,177 articles were excluded during the abstract screening phase, resulting in 354 articles eligible for full-text review. Following full text review, 326 of these articles were excluded because: 1) the article was not in English (n = 3), 2) the article could not be found (n = 1), 3) the article was purely conceptual and did not include data (n = 95), 4) the article did not include a focus on implementation science (n = 65), 5) the article was not focused on a specific EBI (n = 2), 6) use of the word empowerment as a discussion of power in the article was not meaningful or substantive (n = 53), 7) the article did not have a specific health focus (n = 2), 8) the article did not sufficiently examine power or power dynamics (n = 76), and 9) the article’s use of the term “power” did not align with the definition outlined for this review (n = 29). We note that some articles met multiple exclusion criteria. This process is outlined in the PRISMA flow diagram (Fig. 2).

Fig. 2.

PRISMA flow diagram of article selection

Description of articles

Our final sample for review included 28 articles published between 2006 and 2022 (Table 3). Data reported across the eligible studies were from 22 different countries. Most studies used qualitative methods (86%), while the remaining studies employed mixed-methods approaches. Study settings varied, with most conducted in clinical or hospital settings (79%). Smaller proportions were carried out in community-based organizations (21%), schools (7%), and public health departments (7%). About one-quarter took place in diverse community contexts such as correctional facilities, microfinance cooperatives, or prisons (25%). As some studies were implemented across multiple settings, percentages exceed 100%. While we have provided high-level descriptions of the EBIs of interest for each study, the team did not evaluate the underlying evidence as we deemed it out of scope for this inquiry.

Table 3.

