Abstract
Purpose:
Skills in selecting and designing strategies for implementing research-supported interventions (RSIs) within specific local contexts are important for progressing a wider RSI adoption and application in human and social services. This also applies to a particular role in implementation, the implementation support practitioner (ISP). This study examines which strategies have been reported as being used by ISPs across multiple bodies of research on implementation support and how these strategies were applied in concrete practice settings.
Methods:
A systematic integrative review was conducted. Data analysis utilized the Expert Recommendations for Implementing Change compilation of implementation strategies.
Results:
Studies reported on 18 implementation strategies commonly used by different ISPs, who require mastery in selecting, operationalizing, and detailing these. Two further strategies not included in the ERIC compilation could be identified.
Discussion:
Given the use of primarily more feasible implementation support strategies among ISPs, their potential as agents of change may be underutilized.
Keywords
Implementation scientists have shown an interest in how to best support frontline staff, that is, those involved in the delivery of services to children, adults, families, and communities, in their uptake of research-supported interventions (RSIs) since the early beginnings of the field (Harvey et al., 2002; Kitson et al., 1998; Schoenwald et al., 2004). RSIs—be they programs, strategies, procedures, or policies—are those that have been “evaluated using acceptable standards of scientific evidence and found to yield generally positive outcomes” (Thyer et al., 2017, p. 86) for their target populations. Obtaining a widespread uptake and sustainable use of these RSIs remains a challenge in real-world human and social services because individuals, agencies, and systems often lack sufficient skill to overcome barriers to their implementation.
This has been a central topic of debate also in social work. In this profession, evidence-based practice (EBP) is increasingly viewed as central in tackling societal challenges such as child maltreatment, homelessness, family violence, or unequal access to health care (Barth et al., 2017; Nurius et al., 2017). With EBP, we mean the process of “integrating individual practice expertise with the best available external evidence from systematic research as well as considering the values and expectations of clients” (Gambrill, 1999, p. 346). RSIs represent the best available external evidence in this context, which—to establish EBP—would need to be considered and used in the light of client preferences and professional expertise.
Knowledge and skill deficiencies have consistently emerged from studies as a key factor that prevents human and social service workers from establishing this practice (Ekeland et al., 2018; Finne, 2020; Goel et al., 2018; Grady et al., 2017; James et al., 2019; Lery et al., 2015; Scurlock-Evans & Upton, 2015; Shapira et al., 2017; van der Zwet et al., 2019; Wike et al., 2019). In recent years, educational institutions have increasingly aimed to accommodate RSIs and EBP by, for example, adjusting academic or fieldwork curricula (Bertram et al., 2014, 2018; Drisko & Grady, 2019; Mennen et al., 2018). However, since the evidence on techniques for effectively teaching EBP remains scarce (Drisko & Grady, 2019; Spensberger et al., 2020), it is unlikely that such efforts alone will succeed in growing a much needed workforce of professionals who can progress a more widespread use of the EBP model in human and social service settings. Furthermore, it is well-documented in the literature that training, in the form of teacher-centered didactic education sessions, can be effective in building theoretical knowledge and shaping attitudes and beliefs, but it is often insufficient to achieve practical application of RSIs in real-world settings (Beidas & Kendall, 2010; Hecht et al., 2016; Prior et al., 2008). This has increased the attention for the importance of how RSIs are implemented in human and social services, best represented by the field of implementation research, that is, “the study of methods to promote the uptake of research findings into routine practice” (Bhattacharyya et al., 2009, p. 491). Insights from this field of inquiry highlight that the transfer of knowledge gained through training to the conditions of human and social service organizations and systems is influenced by a variety of factors—for example, the climate characterizing the setting into which new knowledge is embedded or the support for knowledge transfer available in this setting (Blume et al., 2010; Grossman & Salas, 2011; Jackson et al., 2018). Factors like these have to be considered and addressed in order to enable real-world RSI transfer and application.
This implies that next to understanding the RSI itself, stakeholders directly involved in its implementation also require implementation expertise, that is, knowledge and skills in selecting and designing strategies that can support RSI implementation within specific local contexts. This type of expertise has been highlighted as one of the particularly important elements in progressing a wider adoption and application of RSIs in human and social services (Brownson et al., 2018), especially in the field of social work, where implementation science has taken hold only slowly (Barth et al., 2017; Cabassa, 2016). Only recently has the question about how to build this implementation capacity in service agencies and their staff gained increasing attention, reflected in publications that report on theory development (Leeman et al., 2017), the design of implementation courses (Moore et al., 2018; Mosson et al., 2019; Park et al., 2018; Provvidenza et al., 2020), or reviews of the existing literature on strategies and approaches used in implementation capacity building (Leeman et al., 2015; Stander et al., 2018).
Implementation Support
One of the most frequently used approaches to building implementation capacity is the provision of ongoing implementation support (Katz & Wandersman, 2016; Leeman et al., 2015). In the absence of a commonly shared definition of the concept of implementation support, we define this support broadly, as any activities and processes aimed at assisting leadership and staff of human and social service organizations and systems in implementing, sustaining, and scaling RSIs for population impact. Multiple streams of research exist that focus on implementation support, including studies on, for example, facilitation (Cranley et al., 2017), coaching (Artman-Meeker et al., 2015), consultation (Nadeem, Gleacher, & Beidas, 2013), or technical assistance (Dunst et al., 2019), indicating a high degree of variability in the terms used to describe implementation support activities.
In human and social service practice settings, the establishment of organizations dedicated to providing implementation support, for example, purveyors (McWilliam et al., 2016) and intermediary organizations (Franks & Bory, 2015; Proctor et al., 2019), technical assistance centers (Bumbarger & Campbell, 2011), or centers of excellence (Biegel et al., 2003; Mettrick et al., 2017), has contributed to an increasing professionalization of this support and considerable public investment in these organizations (Proctor et al., 2019). Especially in social work, their role has been described as integrative due to their potential to better interlink academic training efforts with community services and to develop a growing workforce in need of skills to implement EBP (Shapiro, 2018). As a consequence, growing numbers of professionals are providing implementation support. These “implementation support practitioners” (ISPs; Albers, Metz, & Burke, 2020) are the first focus point of this article.
ISPs
While multiple reviews exist aimed at describing the activities of ISPs, these mostly focus on a single particular role, for example, on technical assistance providers (Dunst et al., 2019; Taylor et al., 2014), knowledge brokers (Taylor et al., 2014), or facilitators (Elledge et al., 2018). A similar picture emerges for conceptual studies aimed at identifying and characterizing the key functions and mechanisms of implementation support as provided by, for example, coaches (P. A. Snyder et al., 2015), consultants (Edmunds et al., 2013; McLeod et al., 2018; Nadeem, Gleacher, & Beidas, 2013), or other change agents (Glisson & Schoenwald, 2005).
However, a comparison of how these key functions have been defined (see Table 1) illustrates that they are similar in focus and aim centered on easing implementation problems and helping others in improving the uptake and implementation of RSIs.
Definitions of Key Functions of Different Implementation Supports.
