Bridging the research to practice gap: a systematic scoping review of implementation of interventions for cancer-related fatigue management

The population for this review were cancer survivors (regardless of age, gender, tumour and treatment type) at any stage of their cancer trajectory that have experienced fatigue as a result of their cancer or cancer treatment. The taxonomy of implementation strategies developed by the Cochrane Effective Practice and Organisation of Care (EPOC) Group [25] were used to determine the definition and inclusion of implementation studies in the review (see Additional File 2). These EPOC implementation strategies were developed for interventions that targeted and produced changes at the healthcare organisation level, healthcare professional level and the health service level and thus were considered relevant for this review.

For inclusion, studies were required to meet the following criteria: 1) have the implementation of an intervention/program/guideline as a primary goal; 2) have cancer-related fatigue as a primary symptom of interest; 3) incorporate at least one of the EPOC implementation strategies; 4) be published in English; and 5) have full-text available.

No restrictions were placed on types of study designs eligible for inclusion. As the key interest of our systematic scoping review is to describe implementation outcomes of CRF management interventions, we included original research articles (i.e., randomised controlled trials, observational studies, qualitative studies, mixed methods, abstracts, program evaluations) and other grey literature (e.g., evaluations of modules, online programs and institutional/government interventions). Descriptive articles (i.e., commentaries, editorials, recommendation reports/articles) were excluded.

Four databases (PubMed, Cochrane CENTRAL, EMBASE and CINAHL) were searched (up to December 2020) as outlined in Additional File 3. Free text terms and relevant subject headings (i.e., MeSH, EMTREE) for “cancer-related fatigue” (cancer fatigue, fatigue) and “interventions” were used. These terms were also combined with implementation study terms (e.g., “implementation”, “translation”, “program development”, etc.) using the Boolean logic operators (or, and). Reference and citation lists of relevant articles were also hand searched for eligible studies that met the inclusion criteria. Titles and abstracts of articles retrieved from the search strategy were independently screened by two authors (CS, OAA). The same two authors then assessed the eligibility of relevant full-text articles for inclusion in the review. Disagreements were resolved through consensus among the two authors, with a third author (RC) as arbiter where required.

Data extraction was conducted by one author (OAA) and checked for accuracy by a second author (RC). Key information extracted included: study characteristics (i.e., author, publication year, study design, purpose, participants, sample size); intervention characteristics (i.e., setting, context, model of care, resources used, intervention description); implementation framework or theory used; implementation strategies used; and implementation outcomes. A model or framework was considered specific to implementation if it described the process of translating research into practice, explained the influences of implementation outcomes, or evaluated implementation processes [26]. Implementation strategies were categorised using the components from the EPOC taxonomy (see Additional File 2).

The RE-AIM framework [27] was used to catalogue the outcomes of strategies, methods or techniques designed to change clinician or patient behaviours related to CRF. RE-AIM was initially developed to balance emphasis on internal and external validity and to expand on assessments of interventions beyond efficacy [28‐30]. The RE-AIM dimensions include reach (R), effectiveness/efficacy (E), and maintenance (M)–which operate at the individual-level (i.e., rate or participation, intervention success rate, and endurance of individual behaviour respectively); and adoption (A), implementation (I), and maintenance (M), which focus on the organisation level [31]. In our review, data were extracted using a widely used [28, 30] RE-AIM coding sheet for systematic reviews published on the RE-AIM website (http://​www.​re-aim.​org/​resources-and-tools/​measures-and-checklists/​ - Additional File 4). Due to the heterogeneity of included studies, a narrative synthesis was conducted. As the purpose of this review was to provide an overview of existing evidence regardless of methodological quality or risk of bias, no quality assessment was conducted, consistent with the PRISMA-ScR guidelines.

Included studies are described in Table 1. Of the six studies included, three were conducted in the USA [34, 36, 37], two in China [32, 35], and one in Canada [33]. Study design varied and consisted of one clinical audit implementation study [32], one qualitative program evaluation [37], three pre-test, post-test single group observation studies [33, 35, 36], and one longitudinal 3-group quasi-experimental comparative study [34]. Three studies [32, 33, 35] examined the impact of implementation efforts on the adoption of CRF evidenced-based guidelines, of which two [32, 33] reported the impact of implementation strategies at the organisational and health professional level, and one [35] described the impact of implementation efforts on oncology nurses and patients. The remaining three studies [34, 36, 37] investigated the efficacy of CRF interventions implemented at the patient level. Participants in each of the three efficacy studies [34, 36, 37] had mixed tumour types (i.e., breast, colon, lung, gastrointestinal, prostate, ovarian, uterine, myeloma, non-Hodgkin and Hodgkin lymphoma cancers). CRF interventions described in these three studies were physical activity and exercise [36], clinician and patient education [34], or combined exercise and education [37]. Of the two studies with education components [34, 37], only ‘information giving’ education strategies were described (e.g., information sessions, printed material). Types of physical activity described were aerobic exercise (e.g., walking, treadmill, Nu-Step), resistance training (e.g., resistance bands), balance and stretching, yoga and aquatic exercises [36, 37]. Of the three studies that implemented CRF interventions [34, 36, 37], two [34, 36] reported the impact of CRF interventions on patient fatigue outcomes (e.g., reduced fatigue).