Study characteristics

Year	First Author and citation	Title	Evidence-based intervention	Site of data collection	Implementation setting	Methods	Theory, model, or framework named	Definition of power provided
2018	Altaweli [29]	Ethnographic study of the use of interventions during the second stage of labor in Jeddah, Saudi Arabia	Evidence-based procedures during the second stage of labor	Saudi Arabia	Clinic or hospital	Qualitative	No	No
2014	Baeza [30]	Evidence-based practice: Reflections from five European case studies	Guideline-based stroke care	England, Sweden, Poland	Clinic or hospital	Qualitative	No	Yes
2017	Bahr [31]	Interprofessional Health Team Communication About Hospital Discharge: An Implementation Science Evaluation Study	Structuredcommunication about readiness for discharge from hospital	United States	School; Clinic or hospital; Public health departments	Qualitative	Yes	No
2017	Banks [32]	The researchers’ role in knowledge translation: a realist evaluation of the development and implementation of diagnostic pathways for cancer in two United Kingdom localities	Cancer diagnosis and referral pathway	England	Clinic or hospital	Qualitative	No	No
2020	Bjonness [33]	I’m not a diagnosis: Adolescents’ perspectives on user participation and shared decision-making in mental healthcare	Evidence-based shared decision making in adolescent mental healthcare	Norway	Clinic or hospital	Qualitative	No	Yes
2019	Byrnes [34]	EXploring practice gaps to improve PERIoperativE Nutrition CarE (EXPERIENCE Study): a qualitative analysis of barriers to implementation of evidence-based practice guidelines	Enhanced Recovery After Surgery (ERAS) guidelines	Australia	Clinic or hospital	Qualitative	Yes	No
2017	Conway [35]	The barriers and facilitators that indigenous health workers experience in their workplace and communities in providing self-management support: a multiple case study	Evidence-based chronic disease self- management interventions	Australia	Other: Community	Qualitative	No	Yes
2018	Edge [36]	An Assets-Based Approach to Co-Producing a Culturally Adapted Family Intervention (CaFI) with African Caribbeans Diagnosed with Schizophrenia and Their Families	Family Intervention psychosocial evidence-based intervention	England	Clinic or hospital; Community-based organization	Qualitative	No	No
2021	Eekholm [37]	‘Stolen Time’-Delivering Nursing at the Bottom of a Hierarchy: An Ethnographic Study of Barriers and Facilitators for Evidence-Based Nursing for Patients with Community-Acquired Pneumonia	Evidence-based guidelines for nursing care for older patients admitted with community-acquired pneumonia	Denmark	Clinic or hospital	Qualitative	Yes	No
2021	Ekawati [38]	Challenging the status quo: results of an acceptability and feasibility study of hypertensive disorders of pregnancy (HDP) management pathways in Indonesian primary care	Guideline-based hypertensive disorders of pregnancy management pathways	Indonesia	Clinic or hospital	Mixed-methods	Yes	No
2013	Fasula [39]	Project power: Adapting an evidence-based HIV/STI prevention intervention for incarcerated women	HIV/STI prevention intervention, Project Safe	United States	Other: Carceral facilities	Qualitative	Yes	No
2012	Gillard [40]	Informing the development of services supporting self-care for severe, long term mental health conditions: a mixed method study of community based mental health initiatives in England	Policy guidance on supporting self-care for people with long term, severe mental health problems	England	Clinic or hospital; Community-based Organization	Mixed-methods	No	Yes
2015	Hamdani [41]	Model for Service Delivery for Developmental Disorders in Low-Income Countries	World Health Organization’s evidence-based mental health gap intervention guide	Pakistan	Other: Community	Quantitative	Yes	No
2019	Hartmann [42]	Generating CHARISMA: Development of an Intervention to Help Women Build Agency and Safety in Their Relationships While Using PrEP for HIV Prevention	HIV pre-exposure prophylaxis (PrEP)	South Africa	Clinic or hospital	Mixed-methods	No	Yes
2007	Harwood [43]	Nurses’ perceptions of the impact of a renal nursing professional practice model on nursing outcomes, characteristics of practice environments and empowerment–Part I	Center-level guidelines for nephrology nursing professional practice models	Canada	Clinic or hospital	Quantitative	No	Yes
2015	Herepath [44]	A realist analysis of hospital patient safety in Wales: applied learning for alternative contexts from a multisite case study	Three focal interventions from the 1000 Lives+ patient safety programme: Improving Leadership for Quality Improvement, Reducing Surgical Complications and Reducing Health-care Associated Infection	Wales	Other: Community	Qualitative	Yes	Yes
2006	Jones [45]	The Building Wellness project: a case history of partnership, power sharing, and compromise	Evidence-based screening, treatment, and support for depression	United States	School; Clinic or hospital; Community-based organization	Qualitative	No	No
2015	Kueny [46]	Facilitating the implementation of evidence- based practice through contextual support and nursing leadership	Evidence-based interventions for fall prevention	United States	Clinic or hospital	Qualitative	No	No
2020	Matheson [47]	Strengthening prevention in communities through systems change: lessons from the evaluation of Healthy Families NZ	Evidence-based interventions addressing nutrition, physical activity, tobacco use, and alcohol intake.	New Zealand	Clinic or hospital	Qualitative	No	No
2013	McCalman [48]	The transfer and implementation of an Aboriginal Australian wellbeing program: a grounded theory study	Family Wellbeing Program focused on pathways to employment and mental health management	Australia	Clinic or hospital; Community-based organization	Qualitative	Yes	Yes
2020	Mugo [49]	From research to international scale-up: stakeholder engagement essential in successful design, evaluation and implementation of paediatric HIV testing intervention	The Counseling and Testing for Children at Home intervention for pediatric index case identification for HIV	Kenya	Clinic or hospital	Qualitative	Yes	Yes
2021	O’Hara [50]	Multi-level implementation factors that influence scale-up of methadone maintenance treatment in Moldovan prisons: A qualitative study	Methadone maintenance treatment	Moldova	Other: Prison	Mixed-methods	No	No
2022	Piat [51]	Translating mental health recovery guidelines into recovery-oriented innovations: A strategy combining implementation teams and a facilitated planning process	National guidelines for recovery from mental illness	Canada	Community-based organization; Other: publicly-funded organizations	Qualitative	No	No
2016	Steinmo [52]	Bridging the gap between pragmatic intervention design and theory: using behavioral science tools to modify an existing quality improvement program to implement “Sepsis Six”	Evidence-based “Sepsis Six” guidelines to reduce mortality from sepsis	England	Clinic or hospital	Qualitative	Yes	No
2009	Swendeman [53]	Empowering sex workers in India to reduce vulnerability to HIV and sexually transmitted diseases	Evidence-based intervention for HIV and sexually transmitted disease prevention	India	Clinic or hospital Community-based organization; Public health departments; Other: microfinance cooperatives and political organizations	Quantitative	No	Yes
2020	Tadaumi [54]	A qualitative study of factors that influence midwives’ practice in relation to low-risk women’s oral intake in labour in Australia	Evidence-based practice regarding oral food and fluid intake during labor	Australia	Clinic or hospital	Qualitative	No	No
2013	Turner [55]	Barriers to and enablers of evidence-based practice in perinatal care in the SEA-ORCHID project	Evidence-based clinical practices to support perinatal healthcare	Thailand, Malaysia, Philippines, Indonesia	Clinic or hospital	Qualitative	Yes	No
2020	Worum [56]	Physiotherapists’ perceptions of challenges facing evidence-based practice and the importance of environmental empowerment in fall prevention in the municipality - a qualitative study	Evidence-based fall prevention programs	Norway	Clinic or hospital	Qualitative	No	Yes

