This scoping review aims to systematically examine and synthesize existing research on the cross-sectoral implementation of rehabilitation in Nordic health systems, focusing on factors that influence implementation processes to inform and support future implementation efforts.

To be eligible for inclusion in this review, the studies had to: •

Target an adult population aged 18 years or older

The following databases were searched to identify relevant studies: MEDLINE (via OVID), EMBASE (via OVID), CINAHL (via EBSCOhost), Physiotherapy Evidence Database (PEDro), Occupational Therapy Systematic Evaluation of Evidence (OTseeker), Klinisk Sygepleje (via Idunn) and Web of Science. No limitations or restrictions were applied regarding language or date of publication. All databases were systematically searched in November 2023. The final search strategies can be found in Supplementary File S1. In addition, the full-text articles were hand-screened for additional records as well as run through Web of Science, adding a snowballing effect to the database search. All studies identified through the search strategy were uploaded into Covidence for study selection. Here, all duplicates were excluded.

The search strategy aimed to identify published, peer-reviewed primary studies. Grey literature was excluded to keep the scope of the review manageable and to ensure a transparent and reproducible search process. An initial search in Cochrane was conducted to expand the search and index terms. These keywords and index terms were tailored to each information source. Two information specialists peer-reviewed the search strategy in accordance with the PRESS guidelines (25).

A four-part PICO model consisting of Population, Intervention, Context/Setting, and Outcome (26) inspired and guided our search. For each component of the PICO model, a combination of Medical Subject Headings (MeSH), keywords, free-text terms, and word variants was employed to ensure comprehensive coverage. For the Population component, the focus was on adults aged 18 years and older residing in the Nordic countries, with country-specific identifiers such as Denmark, Danish, Sweden, Swedish, Norway, Norwegian, Finland, Finnish, Iceland, Icelandic. The Intervention component targeted rehabilitation-related concepts, including terms such as rehabilitation, exercise, physical activity, activities of daily living, leisure activities, recreational therapy. For the Context/Setting, the emphasis was on cross-sectoral care and transitional processes, with terms such as cross-sector, hospital discharge, continuity of patient care. Finally, the Outcome component focused on implementation processes and evaluations, with terms including implementation, outcome assessment, program development, quality control, delivery of health. This comprehensive search strategy was designed to capture relevant literature addressing the implementation and evaluation of rehabilitation interventions across sectors in Nordic adult populations (Supplementary File S1).

To enhance consistency among reviewers, each reviewer initially screened 20 papers to become familiar with the inclusion and exclusion criteria and then discussed any questions or concerns regarding the screening process with HB. Following this, the inclusion and exclusion criteria were refined, particularly regarding the definitions of rehabilitative interventions and implementation evaluations. The exclusion criteria encompassed protocols, opinion papers, conference proceedings or similar, abstract-only and reviews and palliative rehabilitation articles. At least two reviewers independently (HB, AL, IU, TGH, SGK, SFH) screened titles and abstracts for eligible studies. Any disagreements were resolved by consensus between the two reviewers. If consensus could not be achieved, a third reviewer was consulted. Studies that passed the initial screening were then reviewed in full text by two reviewers (HB, SGK), who independently decided whether the study should be included. Disagreements were resolved by consensus between the two reviewers. Decisions were discussed and resolved with SGR if consensus could not be achieved.

The authors followed Levac et al.'s (26) recommendations for advancing scoping methodology to chart the data. A data charting form was developed to extract data from the studies included (Supplementary File S2). The items selected for extraction were chosen to give a thorough insight into each study on design, characteristics, and results and enable a basis for comparison between studies. Items for extraction are country of origin, year of publication, aim, study participants, sectors included, setting, methods, main conclusion, framework, top-down or bottom-up approach, intervention description and the implementers. Two reviewers (HB, SGK) independently charted the data, and then each reviewer verified the other's charting. Disagreements were discussed, and if consensus could not be reached, a third reviewer (SGR) was consulted. The charting process was iterative, and when it became evident that the organizational approach was an element of importance, an additional column with “Top-down, Bottom-up” was added. “Top-down” or “Bottom-up” referred to the implementation process' origin and lead, as to who decided and initiated the implementation. A top-down implementation is decided upon and administered from the top of an organization, for example a national cancer reform (27). A bottom-up implementation is driven and co-designed by the front line healthcare e.g., a patient trajectory where the healthcare professionals define challenges and propose solutions (28).