Hence, within the field of implementation science, the scattering of the literature across multiple “schools of thought” may create silos that increase the risk of research waste and a lack of cross-school learning. Enabling such learning may help to strengthen and refine the evidence base for implementation support work. Furthermore, for stakeholders operating within real-world implementation practice, it is important to understand the characteristics of quality implementation support and how it can be provided. For the social work profession, it has been argued that quality implementation is key in achieving social progress reflected in, for example, improved youth development, reduced homelessness, or the elimination of racism (Cabassa, 2016; Gehlert et al., 2017). Guidance on how to support such implementation should therefore be based on a consolidated evidence base that utilizes insights about all relevant types of implementation support, independent of their label as, for example, facilitation, consultation, coaching, or knowledge brokering.
A previously developed theoretical program logic reflects this thinking (Albers, Metz, & Burke, 2020). It presents ISPs as individuals who, through the use of unique combinations of knowledge, skills, and attitudes tailored to the contextual settings in which they work, can help individuals, agencies, and service systems to effectively utilize RSIs, maintain their use, and generate positive implementation, service, and client outcomes. Subsequently, a systematic integrative review was conducted to deepen the preliminary thinking reflected in this logic and detail its key components. One of these components is the skills required by ISPs to provide implementation support represented in their use of implementation strategies. These strategies are the second focus point of this article.
Implementation Strategies
Skill can be defined as the ability to do something well based on one’s knowledge, practice experience, or aptitude. It represents an ability to assess the unique context in which an individual operates, to integrate technical with practice knowledge, and to use this information flexibly and creatively in situational decision making (Devaney et al., 2017).
Implementation strategies, “the specific means or methods for adopting and sustaining” RSIs (Proctor et al., 2013, p. 1), represent a central body of knowledge as the “doing” of implementation. Examples of common implementation strategies are training, audit and feedback or learning collaboratives. While further research on the effectiveness of implementation strategies is required (Powell et al., 2019), initial studies indicate that the choice of implementation strategy may affect outcomes, including, for example, knowledge and skill acquisition among policy makers (Sarkies et al., 2017) as well as frontline practitioners (Dimeff et al., 2015), implementation (Beidas et al., 2017; Martino et al., 2019), and client outcomes (Chadwick et al., 2015).
One of the most referenced taxonomies of implementation strategies in the field of implementation science is that developed by Powell, Waltz et al. (2015), the Expert Recommendations for Implementing Change (ERIC) compilation. It differentiates between 73 distinct implementation strategies and provides definitions for each. As such, the taxonomy represents a menu available to ISPs as a means to introduce, teach, and practice implementation as they collaborate with different stakeholders in human and social services. Hence, it could be used to describe the potential skills that ISPs may require to adequately support others in their implementation practice.
However, scholars have highlighted that the effectiveness of implementation strategies is dependent on their potential to sufficiently address the barriers and/or facilitators that influence an implementation (Grimshaw et al., 2012; Powell et al., 2014), assigning great importance to the thorough design of implementation strategies. Taxonomies such as the ERIC compilation are valuable tools in this regard because they provide both the basic elements of strategy design and a common strategy language that stakeholders can use in jointly assembling and applying local implementation approaches. Their basic elements though—the different, predefined implementation strategies—still represent relatively broad labels for implementation activities. For ISPs working in often complex settings of routine practice and policy, this leaves ample room for further strategy operationalization and tailoring. For example, consultation, included in the ERIC compilation as a form of expert-based implementation support, has previously been discussed as a “black box” that contains further activities and functions (Nadeem, Gleacher, & Beidas, 2013), illustrating the relevance of examining the concrete details of using a strategy. Furthermore, the work of ISPs must take into account the dynamic and highly relational nature of policy and practice implementation involving multiple layers of context and differing norms and values among stakeholders (Carey et al., 2019; Huzzard, 2020; Norris et al., 2017).
Hence, the design of strategies is not a straightforward, simple selection process but one of tailored strategy operationalization and integration. Taxonomies of implementation strategies alone are therefore insufficient to understand the breadth and depth of the skills that ISPs require in providing implementation support. To achieve this understanding, it is also necessary to investigate how strategies have been applied in implementation practice. This was the aim of this integrative review. In order to describe the skills required by ISPs, it examines (a) which strategies have been reported as being used by ISPs across multiple bodies of research on implementation support and (b) how these strategies were applied in concrete practice settings.
Method
This research is based on the conduct of a systematic integrative review, involving a narrative synthesis of findings guided by thematic analysis.
The integrative review method “allows for the combination of diverse methodologies” (Hopia et al., 2016, p. 663), that is, the inclusion of studies that use different and diverse designs. It belongs to a group of emerging knowledge synthesis methods for integrating rich contextual data (Saini & Shlonsky, 2012; Tricco et al., 2016) and addressing complex practice and policy questions (Greenhalgh et al., 2018). With theory building being among its key purposes (Kastner et al., 2016; Tricco et al., 2016), an integrative review was particularly suitable for this project’s goal of refining a preliminary ISP program logic previously developed by the authors (Albers, Metz, & Burke, 2020).
While novel methods suffer at times from a lack of clarity on how to apply them (Tricco et al., 2016), all operational steps have been defined for the integrative review in the form of a framework developed by Whittemore and Knafl (2005). This framework was applied to this review, which was conducted in five stages including (1) problem identification, (2) literature search, (3) data evaluation, (4) data analysis, and (5) data presentation. With Step (1), presented above, we identified a gap in our understanding of what and how ISPs use implementation strategies and thus of the skills that they require to provide implementation support in social and human services. All other review steps taken are presented in the following sections.
Literature Search
Integrative reviews are a method for systematically searching and synthesizing a targeted body of publications that is selected based on its potential for revealing new patterns and insights. While a comprehensive, representative set of the literature should be the basis of an integrative review (Torraco, 2016), with its key purpose being research integration for theory and framework development, it is not necessary to include every study ever published in the area of interest (Snyder, 2019). Rather, the aim is to achieve a level of “conceptual saturation” (Brunton et al., 2012, p. 120) through studies representative of, in this case, the work of ISPs.
With this in mind, a combination of five different search strategies was used to identify relevant publications for this review. Systematic searches of nine electronic databases (ASSIA, CINAHL, Criminal Justice Abstracts, ERIC, Family and Society Studies Worldwide, Medline, PsycInfo, Scopus, and SocIndex) were conducted and combined with searches of 42 different organizational websites, a targeted call for publications among selected experts, an open call through electronic media, and reference checks of all included studies. Searches were conducted between February and April 2019. A detailed account of all strategies has been included in the electronic results addendum (ERA; Albers, Bührmann, et al., 2020).
Qualitative as well as quantitative peer-reviewed primary studies that focused on the role of ISPs were eligible for inclusion. These could be studies of knowledge brokers, facilitators, consultants, coaches, intermediaries, improvement advisers, mentors, and so on. Given the diversity in terminology reflected in just these examples, further labels for implementation support roles were considered on a case-by-case basis if they emerged from studies. Studies of support provided through individuals in formal leadership roles or opinion leaders were excluded. No geographical limitations were defined. Publications had to be written in English, German, Danish, Swedish, or Norwegian. The detailed inclusion and exclusion criteria applied are described in the ERA (Albers, Bührmann, et al., 2020).
All work was conducted by a research team with five members, including four research assistants and the lead author. The team combined methodical expertise in systematic literature reviews (B.A. and C.V.), with content expertise related to social work (B.A. and L.Ba.), health services (C.V., P.D., and L.Bü.), and education (B.A.). The team used double-screening during both title and abstract and full-text screening, with conflicts being solved by a third team member. Weekly calls with all team members were used to discuss any issues emerging from screening, data extraction, or coding activities.