Defining power

Of the 28 articles in the sample, 11 (39%) presented a definition of power. Of the 11 articles that defined power, 6 referenced a formal definition [30, 33, 42, 43, 49, 56]. For the 17 articles that did not present a definition of power, they were determined to have met the criteria for sufficient attention to power or power dynamics through the described efforts to evaluate and/or intervene upon power. For each included article, Table 4 describes actions taken within the implementation effort to assess and/or address power.

Table 4.

Operationalizing power in implementation efforts

	LEVEL 1: EVERYDAY POWER Supports and challenges faced in routine interactions	LEVEL 2: POLICY POWER Governmental and institutional laws and policies that constrain or support individuals	LEVEL 3: STRUCTURAL POWER Systems that impact the link between individual-level and policy-level action	LEVEL 4: ETHICAL POWER The types and spread of norms, ideology, and values in the focal society
Jones 2006 [45]	Engaged community members in developing action plans Designed an approach to power-sharing that focused on leveraging diverse expertise and using collaborative decision-making in designing intervention strategies		Attempted to shift power from healthcare and community service systems to community members and frontline providers by ensuring that community voices were central in developing culturally appropriate resources and programs Altered research processes to support new forms of executing partnerships between academics and practitioners, e.g., through new models of leadership
Swendeman 2009 [53]	Increased agency in sexual practices (e.g., negotiating condom use) and workplace negotiations (e.g., changing work contracts) among sex workers Focused on building individual consciousness around issues facing sex workers Supported self-protection and community mobilization/social support among sex workers Reduced economic vulnerability and increased political participation among sex workers		Engaged in advocacy with stakeholders and powerbrokers, such as local politicians, police, and brothel owners, reframed sex work as valid work, and championed workers’ rights to reduce stigma and marginalization Challenge led by community partners to address structural power dynamics that traditionally disempowered sex workers Created microfinance systems to help women improve their negotiating power in household decision-making	Set an explicit goal (even within this work) of normalizing sex work
McCalman 2013 [48]		Attempted to influence policy by engaging policymakers so that they would craft policy in a different way	Attempted to center Aboriginal people’s participation, responsibility for and control of their own affairs, and the associated ripple effects to family members, organizations, and communities
Baeza 2014 [30]	Examined negative impacts of hierarchical health systems on implementers perceived to be at the lower end of the clinical hierarchy		Emphasized ways hierarchical systems concentrate power for “high-status” practitioners through the use of EBI (vs. relying more on tacit knowledge)
Hamdani 2015 [41]	Identified an opportunity to empower family volunteers to deliver evidence-based care to their children and other local children and train other volunteers. Highlighted how the project worked to reduce stigma related to developmental disorders so that these children would not be excluded from community life.
Kueny 2015 [46]	Examined the role of nurse managers in daily operations, particularly in empowering nurses to take initiative in EBI changes. Identified a crucial role played by nurse managers in facilitating communication between staff nurses and hospital administration, influencing the daily work environment, and ensuring that EBIs were implemented effectively in their units	Examined the impact of external regulatory forces, such as the Joint Commission and national patient safety goals, on nurse managers’ practice change efforts. These policies significantly impact how the EBI is adopted at the unit level Noted nurses’ involvement in shared governance committees as positive among nurse managers, but their level of power for decision-making was sometimes unclear and limited	Focused on the impact of culture, structure, and resources on nurses’ ability to improve care in their units. Highlighted support from nurse manager-comprised management teams that provided mentorship and support for change. Noted nursing committee membership as a way to empower nurses and create practice change. Identified drivers of change, including institutional avenues to receive direct nurse input, empowerment to review and implement EBIs, and open communication across administrative levels.
Conway 2017 [35]	Highlighted patients’ social needs (e.g., housing and employment), which often overshadowed health concerns. Identified client empowerment as a facilitator in the success of self-management. Empowerment, in the participants’ terms, meant being able to tell their own story, identify their own problems and goals, and take control of their conditions through education and support		Attended to the systemic barriers that Indigenous health workers face, particularly the institutional racism within mainstream healthcare services, which discourages Indigenous patients from accessing care. Addressed the lack of support structures within healthcare systems that impact Indigenous health workers’ ability to provide self-management support effectively
Edge 2018 [36]	Highlighted power dynamics adolescents face in routine interactions with healthcare professionals Explored adolescents’ sense of powerlessness stemming from being diagnosed without input and feeling pushed into treatment Acknowledged that researchers may be regarded as agents of power and surveillance and sought to build trust by reducing the distance between community members, health care providers, and academics. Focused on flattening hierarchies between patients and clients		Highlighted the importance of naming and creating learning opportunities to mitigate racial hierarchies, e.g., potential power differences between White therapists and Black clients. This also connected to hierarchy in the healthcare system, e.g., the authority of healthcare professionals over patients, impacting the treatment process