The charted characteristics are synthesized and presented narratively and in tables. All decisions were discussed between HB and SGK, and if there were any disagreements, SGR was consulted.

The content analysis with a deductive-inductive approach inspired by Elo & Kyngäs (29) was applied. The analysis was divided into three phases: preparation, organizing, and reporting. See Figure 1. The content analysis aims to attain a condensed and broad description of factors influencing the implementation, with categories describing this (29). During the preparation phase, reading the papers, it became evident that factors influencing the implementations process were distributed across various sections in each paper. These factors were identified as either facilitators, barriers or factors with an unspecified impact. All relevant factors were extracted grouped into subcategories. These subcategories were then refined into main categories. To ensure validity, the categorization process involved iterative comparison between individual factors and the emerging categories. Factors that were not clearly labeled in terms of their influence on the implementation were excluded from the results, as their impact remains unclear. The main categories for the factors acting as facilitators or barriers are depicted in the “results” section.

The search took place on November 1, 2023. Initially, 4,630 citations were screened, with 4,425 deemed ineligible. We conducted full-text screening on 205 papers, of which 169 were excluded. Ultimately, 36 papers were included in the final stage (27, 28, 30–63). The primary reason for exclusion at this stage was that the studies were either only abstracts, only focused on one sector, or the studies lacked a focus on implementation. See Figure 2 The PRISMA flow chart.

The 36 papers included were published from 2000 to 2023, with 28 published from 2014 and onwards. Interventions were allocated across four of the five Nordic countries. Denmark (n = 15) (32, 33, 36, 39, 41, 45, 48–50, 56–58, 61–63), Sweden (n = 13) (27, 31, 34, 35, 37, 38, 40, 42, 44, 46, 47, 51, 54), Norway (n = 7) (28, 43, 52, 53, 55, 59, 60), and Finland (n = 1) (30). No studies were identified from Iceland. In terms of study design, six were quantitative, twenty qualitative, and ten employed mixed methods. Some papers reported results from the same original intervention but with different aims (37, 38, 46, 47, 59, 60).

All interventions ultimately aimed to improve patient outcomes, but the interventions targeted various professions, which then influenced patient rehabilitation. In 28 studies, healthcare workers performed the implementation. The healthcare workers included were physiotherapists, occupational therapists, dieticians, psychologists, nurses, and nurses' assistants. A broad variation of other professions involved in implementation were physicians, general practitioners, pharmacists, managers/leaders, employment specialists, social insurance officers, discharge planning coordinators and researchers. Few studies included the patient as a part of the implementation process. Six studies included patient interviews as part of the evaluation (44, 48, 50, 59, 62), and two studies used questionnaires for patient evaluation (39, 41). A total of 29 studies did not ask for patients' opinions on how they were affected. See supplementary File S2.

Twenty-two papers (28, 32, 35, 39, 41, 43–51, 55, 56, 58–63) described interventions moving from the secondary sector to the primary sector, while only one paper described an intervention moving from the primary sector to the secondary sector (27). The secondary sector refers to hospitals of varying sizes, often involving one or more wards. The primary sector encompasses a wide range of settings, including in- and outpatient rehabilitation units, home care, and specialized units.

Thirteen papers described an intervention with more parallel characteristics, implemented in either two sectors (30, 33, 37, 38, 40, 42, 52, 54, 57) or within two entities of a single sector (31, 34, 36, 53). The implementation was presented as more parallel than integrated across sectors, without any obvious element bridging them. Eight of the 13 papers described parallel interventions in the primary and secondary sectors (30, 33, 37, 38, 40, 52, 54, 57), e.g., a screening tool for anxiety and depression being implemented simultaneously in both a hospital and a municipality setting (33). Five of the 13 papers described interventions implemented only within the primary sector (31, 34, 36, 42, 53), but from one entity to another within the sector, for example in a municipality setting between a mental health care unit and a vocational service unit (42).

In 12 papers, the organizational approach to the implementation process was described as top-down (27, 30–32, 37, 39–42, 46, 52, 62). Three papers described a bottom-up approach (28, 44, 51), while two papers combined the two approaches (38, 54). Nineteen papers did not specify the approach (33–36, 43, 45, 47–50, 53, 55–61, 63). With regard to which organizational approach facilitates more successful implementation, all three papers employing a bottom-up approach report only limited success (28, 44, 51). As one paper highlights, “bottom-up process and enthusiasm are not enough” (p.8, Røsstad, 2015) (28). Conversely, top-down approaches appear to support the achievement of greater implementation success (27, 31, 32, 39, 41).