The flow of studies through the screening process is outlined in the ERA in figure ERA1 (Albers, Bührmann, et al., 2020).
Data Evaluation
Taking into account the purpose and character of this review and the available resources and instruments, a decision was made to focus the quality assessment of studies on randomized controlled trials only. In the absence of tools for assessing the risk of bias in rigorous implementation studies, we focused on exploring how developed studies of the still relatively new field of implementation support were using a framework developed by Hodder et al. (2016). It enabled us to consider the quality of included trials based on their integration of theory, consistency in terminology use, and balanced consideration of both benefits and harms of implementation interventions. These considerations were summarized and translated into five questions, which were answered for all included randomized controlled trials. The tool itself and the assessment of included implementation studies are presented in the ERA (Albers, Bührmann, et al., 2020).
The quality of all other studies included in this review remained unassessed. This was deemed acceptable in the current context, partly because implementation support is a novel field of inquiry, partly because this review was conducted for the purpose of theory development. Under these conditions, it can be valuable to review the entire range of existing evidence because even weaker studies may contain important information (Petticrew & Roberts, 2003).
Data Analysis
Descriptive study data were extracted using a standardized extraction form with 19 items to extract, including study design and aim, method, geography, sector, setting, sampling strategy, sample size, clinical intervention, ISP information, outputs, and outcomes. Each study was extracted by one research team member, and all data extraction quality was assured by the lead author.
Following this data extraction, all studies were uploaded to dedoose, an online qualitative data analysis platform for the coding of strategy use. This coding was guided by the 73 implementation strategies that comprise the ERIC compilation (Powell, Waltz, et al., 2015). The label of each strategy, for example, “distribute educational materials,” was entered into dedoose together with its basic definition. In this way, each strategy represented a single code. Research team members assigned these codes to text excerpts, which could be single words, one or more sentences, or entire paragraphs describing ISPs’ use of different strategies. Multiple codes could be assigned to one text excerpt. Strategies, which did not fit with any of the compilation labels, were coded as “other” and later analyzed separately.
The coding scheme was tested by three research team members on a sample of five studies, and results were discussed for further improvement of the scheme. It was then used with all studies, each of which was coded by one research team member. Of these studies, 42% were double coded by a second member of the research team, with any conflicts solved by the lead author. The coding of the remaining 58% of studies was quality assured by the lead author, through a review of all text excerpts during data analysis, and a recoding of these as necessary. Simultaneously with coding, all research team members generated memos to capture ideas related to specific aspects of strategy use, gaps, or questions emerging from the textual data.
Coding results for all studies were exported from dedoose by coding category, generating 44 different Excel spreadsheets each of which contained the raw coding data for a single code (i.e., ERIC strategy) used, including (a) publication identifiers, (b) the text excerpts relating to a particular code, and (c) the ISP category to which a publication belonged. Memos (N = 277) were exported separately grouped under nine different topics, each of which had been defined by different research team members during the coding process. The lead author reviewed each spreadsheet separately to assure the quality of all coding and highlight key content for each excerpt.
As part of the subsequent data comparison, the lead author reviewed all text excerpts in detail, reassigned codes (i.e., strategy labels) as relevant, identified preliminary ideas for central/common aspects of strategy use, and finally refined these through constant comparison with other text excerpts belonging to the same code. Since excerpts related to the implementation strategies “facilitation” and “providing ongoing consultation” showed to contain the same key elements and thus were not distinctly different from each other, a decision was made to merge these into just one code: “providing ongoing consultation/facilitation.” Furthermore, while a large number of findings were available, some of these were derived from only a single or very few included studies. To ensure that the content of the revised ISP program logic could be built on an acceptable minimum of scholarly agreement, it was therefore decided to only consider findings if they emerged from a minimum of 10 studies.
Results
Overview of Studies
A total of 109 publications formed the final sample of literature, representing 99 separate studies. A full list of these publications is included in the ERA (Albers, Bührmann, et al., 2020). The four most prominent ISP roles in this literature were facilitators (n = 22), consultants (n = 21), technical assistance (TA) providers (n = 12), and knowledge brokers (n = 10), followed by studies on coaches (n = 8), implementation teams (n = 4), and intermediaries (n = 3). Five studies examined a mix of different roles, and a further 14 studies either did not label the ISP role in focus (n = 5) or used other terms to coin it (n = 9).
Studies were primarily conducted in North America (United States, n = 67; Canada, n = 15), followed by the UK (n = 7), Central Europe (n = 4), Australia (n = 2), and New Zealand (n = 1). One further study involved multiple international locations, whereas geographical information was missing from two studies. The vast majority of studies was conducted within health services (n = 67), and a smaller number related to social welfare (n = 15), education (n = 14), and crime and justice (n = 1). A mix of sectors was involved in two studies. The type of RSIs implemented and supported included research-supported practices (n = 40), programs (n = 28), guidelines (n = 11), and policies (n = 2). Quality improvement approaches were the focus of a further 18 studies.
All articles included were published between 2000 and 2019 with 2015 being the median year, confirming the newness of this field of inquiry and indicating a growing research interest in implementation support in more recent years. Quantitative (n = 48), qualitative (n = 31), and mixed methods designs (n = 30) were used across the publications included. Studies used quasi-experimental (n = 4), pre–post evaluation (n = 22), or case study methods (n = 25). Cross-sectional designs were used in 12 studies, and a small group of studies used a multiple baseline (n = 2), a single-case approach (n = 3), and a cohort design (n = 1). Finally, 30 studies were randomized controlled trials, with the majority being randomized at the cluster level (n = 24).
In assessing their quality (see the ERA for details), it was apparent that less than 50% (n = 13) reported on the use of a theory, model, or framework to inform the design of the implementation support examined. This included the Reach, Effectiveness, Adoption, Implementation, Maintenance framework (n = 4), Getting to Outcomes (n = 2), the Consolidated Framework for Implementation Research (n = 1), the Promoting Action on Research Implementation in Health Services framework (n = 2), and the interactive systems framework (n = 1) together with organizational culture theory, the theory of planned behavior, the theory of continuous quality improvement, and Roger’s diffusion of innovations theory. However, the use of such theories and models was often described in passing, with only a few study reports providing detailed, theoretically informed explanations of how a support intervention was anticipated to work (Peterson et al., 2015; Quanbeck et al., 2018; Williams, Glisson, et al., 2017). Only three studies included descriptions of the use of local empirical evidence to ensure that the implementation support studied would fit with local conditions, two of which relied on previously collected data on the use of the intervention (Acolet et al., 2011; Calo et al., 2018) and one on the active codesign of the implementation support with stakeholders who had developed the guideline to be implemented locally (Quanbeck et al., 2018). No study referenced the distinct use of a strategy taxonomy in designing the implementation support examined.
Implementation Strategies Used by ISPs
In the studies reviewed, ISPs used 37 of the 73 implementation strategies that form the ERIC compilation (Powell, Waltz, et al., 2015).
Of these, 18 emerged as particularly common because each was named and/or detailed in more than 10 studies. Table 2 lists these strategies, the number of publications in which they appeared, and the cluster of strategies they belong to, as defined in a previous study (Waltz et al., 2015), aimed at validating the ERIC implementation strategies and grouping these into clusters. The ERA (Albers, Bührmann, et al., 2020) contains a table that summarizes the use of these implementation strategies by ISP role, indicating no clear differences in strategy use between these roles and confirming that there are strong similarities in their key activities.