Hartmann 2019 [42]	Discussed the power dynamics in intimate relationships, particularly how male partners influence women’s ability to use HIV prevention methods consistently. The dynamics ranged from opposition to support, with some male partners feeling threatened by women’s autonomous use of HIV prevention products. Extended a model based on the idea that intimate partner violence is intended to disempower female partners, and counseling can help increase a woman’s sense of power over her life and help her develop partner communication and problem-solving skills to reduce violence and increase safer sex in her relationships.		Discussed institutional power within healthcare settings and research trials. Highlighted how healthcare structures must engage male partners and challenge harmful gender norms to increase PrEP adherence. Attempted to shift power dynamics by using structures, including research clinics and community outreach activities, and educating men and supporting women’s empowerment within relationships. Engaged individuals demonstrating leadership in the workshops to become ‘community action team members’ and use their knowledge to educate their communities further. Created safe spaces for men to consider and challenge traditional masculinities and gender relations. Paired participatory workshops that generally have short-term impacts with community mobilization and public awareness activities that worked toward more medium and long-term impacts on gender equality in communities.
Worum 2020 [56]		Described the ways policies and political decisions shape the ability of physiotherapists to implement EBIs in fall prevention. Includes attention to the way a lack of understanding long-term benefit, combined with little immediate economic benefit led to low priority placed on fall prevention. Also described public policies, politicians, and organizational culture as influencing factors for EBI implementation.	Detailed how leadership and organizational culture influence physiotherapists’ ability to use EBIs. Leadership plays a critical role in empowering physiotherapists by providing the necessary resources, support, and a culture that values continuous professional development and EBI implementation. Leaders who understand the importance of EBIs can increase organizational readiness and foster an empowering environment
Bjonness 2020 [33]	Described how power dynamics influenced shared decision-making (or lack thereof) related to treatment and provider-patient concordance. Also discussed ways that diagnoses reduced patients to a set of problems. The adolescents in the study discussed the power dynamics in their day-to-day interactions with healthcare professionals. An example was being diagnosed or labeled without input, and the feeling of being pushed into treatment they did not initiate, which created a sense of powerlessness.		Highlighted the hierarchical nature of the healthcare system, particularly in how the authority of healthcare professionals over adolescents affects the treatment process. Related, organizational culture was noted to influence teamwork and clinicians’ attitudes and leadership.
Matheson 2020 [47]		Described the government-funded Healthy Families NZ initiative, which is part of a national policy approach to strengthening the health prevention system. The centering of industry needs and exclusion of community voice in local alcohol licensing processes was highlighted as an example of how a lack of community agency impacted implementation of prevention efforts	Described the institutional structures involved in implementing the initiative, including groups comprised of leaders from various sectors, including health, local government, business, and education, who work to influence transformational change at the community level. Disrupting the status quo and addressing power imbalances was part of the work. One of the outcomes it achieved was to improve information sharing between communities, policymakers, and other decision-makers.
Mugo 2020 [49]		Highlighted how the Kenya Ministry of Health played a critical role in incorporating the study approach into national HIV testing policies. By integrating the findings into the national HIV testing guidelines and endorsing the index case testing model, the ministry used its hierarchical power to influence national scale-up of the intervention. This demonstrates the role of national-level policies in driving the adoption of healthcare innovations
O’Hara 2021 [50]	Examined attitudes of individual prisoners and of individual prison administrators toward methadone maintenance treatment, and impact on treatment access and acceptance	Highlighted the challenges in scaling up methadone maintenance treatment (MMT) despite administrative coordination and support from three high-level ministries. Identified scale-up challenges related to low coverage and resistance, driven in part by prison subcultures and social hierarchies Examined the role of “rigid treatment policies” on perceptions and trust and acceptance of evidence-based solutions.	Examined prison hierarchies and the influence of informal prisoner governance systems on MMT uptake. These informal hierarchies often discouraged or outright blocked prisoners from enrolling in MMT due to cultural norms and economic interests tied to drug trafficking within the prison system. This power structure often undermined the formal governance and policies intended to provide health interventions like MMT
Ekawati 2021 [38]	Highlighted the power imbalance enacted in daily interactions between healthcare providers, i.e., general practitioners, nurses, midwives, and patients. For example, patients often complied with the instructions of doctors without fully understanding or questioning them. Examined acceptability and feasibility of implementing specified care pathways, in relation to primary care providers’ experiences and whether they had a positive experience and a sense of practice improvement including while navigating hierarchical interprofessional relationships	Addressed the influence of regulations from the local health office. Emphasized the need for more supporting policy to reinforce primary care practice	Identified differential clinical resources as a key impediment to implementation, though not as large of a barrier as interprofessional hierarchy.
Piat 2022 [51]	Assessed whether implementation team members valued stakeholder inclusion and felt “free to express their opinions, in an inclusive environment where people shared a strong interest in recovery.” Documented “challenging group dynamics” when conflict emerged between tenants and staff and how this was perceived by implementation team members. For service users/tenants, this was one of the first times they had ever been engaged in planning and implementation.