To examine whether utilizing a framework facilitates implementation, we searched the papers for the use of frameworks. Seventeen papers utilize implementation frameworks considering both context and individual perspectives. See supplementary File S2—data extraction chart. The frameworks are RE-AIM (30, 56), PARiHS (35, 49), Normalization Process Theory (28, 33), ERIC (38), CFIR (37), Model of strategic change management developed by Pettigrew and Whipp (31), Realistic Evaluation (63), Framework by Nielsen et al. (47), implementation outcomes described by Proctor et al. (40, 52), The Theoretical Domains Framework (61), Process evaluation by Saunders et al. (34, 62) and Process evaluation by Thorne (36).

Only one of the papers described using an implementation framework for both planning and evaluating the intervention (56). The majority of the papers employ a framework for evaluation only (30, 33, 35, 36, 40, 63) or for data analysis (28, 37, 38, 47, 61, 62). The remaining papers utilize the implementation framework for a theoretical framework (31, 49) or to structure or inspire the interview guide (34, 52). A search through the 36 papers' results and conclusions sections did not convey more success using a framework.

Factors reported as facilitators, barriers, or both were grouped into subcategories during analysis and later organized into two main categories: Organization & Resources and Collaboration & Communication (see Figure 3). Only the main categories are described narratively, while subcategories are shown in Figure 3 to illustrate the analytical process.

The category of Organization & Resources encompasses factors that affect the implementation within the organizations, such as organizational structure, the teams and leadership roles, the operational procedures, and human resources, such as staffing and management. Financial resources, IT systems, and resource allocation are also in this category. As are the more evaluating factors.

Collaboration & Communication encompasses how teams, departments, and sectors work together, internal and external communication, and interaction with other stakeholders in the implementation. It also includes structures for discussions and decision-making, as well as other tools and strategies for a collaborative work environment.

The alignment or especially misalignment of the intervention within the context is a frequently mentioned hindering factor. It is attributed to, e.g., inflexible policies, the structure of the intervention, the size of the workplaces, a patient population that differs from what the intervention requires (27, 30, 31, 33–36, 39, 42, 44, 49, 52, 53, 63) or the organizations working together differing in aims and perspectives (43, 45, 50, 62). A facilitator that can affect this factor is pilot testing or trialability, mentioned by several papers as part of the pre-implementation phases (30, 37). Additionally, involvement from the frontline personnel at all stages is a crucial facilitating factor for the implementation and for the intervention to be integrated and contextually appropriate (32, 33, 38, 46, 49, 50, 53, 56, 57, 61).

Time & Resources are factors of great influence, both acting as barriers and facilitators to the implementation process. Commonly cited barriers arise in the form of a lack of resources, high staff turnovers, and time-consuming tasks like forming new teams or recruiting new staff (28, 31, 37, 40, 44, 51, 54, 56, 57, 62). Facilitators are commonly seen in the form of management engagement, where management is often responsible for providing additional resources, such as adjusting task lists, supporting staff in dedicating the required time, and allocation and helping to prioritize time (28, 37, 56, 61, 62).

The time factor can act as a facilitator when there is sufficient time for planning and conducting the implementation (48, 61, 63) and as a barrier when time is perceived as insufficient (27, 28, 30, 31, 35, 36, 39, 40, 46–49, 52). Insufficient time can lead to a feeling of inadequate preparation among implementers and uncertainty about what to expect (28, 34, 36, 46, 53, 55) and can impact the ability to reflect on challenges, causing feelings of professionalism to be compromised (35, 36).

The workplace itself is a factor. A barrier is a workplace culture resistant to change (34, 36, 39, 42, 49). Support or encouragement from the workplace, including project leads, leaders, and colleagues, can act as a barrier when absent (35, 36, 60) and as a facilitator when present (47, 63). Additional facilitating support encompasses technical, organizational, and emotional assistance (47, 48). Having a person designated as the facilitator on-site (49) can aid in creating the sense of being part of a larger common agenda (49). Researchers can facilitate implementation by providing the right research evidence that aligns with the clinical practice needs (35). Project leads or trainers contribute to a barrier through insufficient informal workplace training, insufficient continuous support, and a lack of follow-up (33, 63).