Common Implementation Strategies Reported as Being Used by Implementation Support Practitioners.
a This column only contains strategies described in 10 or more studies. Additional strategies (n = 19) identified in texts, but in less than 10 studies each, included the following: conduct local consensus discussions (n = 6); conduct cyclical small tests of change (n = 7); facilitate relay of clinical data to providers (n = 1); increase demand (n = 4); purposely reexamine the implementation (n = 8); develop and implement tools for quality monitoring (n = 6); develop and organize quality monitoring systems (n = 5); identify and prepare champions (n = 6); use train-the-trainer strategies (n = 6); recruit, designate, and train for leadership (n = 6); provide local technical assistance (n = 4); obtain and use patient/consumer and family feedback (n = 5); access new funding (n = 4); remind clinicians (n = 3); start a dissemination organization (n = 2); stage implementation scale up (n = 1); revise professional roles (n = 1); provide clinical supervision (n = 1); and centralize technical assistance (n = 1).
b The Latin numeral added within a squared parenthesis is the number assigned to a strategy by Waltz et al. (2015). The Roman numeral describes the location of this strategy in one of four quadrants, each of which represents a particular combination of perceived strategy importance and feasibility in the study by Waltz et al. (2015). These numerals are used in this article to highlight the distribution of identified ISP strategies in Figure 1.
Table 2 shows that the majority of strategies described as being used by ISPs as part of their support work fell within the cluster of training and educating stakeholders (n = 8). These were followed by evaluative and iterative (n = 4) and interpersonal strategies (n = 4). A further two strategies focused on adapting and tailoring. No strategies belonged to the clusters of “engaging consumers,” “changing infrastructure,” or “utilizing financial strategies.”
The analysis of the literature also showed that implementation support was multifaceted and, in most cases, built on the combined use of multiple implementation strategies. An average of 4.5 different implementation strategies were reported as being used by ISPs across all randomized, controlled trials (n = 30), with a maximum of nine and a minimum of two strategies (SD = 1.9). These represented the entire range of strategies listed in Table 2.
Opening the Black Box of Strategy Use by ISPs
Table 3 provides an overview of the aspects of strategy use that could be identified through the literature. Identifiable types of use for each strategy are listed together with techniques that were reported to support strategy operationalization and components that were apparent as forming strategy use across multiple publications. Since aspects of using strategies with an educational purpose, that is, conduct educational meetings/ongoing training/outreach visits, develop/distribute educational materials, and create learning collaboratives, were very similar, these were integrated into one strategy labeled “education” in Table 3.
Types, Techniques, and Components of ISPs’ Implementation Strategy Use.
Note. ISP = implementation support practitioner.
a “Education” covers aspects identified across the following strategies: conduct educational meetings, distribute educational materials, conduct ongoing training, conduct educational outreach visits, make training dynamic, develop educational materials, and create a learning collaborative.
In the following, these aspects of strategy use are outlined in greater detail.
Train and educate stakeholders
When ISPs are providing consultation/facilitation, five common components emerged across studies: (1) identifying the support needs of those involved in the consultation/facilitation efforts, for example, through formalized, periodic needs assessments (Duffy et al., 2012), structured interviews (Bice-Urbach & Kratochwill, 2016), or by explicitly inviting stakeholders to articulate their support needs in each session (Akin, 2016; Chilenski et al., 2016); (2) educating and professionally supporting these stakeholders, for example, through processes such as learning from others (Akin, 2016), role-plays (Barac et al., 2018), didactic teaching (Beidas et al., 2013; Chaffin et al., 2016), answering questions (Chilenski et al., 2016; Hurtubise et al., 2016; Kelly et al., 2000), or offering advice (Rosen et al., 2012); (3) monitoring the progress and/or performance of stakeholders, for example, by measuring fidelity (Bice-Urbach & Kratochwill, 2016; Caron & Dozier, 2019; Eiraldi et al., 2018; Murray et al., 2018), program outcomes (Funderburk et al., 2015; Olson et al., 2018), or progress toward other implementation or service goals (Chilenski et al., 2016; Holtrop et al., 2008; Preast & Burns, 2018); (4) identifying implementation barriers and problems faced as part of the change efforts, typically related to learning a new practice (Dusenbury et al., 2010; Eiraldi et al., 2018; Kauth et al., 2010; Nadeem, Gleacher, Pimentel, et al., 2013) and/or enabling its implementation within a particular local context (Rosella et al., 2018; Saldana & Chamberlain, 2012; Tierney et al., 2014); and (5) identifying potential solutions to these problems, including next steps to initiate these. This final step was at times characterized as “troubleshooting” (Chaffin et al., 2016; Hodge et al., 2017; Meropol et al., 2014) signaling a more urgent and ad hoc type of character that this strategy could take. Other studies emphasized that consultation/facilitation depended on a climate conducive to jointly identifying and solving problems among multiple stakeholders (Dogherty et al., 2012; Hurlburt et al., 2014), thereby indicating that it could take time for ISPs to create adequate conditions for using this strategy.
Studies reflected that consultation/facilitation was provided to individuals (Bradshaw et al., 2012; Rosen et al., 2012) as well as groups of stakeholders (Kousgaard & Thorsen, 2012; Murray et al., 2018) and occurred either in-person (Anyon et al., 2016; Dobbins et al., 2018; Eiraldi et al., 2018), or remotely (C. H. Brown et al., 2014; Gustafson et al., 2013; Kauth et al., 2010). The use of videotaped work samples or the direct observation of the work of those supported was described in multiple studies as a key cross-component activity for ISPs (Akin, 2016; Bice-Urbach & Kratochwill, 2016; C. H. Brown et al., 2014; Caron & Dozier, 2019; Dusenbury et al., 2010; Funderburk et al., 2015).
While the description of ISPs’ educational activities often lacked detail, study reports displayed a broad spectrum of types of meetings, ranging from formal, comprehensive series of meetings, to informal quick check-ins. On the formal end of the spectrum, ISPs conducted, for example, an “onsite 6-hour experiential workshop” (Carson et al., 2014, p. S14), “a series of six learning sessions” (Anaby et al., 2015, p. 3), a “two-day initial training” (Bradshaw et al., 2012, p. 181), or “study days” (Gerrish et al., 2011, p. 2010). The use of less formal gatherings aimed at educating stakeholders included, for example, “addressing the subject informally during lunch breaks” (Aasekjær et al., 2016, p. 35), offering one-to-one tutoring (Bice-Urbach & Kratochwill, 2016; Kaasalainen et al., 2015), integrating educational elements into an agency’s routine staff meeting (Byrnes et al., 2018; Graaf et al., 2017), or initiating ad hoc informal training sessions upon request by those supported (Tierney et al., 2014). Study reports that presented the techniques used to educate stakeholders in these meetings reflected a consistent combination of didactic and dynamic, interactive teaching elements (Becker et al., 2013; Beidas et al., 2012; Brownson et al., 2007; Chaffin et al., 2016; Dobbins et al., 2018; Ryba et al., 2017; Tierney et al., 2014; Yano et al., 2008), the latter of which typically aimed at integrating the concrete and individual/local work experience of those supported.