Of the 28 articles in the sample, 18 (65%) described the ways they operationalized power in relation to implementation strategies or the EBI. We mapped these operationalizations onto Fung’s conceptualization of power at four levels as shown in Table 4. Of the 18 articles, 12 addressed Level 1 (Everyday Power), 7 addressed Level 2 (Policy Power), 12 addressed Level 3 (Structural Power), and 1 addressed Level 4 (Ethical Power). The table is intended to illustrate the focus of action at different levels, but in reality, the interventions may have been blended in a way that made the different target levels invisible to the study team. Many of the articles focused on opportunities to both redistribute and generate power. A manuscript by Swendeman and colleagues [53] offers a compelling illustration of alignment across levels of power. The project focused on empowering sex workers in India through a multi-prong intervention, with the goal of ultimately reducing vulnerability to HIV and sexually transmitted diseases. In terms of everyday power, the intervention targeted sex workers’ agency in negotiation of sexual practices and workplace conditions, consciousness-raising around key issues, community mobilization and social support among sex workers, economic vulnerability, and political participation. At the structural level, the intervention attempted to reframe perceptions around sex workers and workers’ rights, address underlying power dynamics, and use microfinance to improve sex workers’ negotiating power within their households. Finally, at the ethical level, the intervention sought to move towards normalizing sex work. Qualitative research on this structural intervention examines interconnectedness between levels of power and highlights the specific pathways through which collective political mobilization, the development of destigmatized sex worker identity and structural change (outcomes of attention to third and fourth levels power) practically influence everyday empowerment in negotiating work (first level) [57].