Managers play a key role in the implementation process. They can be a barrier in the form of lack of support, due to absence from daily practice, or failing in providing leadership (33, 39, 46, 53, 62, 63). On the other hand, managers are frequently mentioned as facilitators (28, 33–35, 49, 53, 54, 56, 60–63). Facilitating management involves showing commitment and drive, being encouraging and responsive, as well as being realistic and concise. Formal requirements and mandatory participation facilitate implementation (33, 62). The factors and the number of papers that include each factor are depicted in Figure 4.

Knowledge and competence among staff regarding the intervention are identified as a factor that is a barrier to implementation when there is a lack of knowledge and a feeling of lack of competence towards the intervention (31–33, 37, 44, 49, 55) or lack of knowledge of each other's work (42, 56, 62). However, it can be a facilitator through educational activities (33–35, 38, 52, 54, 62, 63), including workshops, meetings, and education on relevant evidence. Knowledge sharing between staff at meetings or audits was a valuable resource; discussions and communication made the interventions meaningful for staff (31, 34, 35, 39, 49, 53, 56, 57, 59, 61), as well as training in a real environment (35, 47, 52, 63), and learning through their own experience (33, 35, 55, 63). These activities helped staff feel prepared to implement the intervention.

Attitudes among staff are a factor that holds both barriers and facilitators. Attitudes can be a barrier when there is a lack of motivation, a perception of no need for change or resistance towards change, and skepticism towards other professions and their capabilities that complicate collaboration (31, 34, 39, 42, 44, 45, 50, 54). Attitude is a facilitating factor when it includes mutual interest, active engagement from all parties, increasing awareness, positive expectations, respect towards the practitioners’ roles from managers and researchers, and a sense of responsibility towards practicing improvement (31, 33–36, 41, 46, 54, 56, 57, 61, 62).

Communication is a factor that can facilitate collaboration (36, 39, 45–47, 54, 60, 62). This includes more structured communication, guidelines, and communication at all levels within and across organizations. The setting for communication must be developed in the specific context of cross-sectoral work. This is also the case for documentation flow and systematic processes, such as referrals that were a facilitator when functioning (34, 35, 39, 41, 43, 44, 52, 53, 62) and an important factor for cross-sectoral communication. Working towards a shared goal can be both facilitating when it occurs (34, 39) and a barrier when not (44, 45, 50, 54). Communication can be a barrier when respect is lacking, e.g., researchers have to respect stakeholders' knowledge and expertise in their respective fields (35, 59).

Patients and their families can be a factor that serves as a barrier when they are dissatisfied with the information provided, as it was either repetitive or contradictory to previous information received (33, 63). The factors and the number of papers that include each factor are depicted in Figure 5.

We consider the broad scope of rehabilitation focused on care, physical, mental, and/or social outcomes a strength in our effort to identify factors affecting cross-sectoral implementation. The papers in this scoping review, although covering various rehabilitative areas, all contribute to valuable knowledge for future implementations efforts. A strength of this study is that we not only searched in the larger databases but also in databases targeted at rehabilitation, such as PEDro, OTseeker, and journal of Klinisk Sygepleje (Clinical Nursing), and in the search for relevant papers, included a citation search in Web of Science. Also, we adhere to the PRISMA-ScR recommendations and reporting standards, the consultation with information specialists in deciding the search strategy, and the use of multiple reviewers in all phases.

This review has limitations. Firstly, despite using an extensive search strategy, we may have missed some publications as the research papers do not always clearly describe the setting or focus on implementation and a more precise and established terminology could have increased the number of relevant studies found. However, the potential loss of relevant studies was minimized by hand-screening for additional records as well as by using Web Of Science. Secondly, only published studies were included, and grey literature and unpublished studies were excluded. While this decision supported feasibility and transparency of the review process, it may have biased the findings toward academic and clinical perspectives, underrepresenting practice-based or policy-level insights. Thirdly, sector-specific studies were excluded, which means that contextual factors unique to single-sector settings were not captured. Whether these differ substantially from those influencing cross-sectoral implementation is unknown and warrants further research. Fourthly, a third of the included studies, despite having a cross-sectoral design, described a more parallel implementation without a clear element bridging sectors. These studies are presented and included. Thus, the facilitators and barriers for cross-sectorial implementation are still experienced in a multiple-sector setting. Several factors that seem to affect the implementation were not labelled or distinguished by the authors as to what effect they had on the implementation and could possibly hold valuable knowledge. Finally, consistent with scoping review methodology, we did not appraise the quality of the studies included. This lack of critical appraisal means the strength of evidence underlying individual findings cannot be assessed within this review. Readers interested in evaluating the robustness of particular studies are therefore encouraged to explore the articles included independently.