Use evaluative and iterative strategies
The use of audit and feedback involved multiple activities for the ISP, including sourcing relevant administrative or statistical data or self-collecting such data (Burns et al., 2008; Dickinson et al., 2014; Jacobson et al., 2019; Meropol et al., 2014), summarizing these data in relevant and operational ways (Calo et al., 2018; Chaple & Sacks, 2016; Chinman et al., 2018; Mold et al., 2008), presenting them for stakeholders (Holtrop et al., 2008; Jacobson et al., 2019), and promoting a discussion of findings enabling stakeholders to set priorities for improvements (McCullough et al., 2017; Preast & Burns, 2018). Multiple studies also included descriptions of ISPs tracking and assessing fidelity data to better understand whether interventions were implemented as intended by their developers (Brunette et al., 2008; Chinman et al., 2018; Eiraldi et al., 2018; Gunderson et al., 2018; Kirchner et al., 2014), highlighting that evaluative functions were included in implementation support.
ISPs also took part in assessing the needs for different clinical or educational interventions (Dogherty et al., 2013; Fortney et al., 2018; Ward et al., 2017; Waterman et al., 2015). This could occur in formal ways, for example, in using a particular method such as the nominal group technique (Quanbeck et al., 2018) or as part of purposely scheduled site visits (Anaby et al., 2015; Rosella et al., 2018). In addition, study reports included descriptions of informal approaches to needs assessments, necessary because stakeholders’ support needs changed over time and required to be continuously updated. In these cases, the identification of needs was routinely integrated into ISP observation of practice and the regular contact they had with those supported (Becker et al., 2013; Duffy et al., 2012; Rivard et al., 2010).
A slight broadening of scope could be observed for the strategy to develop a formal implementation blueprint. Originally defined with a focus on planning the implementation of a clinical intervention, the literature documents that ISPs were also involved in developing both more targeted plans in the form of, for example, “coaching delivery plans” (Rushovich et al., 2015, p. 373) or “coping plans” (Sanetti et al., 2018, p. 52), describing how to handle potential implementation barriers, and higher level plans described as, for example, “project charters” (Lavoie-Tremblay et al., 2012, p. 421). In only a few cases did study reports outline what this planning implied, for example, using “a checklist […] to guide the decision-making process (e.g., identifying [intervention] target population and exclusion criteria, [intervention] staff members’ roles and responsibilities, and how to monitor the implementation process)” (Ritchie et al., 2017, p. 5).
Multiple studies described ISPs as being involved in readiness assessments, which implied ISPs surveying (Holtrop et al., 2008; Russell et al., 2010) or interviewing stakeholders (Peterson et al., 2015), applying “formative evaluation techniques” (Ritchie et al., 2017, p. 5), or conducting a series of readiness calls (Brown et al., 2014). The elements of readiness assessments and tools used to conduct them were detailed in only a few studies (Russell et al., 2010; Waterman et al., 2015). Other studies reflected that readiness assessments could have a particular scope and focus on, for example, identifying stakeholders’ readiness for training (Gerrish et al., 2011; Worton et al., 2018) or implementation support in general (Feinberg et al., 2008; Peterson et al., 2015; Yazejian et al., 2019).
Develop stakeholder interrelationships
The importance of network weaving to the work of ISPs was reflected in poignant terminology used in study reports, characterizing them as a bridge (Elnitsky et al., 2015), linking agent (Dogherty et al., 2012), boundary spanner (Dogherty et al., 2012; Graaf et al., 2017), connector (Franks & Bory, 2015), liaison (Dogherty et al., 2012), and convener (Worton et al., 2018). ISPs were described as fulfilling this function in two ways. Firstly, at a basic level, they shared their knowledge about relevant services and agencies that those they supported could benefit from in their own practice. This implied, for example, connecting clinical practices to available community services (Brown, Elliott, & Leatherdale, 2018; Brown, Elliott, Robertson-Wilson, et al., 2018a, 2018b; Mader et al., 2016). Secondly, at a more advanced level, ISPs worked to overcome barriers to stakeholder connectivity, thereby bringing individuals, organizations, and systems together for collaborative processes that these otherwise would struggle to establish themselves (Waterman et al., 2015; Worton et al., 2018; Yazejian et al., 2019). This second type of network weaving was described as requiring greater neutrality from ISPs (Worton et al., 2018), enabling them to consider perspectives that exist outside of the context that received their direct implementation support.
Through these types of network weaving, ISPs represented a connection resource to their stakeholders that they themselves needed to maintain and expand continuously. This was reflected in studies describing ISPs as needing to regularly develop their own networks as a form of social capital that was of general benefit to their work and helped to, for example, ensure that later implementation activities ran more smoothly or crucial information was accessible when needed (Waterman et al., 2015).
In using change modeling as an implementation strategy, ISPs pursued primarily two purposes: Firstly, to build specific skills in those supported, for example, clinical skills required by practitioners to deliver an intervention. These were taught through, for example, role-plays (Akin, 2016; Barac et al., 2018; Caron & Dozier, 2019; Dogherty et al., 2012), formal shadowing (Gerrish et al., 2011), walk-throughs (Jacobson et al., 2019), or participant modeling (Funderburk et al., 2015; Graaf et al., 2017; Kinley et al., 2014; Sanetti et al., 2018). Secondly, ISPs were described as using the exact same techniques, which they wanted their stakeholders to use with patients or clients, in their own implementation support. In a study of consultation, ISPs used the principles of motivational interviewing (MI) in their support of clinicians learning to apply MI in their clinical practice (Barac et al., 2018). Similarly, in another study, knowledge brokers used the principles of EBP in as many aspects of their support work as possible (Hurtubise et al., 2016).
In identifying and preparing champions for implementation, ISPs collaborated with formal leaders, labeled as “clinical leads” (Acolet et al., 2011), “senior management” (Dobbins et al., 2018), or just “leadership” (Chaffin et al., 2016; McCullough et al., 2017; Ritchie et al., 2017). A few examples also pointed to ISPs connecting with, for example, full steering committees (Brunette et al., 2008), boards (Ward et al., 2017), or government departments (Dobbins et al., 2018). The function of these connections was predominantly described as to update champions, “ensuring the right individuals are informed” (Dogherty et al., 2012, p. 9) about the state and progress of the implementation. However, implicitly, this also aimed to engage champions and to garner their support (Kirchner et al., 2014; McCullough et al., 2017; Ritchie et al., 2017). In addition, examples could be identified of ISPs presented as explicit and proactive advisors to implementation or system leaders (Brunette et al., 2008; Chaffin et al., 2016).
Organizing clinician team meetings involved ISPs supporting and/or participating in teams explicitly built to shepherd the implementation effort and labeled, for example, project team (van der Zijpp et al., 2016), change team (Fortney et al., 2018; Quanbeck et al., 2018), interdisciplinary facilitation team (Lessard et al., 2016), improvement team (Dickinson et al., 2014), community development team (Saldana & Chamberlain, 2012), or interagency collaborative team (Chaffin et al., 2016). ISPs could be full members of these teams, with all members providing different aspects of implementation support (Chaffin et al., 2016; Fortney et al., 2018), or they could be in an assisting role in which they supported a team in its local work to promote the implementation of an intervention (Jacobson et al., 2019; Saldana & Chamberlain, 2012). This means that the key function of this strategy, described in the ERIC compilation as focused on enabling shared learning among clinicians, had a broader scope in the ISP literature. While shared learning and problem-solving was a goal for teaming (Kaasalainen et al., 2015; Parchman et al., 2013; Saldana & Chamberlain, 2012), for ISPs these teams also represented an infrastructure that made it possible to connect the parts of a system that were involved in and affected by an implementation and to include these in their support activities. The following excerpt illustrates this for a change team that received regular implementation support from consultants: “Multidisciplinary representation on the team was important. […] changing a workflow requires understanding the tasks performed by staff members in all occupations involved and securing their cooperation to make the change” (Jacobson et al., 2019, p. 5). The multidisciplinary nature of teams pointed to in this quote also emerged as a characteristic from other studies (Kaasalainen et al., 2015; Lessard et al., 2016), indicating that ISPs worked with teams with considerable professional diversity, reflected by team members’ educational backgrounds, professions, or organizational roles.