As seen in Fig. 3, a few common themes arose from the studies within a given level of Fung’s framework. Importantly, these targets reflect the goals identified from each study and do not necessarily correlate with their impact on outcomes at a given level. Additionally, some of the studies highlight tensions across levels, e.g., when Level 2 power was addressed through policy change, but Level 1 power dynamics precluded successful implementation in the example of methadone treatment in a Moldovan prison.

Fig. 3.

Common themes related to power in implementation science projects from 28 included articles, across levels described by Fung

Theories, models, and frameworks

Of the 28 articles in the sample, 11 (39%) explicitly named the use of IS TMFs or core approaches that guided their research (see Table 3). Among these, 12 distinct TMFs were referenced; some of which included explicit attention to power (see Table 5).

Table 5.

IS theories, models, and frameworks highlighted by the 11 articles that provided these details

Author/Year	Framework or approach
Turner 2013	Making psychological theory useful for implementing evidence-based practice: a consensus approach [58]
Fasula 2013	CDC’s MAP: “Map of the Adaptation Process: A Systematic Process for Adapting Evidence-Based Behavioral Interventions” [59]
McCalman 2013	Empowerment Research Program (ERP) [60]
Hamdani 2015	Integrated Innovation [61]
Herepath 2015	Plan, Do, Study, Act (P-D-S-A) model for improvement [62, 63]
Steinmo 2016	Plan, Do, Study, Act (P-D-S-A) cycle approach [64]
	Theoretical Domains Framework (TDF) [65]
Bahr 2017	Consolidated Framework for Implementation Research (CFIR) [66]
Byrnes 2019	Integrated Promoting Action on Research Implementation in Health Services i-PARIHS [67]
Mugo 2020	Power and interest model [68–70]
Eekholm 2021	Theoretical Domains Framework (TDF) [65]
Ekawati 2021	Medical Research Council (MRC) [71]
	Practical Robust Implementation and Sustainability Model (PRISM) [72]

Discussion

This review examined the extent to which and how EBI-focused implementation studies explicitly attend to power as an influence on implementation and conceptualized, operationalized, and sought to examine or address power. We found limited explicit attention to power overall, with a clustering of attention around everyday and structural power, with much less attention to policy and ethical power. Additionally, we found limited reference to IS TMFs in studies involving power. The result suggests a literature that increasingly acknowledges the importance of power in terms of health equity [73], but has yet to translate that interest into treating power as a core construct and lever for change.

In our review of over 3,500 manuscripts, we found that only 28 papers explicitly discussed power in relation to EBI implementation, and of these, only 11 presented a formal definition of power. While this does not imply that other studies entirely ignored or failed to address power, it underscores a critical need for researchers to be explicit about the theoretical constructs guiding their work. After all, theory shapes what we see, the questions we ask, and the interpretations we make [74]. The failure to explicitly define or operationalize power perpetuates the risk that vital structural drivers of inequity remain invisible or are treated as external, immutable factors beyond the purview of IS efforts. Similarly, only 11 studies involving power named IS TMFs, suggesting that power remains a peripheral construct in the field, rather than being integrated into IS efforts. While recent TMFs, such as the Consolidated Framework for Implementation Research (CFIR) 2.0, have made efforts to foreground equity and power, researchers must translate these areas of emphasis into targets of disruption or redistribution, rather than simply describing them as contextual elements [75]. Similarly, the Theoretical Domains Framework was adapted to include intersectionality and supports assessments of and interventions upon power in systems under study [76]. We recognize that not all researchers will have the ability to explicitly name attempts to address power.

Our review also highlights that among articles that explicitly defined power, there was a greater focus on everyday or structural power, less attention on policy-focused power, and almost no attention to ethical power. This is consistent with a broader literature on power as connected to social change and the introduction of innovations, which tends to emphasize the stability of systems over change [77]. Per Fung, the most impactful change comes from alignment across all four levels of power [20]. Yet, it is likely that routine interactions and measurable outcomes lend themselves to research efforts more easily than complex, large-scale policymaking efforts or attempts to change ideologies or norms. Additionally, IS projects tend to have short timeframes, making intervention on and evaluation of higher-level change at societal or policy levels a challenge. Similarly, the study of power at a societal level is diffuse and often implicit, or so pervasive that it may often be considered background context rather than a focus of intervention. Notably, the only study that attempted to address ethical power directly was embedded in a larger, long-term social movement, which provided structure for those being studied (i.e., sex workers) to make demands of researchers, including to attend to the role of ethical power in shaping implementation [53]. Thus, it may be crucial to embed power-focused studies within broader, community-based movements, allowing the necessary investment and timeframe to bring about societal change. Such an approach also shifts power away from academics by ensuring meaningful investment in community and practice infrastructure, rather than a short-term academic project [78]. By treating power as fundamental to implementation, IS efforts can challenge and address inequities embedded in health systems. This reframe offers the opportunity to redirect investments from proximal power-related frictions towards the underlying normative and political conditions that sustain those frictions.

Incorporating the study of and intervention upon power in IS efforts requires examining implementation systems and processes while seeking to change them, avoiding reductionist perspectives, incorporating critical perspectives, attending to how power dynamics are reflected in evidence generation and evidence-based decision-making, and moving past the dominant conceptualization of EBIs as technologies incorporated into a broader system [9, 79, 80]. This connects with opportunities to address epistemic injustices, where marginalized groups are perceived as less credible or less capable of generating knowledge, which often leads to imbalances in what knowledge is privileged and how findings are shared [81]. These perspectives enable researchers to raise important questions about how implementation is framed and whose interests it serves or fails to serve [82]. Relatedly, another opportunity arises from participatory approaches to research, as teams can examine the impact of power on issue selection, team composition, study design, outcomes of interest, reflexivity, and the positionality of team members, among other factors [1]. Participatory research emphasizes long-term engagement in communities, a deep understanding of the study context, and investment in relationship building-three strategies that also require a deep examination of and redistribution of power [83]. Another opportunity lies in the use of co-creation/co-design to examine and intervene upon power imbalances, generate new power, and utilize it in agreed-upon ways (e.g., for valuable resources, see [84–86]). Participatory and co-design approaches prompt meaningful reflexivity and attention to positionality on behalf of the research team, offering an important opportunity to examine how patterns of power and privilege play out in IS efforts [87, 88]. Tools such as dynamic journey mapping [89], power mapping [90], and the Power Primer (developed by the National Association of County and City Health Officials) [91] can support IS teams to analyze the ways in which power support or restrain various actors’ ability to achieve their goals.