Adapt and tailor to context
In coding for tailoring strategies, the focus was on identifying descriptions of how ISPs tailored their own support, that is, the consultation, facilitation, or TA they offered. While tailoring was the dominating term used in this context, adapting, modifying, individualizing, fitting, customizing, or matching were alternative wordings used in study reports. Close to a third of these studies (n = 13) provided little detail on the factors informing the tailoring practice of ISPs, while a further 24 studies included some, but still relatively general, detail about what caused and informed tailoring processes. These could be structured into two main groups.
The first group explained tailoring as a response to the needs expressed by or identified for the individuals supported by ISPs, for example, professional development needs (Akin, 2016; Becker et al., 2013; Dusenbury et al., 2010; Shernoff et al., 2015), needs emerging from particular preferences (Anaby et al., 2015; Barac et al., 2018; Dobbins et al., 2018), or from individuals’ organizational role (Gerrish et al., 2011; Gustafson et al., 2013; Rivard et al., 2010). The second group of publications described factors in the inner setting in which the implementation support was provided as causing the tailoring, including organizational culture and climate (Aasekjær et al., 2016; Anaby et al., 2015; Garbacz et al., 2016), policies (Anaby et al., 2015; Brunette et al., 2008) and priorities (Kelly et al., 2000), resources (Chinman et al., 2017), or structures (Brunette et al., 2008; Garbacz et al., 2016). Only one study pointed to outer setting factors, in the form of other health initiatives and organizations operating in the external environment to an implementation, as a potential trigger for tailoring (Tierney et al., 2014). Finally, in a study of intermediaries (Chew et al., 2013), the professional skills and interests of the ISP were described as leading to tailoring in a system in which this role was newly established.
There was very little information about the elements of implementation support that were tailored. Studies containing such detail pointed to, for example, intensity and focus (Meropol et al., 2014), delivery mode (Yazejian et al., 2019), resources (Russell et al., 2010), tools (Rosella et al., 2018), or documentation (Mackenzie et al., 2011) with more fine-grained accounts being a rare exception (Quanbeck et al., 2018).
Finally, ISPs worked to support adaptation. These adaptations aimed to address the clinical needs of specific target populations (Beidas et al., 2013; Shernoff et al., 2015), or their cultural preferences (Chaffin et al., 2016; Hurlburt et al., 2014), and also to address the local context of the implementation, such as resources (Saldana & Chamberlain, 2012) or the interests and preferences of individuals using a new intervention (Rosella et al., 2018; Waterman et al., 2015). The specific activities conducted by ISPs included surveying stakeholders to assess whether interventions were deemed feasible and acceptable and inform adaptation needs (Waterman et al., 2015); sourcing, translating, and applying evidence to guide adaptation (Beidas et al., 2013); liaising between intervention developers and providers to enable adaptation (Chaffin et al., 2016; Fortney et al., 2018; Saldana & Chamberlain, 2012); helping to adapt guidelines, protocols, tools, and other intervention resources used by providers and other key stakeholders (Parchman et al., 2013; Quanbeck et al., 2018; Rosella et al., 2018); and tracking and documenting adaptations to ensure these were sufficiently captured (Yano et al., 2008).
Strategy Feasibility
The 18 common implementation strategies listed in Table 2 also were assigned a unique identifier in the form of Latin numerals added within squared parentheses. These identifiers were previously used in a study conducted by Waltz et al. (2015), aimed at locating each strategy in a diagram (Figure 1) by its degree of importance and feasibility.

The feasibility and importance of implementation support practitioner strategies.
That study, which was based on a concept mapping process involving 35 experts, led to the positioning of each strategy in one of four quadrants—labeled I–IV (the Roman numerals in Table 2)—each of which represents a particular combination of perceived strategy importance and feasibility. This is displayed in Figure 1, in which all 18 ISP strategies presented above and listed in Table 2 have been circled in black. It shows that the vast majority (n = 13) of these falls into Quadrant I. This quadrant represents strategies that were rated as highly important and highly feasible. Three further ISP strategies fall into Quadrant II representing strategies deemed equally feasible as those in Quadrant I but less important. Finally, two strategies—promote network weaving and model and simulate change—belong to Quadrant III displaying strategies viewed by experts as being less feasible and less important. None of the central ISP strategies identified as part of this review fell into Quadrant IV, that is, were strategies viewed to be of higher importance but less feasible.
Other Implementation Strategies
While the ERIC strategies supported the extraction and coding of substantial amounts of text material, not all ISP activities could be covered through this compilation. Forty-three publications examining all ISP roles presented information that initially was coded as “other” and could be synthesized into further two strategies. While the literature on knowledge brokers contributed considerably to the identification of these strategies, studies covering other roles also described their use.
The first strategy was labeled “source, share, and translate evidence of relevance to stakeholders involved in the implementation.” ISPs applying this strategy were described as “culling through the research” (Cameron et al., 2011, p. 30), selecting journal articles, or measures (Anaby et al., 2015; Hurtubise et al., 2016; Waterman et al., 2015), sharing and/or summarizing these resources (Chew et al., 2013; Gerrish et al., 2011), and helping to apply them (Dogherty et al., 2012; Gerrish et al., 2011). This application could relate to the intervention to be implemented (Beidas et al., 2013), but it could also be used for advocacy and other efforts to influence policy or practice (Franks & Bory, 2015). A small sample of studies also highlighted a precondition of using this strategy, namely, that ISPs needed to be avid and competent consumers of research (Dogherty et al., 2013; Gerrish et al., 2012; Hurtubise et al., 2016; Rivard et al., 2010).
The second strategy falling outside the ERIC compilation was “contribute to intervention design” and refers to situations in which ISPs were involved in developing clinical protocols (Aasekjær et al., 2016; Holtrop et al., 2008), behavior support plans (Bice-Urbach & Kratochwill, 2016; Sanetti et al., 2018), guidelines (Byrnes et al., 2018; Gerrish et al., 2011), practice policies (Lemelin et al., 2001), best practice models (Franks & Bory, 2015), or evidenced interventions to be integrated into clinical practice (Dobbins et al., 2018; Hurtubise et al., 2016; Waterman et al., 2015). Few studies provided further detail on the look and feel of this activity.
Discussion
This research identified 18 discrete implementation strategies commonly used by ISPs when providing implementation support in human and social service settings. No clear differences in strategy use could be identified across different ISP roles (e.g., facilitators, knowledge brokers, TA providers, consultants), indicating considerable similarity in their work and confirming that greater research integration in the field of implementation support is relevant.
The range of strategies identified suggests that an ability to train and educate stakeholders; to continuously monitor, evaluate, and adapt implementation; to develop stakeholder interrelationships; and to tailor one’s own implementation support are central skills of ISPs. Furthermore, the reported variability with which strategies were used by ISPs reflects that these strategies can be further broken down into concrete activities, components, and techniques. This highlights that ISPs also require skill in operationalizing and detailing the implementation strategies that they decide to integrate into their implementation support. Finally, with implementation support being generally characterized by the use of multifaceted strategies, the ISP role demands mastery in selecting, combining, and using multiple strategies. Mastery describes an aptitude for accessing strategy knowledge on demand, flexibly linking and tailoring this knowledge to situational and contextual conditions, with the purpose of enabling sustainable learning and skill building in others.
The use of financial strategies, infrastructure changes and the engagement of consumers, were absent in the ISP literature. Two strategies—“source, share, and translate evidence of relevance to stakeholders involved in the implementation” and “contribute to intervention design”—were identified as not being part of the ERIC compilation, yet, relevant to the work of ISPs. These strategies are therefore suggested as additions to the ERIC compilation.
Implications
In scholarly debates about the development of the social work profession, implementation science has been introduced as a field of inquiry that, due to its applied nature, can help to create stronger bidirectional ties between research and practice and thereby enhance the relevance and quality of social work as well as its capacity to address societal problems (Bunger & Hall, 2019; Cabassa, 2016; Gehlert et al., 2017). A central point of attention in these debates is the connectedness of research and practice, with learning collaboratives (Bunger et al., 2016; Stephens et al., 2014), partnerships between academic and social and human service organizations (McBeath et al., 2019; Palinkas et al., 2017), and the research-minded practitioner (DePanfilis, 2014; Liedgren, 2020) being among the suggested solutions for how to create closer linkages between these domains. With this review, we add a further potential solution to this list—the ISP—and propose for it to be considered as a role that can actively bridge the research–practice gap in social work and help human and social service organizations to establish and facilitate the adaptive learning required not only to implement ready-made RSIs but also to apply evidence in the design and improvement of local practice (Mosley et al., 2019).
However, the knowledge about this role when used in social work is scarce, pointing to a need for developing programs, initiatives, and funding structures at the system level to generate a broader experience with utilizing it in different forms and contexts. As part of such initiatives, the specific conditions and characteristics of social work and its providers—human and social service organizations and their staff—should be taken into account. This would include, for example, the central role of peer influence (Wharton & Bolland, 2012; Wike et al., 2014), of inter-agency networks and collaboration (Bergmark et al., 2018; Palinkas et al., 2011, 2012), and of supervisors and managers (Bäck et al., 2020; Bunger et al., 2019) for social workers’ motivation and ability to integrate evidence into their daily routines. These and other individual and organizational factors form the fabric into which implementation support—be it delivered by an individual or a team, an external or an internal unit—would need to be integrated.
At the level of service provision, the breadth and variability of ISP strategy use identified through this review, and the high level of skill required emerging from this use, naturally raises the question of how to select, recruit, or develop professionals for ISP roles in human and social service settings.
With EBP, RSIs, and implementation science still not being widely integrated into training and professional development programs on the one hand and routine human and social services on the other, highly skilled ISPs can be expected to remain in short supply in the future. Moreover, even if an organization identifies a single individual who is an experienced practitioner and familiar with EBP as well as implementation support, relying solely on this one person would create vulnerable implementation capacity at risk of disappearing quickly in the event of staff turnover, budget cuts, or other challenges that happen routinely in human and social service agencies. The literature points to two potential pathways for minimizing such vulnerability.
Providers of human and social services may want to consider whether implementation support could be distributed across a team, whose members contribute with different types of skill, expertise, and experience. This thinking aligns with a central point made in the literature on facilitation, which emphasizes that facilitation represents both a role and a process (Dogherty et al., 2010). This indicates that facilitation does not necessitate the establishment of a single, formal facilitator role and can take place as long as its key functions are appropriately represented by different members of an organization. Within human and social services, this means that implementation support activities such as identifying and preparing champions, consultation, or informing local opinion leaders could be performed by different members of an implementation team. In recent years, this team approach to implementation support has been increasingly discussed in the literature (Higgins et al., 2012; Metz & Bartley, 2020). It was also applied in a small number of studies included in this review (Chaffin et al., 2016; Hurlburt et al., 2014; Lessard et al., 2016) and described as usable within as well as across organizations.
The intraorganizational model involves establishing an internal implementation support team formed by an agencies’ own staff (Lessard et al., 2016). In the cross-organizational model, multiple agencies work together to establish the implementation support team, each contributing different personnel. In this model, decision makers need to be prepared to balance different organizational cultures, interests, and priorities that exist among participating agencies. If these are highly diametric, tensions may emerge among stakeholders (Aarons et al., 2014) and complicate the use of the team approach.
A second pathway toward developing implementation support roles can be to collaborate with an intermediary organization specialized in providing implementation support. While research on these intermediaries remains scarce (Proctor et al., 2019), descriptive studies of their work to support the implementation of RSIs in human and social services exist. These confirm the positioning of intermediaries at the nexus of research, practice, and policy and identify the capacity building as one of their key functions (Cheron et al., 2019; Isett & Hicks, 2019; Smits et al., 2020; Weaver et al., 2017). Human and social service agencies can use intermediary staff as a temporary external resource available during implementation efforts and working to build internal implementation support capacity that, in the longer term, will make the agency independent of the support provided by the intermediary.
Of interest in this context are data that were collected through a recent survey (Proctor et al., 2019), administered with 54 intermediary and purveyor organizations in the United States. The findings from this survey showed that these organizations used a range of 32 distinct implementation strategies across five domains, indicating a greater breadth of strategy use than identified through this review. This breadth was later confirmed as part of a program evaluation, reporting that intermediaries used 31 strategies in supporting the early implementation of three different manualized RSIs in Australia (Albers, Hateley-Browne, et al., 2020). One explanation for this difference may be the underreporting of strategy use in the studies included in this review, a challenge commonly acknowledged in the literature (Bunger et al., 2017; Hooley et al., 2020; Pinnock et al., 2017b; Varsi et al., 2019). Furthermore, both of the above studies included implementation support as delivered by purveyors, that is, companies focused “on the dissemination of a specific EBP [evidence-based practice] with the goal of implementing the EBP with fidelity and good effect” (Franks & Bory, 2015, p. 43). As they depend on the success of their products, these companies may tend to intensify their implementation support and in doing so draw on a greater number of strategies. However, the numerical differences in strategy use may also reflect that implementation support capacity, when professionalized and institutionalized within the organizational settings of intermediaries, makes it possible to develop expertise in applying a broader range of strategies—for example, because a more diverse set of knowledge and skills is available across multiple ISPs working for an intermediary; their access to contacts and networks in academia, practice, or the policy sphere is broader; or their pooled funding allows for testing and developing a broader range of implementation support activities. If so, the collaboration with an intermediary may provide opportunities for tapping into specialized expertise that would be difficult to generate by a single human or social service agency alone.
For intermediary organizations, this presupposes that their staff are sufficiently skilled in applying implementation strategies, that is, in selecting, operationalizing, designing, and tailoring them. While this kind of strategy work requires further investigation and therefore continues to be a high priority for implementation scientists (Powell et al., 2019), scholars confirm its dependency on appropriate methods that allow for, for example, the integration of multiple stakeholder perspectives, the identification of crucial barriers to implementation, or the assessment of available resources for strategy development. Concept and intervention mapping, group model building, and conjoint analysis have been identified as such methods (Fernandez et al., 2019; Powell, Beidas, et al., 2015), all of which should belong to an ISP’s toolbox. In addition, the use of theory has been highlighted as an important feature of implementation strategy design (Lyon et al., 2019), further adding to the knowledge and skill level required by ISPs and highlighting the importance of continuously promoting their professional development alongside progress in implementation science. Given their organizational capacity, intermediaries may be in a particularly strong position to meet these needs, but also human and social service organizations developing internal implementation support roles or teams should be aware of them and consider how this skill building can become a routine practice.