Implementation scientists incorporating TMFsin their planning processes can explicitly integrate power into their conceptualizations. A recent illustration demonstrated opportunities to increase the utility of IS TMFs and potential impact on equitable outcomes by connecting four forms of power (structural, discursive, instrumental, and epistemic) to five domains of the CFIR [92]. In this way, power can be attended to explicitly, throughout the lifecourse of a project, and in dissemination products. We offer exemplar questions such an integration might offer as a starting point for inquiry. For example, researchers can attend to determinants, including how power is distributed or utilized in a system, both in the past and as a function of implementation. Related to the EBI, an examination of power may direct attention to how the selected EBI (or pool of available EBIs) reflects the differential value placed on knowledge from privileged groups or via privileged methods, priorities held by power groups, and more [93, 94]. For implementation strategies, IS teams can develop and evaluate them in terms of their impact on generating power with and power to, as well as how they may create, sustain, or exacerbate power differentials. Finally, outcomes could be examined in terms of whose interests and definitions of success are privileged. In this study, applying Fung’s framework as a conceptual guide and analytic tool for results extraction offered a valuable framework for identifying opportunities to 1) move away from compliance-focused strategies to inclusive, agency-oriented strategies, 2) examine policy in the context of power across levels by assessing whose interests are served, how resources are distributed, and what the response is from various constituents, 3) ensure systems are inclusive of a wide range of evidence types, 4) structure IS projects to ensure that implementers and decision-makers have sufficient control and ability to shape the project for long-term utility, and 5) embed power-oriented projects into broader ongoing work or long-term studies so deep change in societal norms and expectations can be addressed at multiple levels.

Limitations

Findings must be interpreted in the context of a few limitations. First, the review does not include practice-based expertise (in terms of content or study team membership) from fields such as community organizing, grassroots mobilization, or movement building. As a result, it is likely to overlook critical insights from community practitioners and advocacy organizations whose work meaningfully addresses power but falls outside the realm of academic publishing. For example, efforts like those led by The Praxis Project [95] demonstrate how power can be built and shifted through community-driven strategies—contributions that are often captured in grey literature, reports, or other non-peer-reviewed formats. Incorporating these sources is a crucial next step in capturing a more comprehensive picture of how power is addressed in real-world implementation contexts. Second, this review focused specifically on publications that explicitly discussed power (and related concepts) in relation to implementation of EBIs. While this excludes work that may address power more implicitly or simply did not report those activities, our goal was to identify scholarship that supports conceptual and methodological development in the field that explicitly names, defines, and operationalizes power in ways that can advance IS. Future work should use a broader set of relevant concepts, e.g., related to epistemology or agenda setting. Third, the restriction to English manuscripts and predominant representation of studies from high-income countries suggest that further research is required to examine patterns in low- and middle-income countries, particularly since IS efforts are enacted in meaningfully different ways across these categories [96]. Fourth, the review did not examine the quality of underlying EBIs or impacts on implementation or health outcomes. Instead, we studied more proximal targets of power-focused action. Finally, the review is current as of February 2022 and there has been a recent surge in interest in power in IS; thus an update to the review will be useful soon.

Conclusion

We argue that purposeful, theory-driven attention to power is necessary if we seek to leverage IS as a tool for health equity and social justice. In the same way that assessment of cost is becoming routine in IS [97], examining and intervening upon power can be routinized to support the equitable integration of EBIs into routine practice [7, 98]. Scholarship, such as the Health Equity Implementation Framework [98] and the He Pikinga Waiora Implementation Framework [99], signal welcome shifts toward treating power dynamics—structural, interpersonal, and individual—as crucial forces shaping implementation processes and outcomes. In these ways, power can be attended to not as a feature of implementation context, but as an active driver of change. Our findings underscore the need for implementation scientists to more intentionally and explicitly incorporate examinations of power, with clear theoretical grounding and a focus on intervention, to advance equity and justice within health systems and implementation processes. For those already engaged in this work, this study can serve as a prompt to be more explicit about it to build this area of inquiry. This means incorporating additional work into IS team members’ job descriptions as examining and addressing power is critical if we are to shift IS efforts away from patching flawed systems toward becoming a force to change them for the better.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

SR and NS conceived of the study objectives; SR, JLC, NS, SA, ML, EDJ, PA, and RS, designed the study and analyzed the data. SR, JLC, NS, SA, ML, EDJ, PA, RS, and MS wrote the manuscript. All authors approved the final manuscript.

Funding

This study was supported by the National Cancer Institute (U54CA156732 and 2T32CA057711).

Data availability

The data can be made available upon request to the corresponding author.

Declarations

Ethics approval and consent to participate

N/A.

Consent for publication

N/A.

Competing interests

N/A.