Within the field of implementation science, the range of implementation strategies reported to be used by ISPs should remain a topic for further investigation because it raises a number of critical questions.
At the basic level, the identification of strategies that are not currently included in the ERIC compilation—“source, share, and translate evidence of relevance to stakeholders involved in the implementation” and “contribute to intervention design”—suggests that there may be a need to review the compilation and to consider whether it fully reflects the realities of implementation as it is practiced and researched in different countries today. It was developed at an earlier developmental stage of implementation science, and the broadened implementation literature and experience that is available may help to refine it further. For example, adding strategies could improve the compilation and make it more relevant to social work as new strategies may relate more to interventions used in this profession. Furthermore, given a growth in knowledge about how strategies are used in practice, they could be more clearly delineated from each other in a new version of the compilation, thereby enhancing its clarity and usefulness for both research and practice.
The general literature on implementation strategies has also discussed the persistent challenges of implementers in appropriately matching implementation strategies with implementation barriers, reflected in difficulties with identifying relevant implementation strategies, and with using these with the proper frequency, intensity, and fidelity required to achieve their intended benefit (Eisman et al., 2019; Powell et al., 2019, 2020; Waltz et al., 2019). Given that the findings of this review indicate that ISPs primarily apply more feasible strategies, it is relevant to ask why financial strategies or strategies aimed at changing infrastructure or engaging consumers were less present in the literature. One explanation might be that the strategies used were deemed to be most appropriate, given the context in which support was provided. However, it could also reflect that they were “more immediate and concrete and […] potentially more in the control of those tasked with supporting change” (Waltz et al., 2015, p. 6). This latter interpretation suggests that ISPs may have to neglect potentially effective support strategies if they are out of their control, for example, because their use would require the continuous engagement of senior management or deeper structural changes in an organization. If this is a characteristic of implementation support work, then ISPs—rather than challenging the values, norms, or power structures of the systems in which they work—may be at risk of just “conforming to existing ways of doing things” (Kislov et al., 2016, p. 472). This warrants a deeper examination of the use of strategies by ISPs in order to better understand their potential as true agents of change. Relevant research questions to address in this context would be, for example: What is the rationale underpinning ISPs’ use of particular implementation support strategies? How does the use of strategies by ISPs change when their roles are positioned/set up differently within a service system (e.g., internal vs. external roles, within vs. outside of leadership structures)? What characterizes unsuccessful attempts of using particular implementation support strategies?
Finally, in confirming that the use of strategies did not differ substantially across different ISP roles, and in identifying two additional implementation strategies commonly used across these roles, this review highlights the value of greater implementation research integration. Variability in terminology—centered on, for example, the differences between dissemination, knowledge mobilization, translation, and implementation—has been a characteristic of implementation science since its invention in the early 2000s (Graham et al., 2006; Khalil, 2016; Rabin et al., 2008), partly due to the field’s multidisciplinary roots (Rabin et al., 2008). While a certain level of variability in terminology may be unavoidable, the investigation of different implementation support roles makes it visible that this variability also may have a cost. Different phrases used to coin particular implementation support roles appear to have generated separate streams of research as unconnected “schools of thought,” despite considerable similarities among them. The potential consequence within science is a waste of research resources and an unnecessarily fragmented knowledge base, leading to implications for decision makers in policy and practice for whom the navigation and use of this knowledge base becomes needlessly complicated. Future research activities in this area of implementation science should therefore be conducted from an integrative perspective, that is, draw on the broadest possible evidence base existing across multiple schools of thought and producing knowledge that is independent from particular ISP role labels.
Limitations
Multiple limitations have to be taken into account when considering the findings and implications of this study. Firstly, while searches conducted for this integrative review were systematic, not every single study examining the work of particular ISP roles would have been captured since concept saturation guided its production. Hence, readers interested in the detailed knowledge base that exists for specific ISP roles should consult the literature for these roles separately. Furthermore, readers should keep in mind that implementation science is a relatively new field of inquiry. Literature may exist in which novel implementation terminology has not been used but functions comparable to implementation support have been described. The search terms applied in this review build on the common language used in the field of implementation science to characterize this support. Hence, studies published, for example, in other, unrelated fields of science or before this field emerged, may therefore have been missed.
Secondly, the evidence base presented with this review contains a relatively small share of studies conducted in human and social service settings. The decision to take a cross-sector perspective was intentional in that the evidence base in the human and social service sector was expected to be limited and hence insufficient to achieve concept saturation. Against this backdrop, some of the review findings will be applicable across sectors, that is, social, health, and educational settings, while the utilization of other findings will require a translation that takes into account the specific realities of, for example, human and social service organizations such as resource scarcity, high rates of turnover, or fast-paced working routines.
Thirdly, the results from this review may be affected by an underreporting of implementation strategies in the studies included. While standards for reporting implementation studies have been developed (Pinnock et al., 2017a), these are still relatively new and not necessarily applied across eligible studies. The inclusion of diverse study designs—incorporating a considerable number of detailed qualitative studies of implementation support—aimed at capturing the broadest possible range of strategies used by ISPs. However, readers should be aware that not all aspects of strategy use may be presented here.
Finally, in using the ERIC strategy compilation for examining ISP strategy use, we chose a particular lens for our analysis. The fact that some of its implementation strategies appeared to be suitable to be merged and others could be further broken down into detailed activities indicates that not all of its different strategies may be fully discrete units that can be clearly separated from each other. This raises a question about the compilation’s maturity as an analytical tool and highlights the importance of continually refining taxonomies as the field of implementation science progresses. In the meantime, readers should be aware of a certain blurriness in current implementation strategy definitions. Furthermore, advocates of relational theory have criticized the often-gendered nature of the ways in which we understand organizational phenomena, reflected in, for example, overemphasizing the linearity of change processes or the use of technical strategies in enabling such change (Fletcher, 1998). Viewed from this perspective, the ERIC compilation would represent a masculine biased view on implementation, which neglects the importance of, for example, connection, interdependence and collectivity in implementation processes, and the potential of relational strategies aimed at, for example, creating reciprocal dialogue among stakeholders to improve the use of evidence in practice (Metz et al., 2019). Further studies of ISPs may therefore benefit from utilizing alternative theories and perspectives in their exploration of this still relatively new role in implementation.
Conclusion
Different ISPs use a similar set of implementation strategies, indicating considerable similarity in their work as well as their required skill set and confirming that greater research integration in the field of implementation support is relevant. The breadth and depth of strategies used by ISPs suggests that strong professional development pathways are needed to enable staff of human and social service and of intermediary organizations to build, utilize, and offer implementation support skills. However, the limited range of implementation strategies applied by ISPs indicates that this role may be far from reaching the level of change agency that stakeholders in policy and practice would require for a system wide move toward implementation-informed EBP in human and social services. Further development of the ISP role for this sector is therefore urgently needed.
Footnotes
Authors’ Note
Cecilie Varsi is also affiliated to European Implementation Collaborative, Soborg, Denmark.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