Only three of six studies were informed by an implementation framework or model. Huether and colleagues [36] utilised Cullens and Adams’ Implementation Strategies for Best Practice Guide. Jones and colleagues [33] used the Knowledge to Action (KTA) implementation framework; and Tian and colleagues [35] used the Promoting Action on Research Implementation in Health Services (PARIHS) framework to guide dissemination and implementation of CRF guidelines into clinical practice.

Distinct EPOC implementation strategies used by all included studies are recorded in Table 1.

Implementation outcomes of the included studies are outlined in Table and Additional File 4. Overall, Effectiveness and Implementation were the most highly reported dimensions followed by Reach. Adoption and Maintenance were the least reported dimensions.

Reach is defined as the number, proportion and representativeness of individuals who are willing to participate in a given initiative or intervention [31]. Descriptions of target population (including demographic information), inclusion criteria and sample size were reported in five [33‐37] of six studies. Only one study [35] reported the representativeness or characteristics of participants and non-participants by comparing the sample with broader populations. Program participation rate was reported by two studies [33, 34].

Efficacy describes the impact of CRF interventions on identified outcomes (e.g., fatigue). Fatigue and behavioural outcome measures were reported in four of six studies [34‐37]. Of these studies, all reported reduced CRF severity as a result of the intervention. The ‘Energy Through Motion’ CRF intervention [36] resulted in decreased fatigue severity by an average of two points compared to an increase of 0.69 points in the usual care arm (p = 0.0006). Pre- and post-program scores from Van Gerpen and Becker’s ‘LifeSpring’ CRF intervention [37] demonstrated statistically significant improvements in fatigue (5.58 (SD 2.11) vs. 3.55 (SD 1.86); p < 0.0001). Tian and colleagues’ [35] also produced lower patient CRF scores after their CRF intervention (p = 0.04). Lastly, the ‘Passport to Comfort’ CRF intervention [34] produced significant and beneficial effects on fatigue barriers (p = 0.001) and patient fatigue management knowledge (p = 0.002). No studies reported on cost-effectiveness.

Adoption is defined as the number, proportion, and representativeness of settings and intervention agents who are willing to initiate a program [31]. Indicators for adoption were the least reported outcomes in the included studies. Further, indicators such as the description of targeted locations, inclusion/exclusion criteria of settings and staff, method to identify settings and staff, setting and staff participation rate, representativeness of staff and settings, number of staff participating in intervention delivery, and measures of intervention cost were not reported by any study.

According to Glasgow and colleagues [31], implementation at the setting level refers to the cost of implementation, and whether the intervention was delivered as intended. At the individual level, implementation refers to clients’ use of the intervention and implementation strategies. Intervention completion rates were reported by four studies [32, 33, 36, 37] and ranged from 80 to 90%. No study described methods to ensure fidelity of the intervention. Additionally, only the ‘Energy Through Motion’ CRF program [36] detailed the ongoing implementation cost of the intervention (intervention kits valued at $21.75 USD per patient).

Results of implementation efforts varied across all studies. Implementation strategies utilised in Wang and colleagues’ study [32] resulted in increases in nurse CRF education, nurse assessment of patient CRF upon admission and at regular intervals throughout treatment, and nurse provision of patient education on exercise and other management strategies for CRF. Jones and colleagues [33] reported that their two-hour health and community professional training session resulted in large to very large increases in clinician CRF knowledge (d = 0.98), self-efficacy in CRF assessment (d = 0.88), self-efficacy to intervene for CRF (d = 1.13), and intent to apply CRF guidelines (d = 1.35). Tian and colleagues [35] dissemination of CRF guidelines led to increased clinician knowledge, attitude and CRF management behaviours, and the increased adoption of effective CRF management strategies amongst patients. Borneman and colleagues [36] strategies to address professional and system barriers (e.g., formal fatigue presentations to staff, monthly newsletters, ongoing meetings with nurse practitioners) resulted in organisational change (e.g., routine fatigue assessment added to outpatient clinic sheet, increased supportive care referrals). Although Huether and colleagues [36] and Van Gerpen and Becker [37] described the use of implementation strategies in their respective programs, outcomes of their implementation efforts were not reported.

Maintenance is defined as the extent to which individual behaviour is sustained 6 months or more after the intervention; and whether a program or policy is institutionalised as part of routine organisational practice [31]. Maintenance indicators at both the individual and setting level were not fully reported in any study and only partially reported across four of the six studies [34‐37]. Borneman and colleagues [34] reported individual follow-up and attrition (3 months) after program completion; however, did not provide follow-up data at ≥6 months post-intervention Tian and colleagues [35] stated that innovations and strategies developed from their implementation efforts were maintained for 2 months after project completion, but were discontinued due to lack of staff time and funding. Borneman and colleagues [34] noted their dissemination of the intervention was conducted at the study institution and that plans were underway to disseminate the intervention into other community centres.

At the time of writing this review, the ‘LifeSpring’ CRF intervention described in Van Gerpen and Becker’s evaluation [37] is currently maintained and institutionalised at the US Bryan Health Medical Center. Additionally, the ‘Energy Through Motion’ CRF intervention [36] appears to be institutionalised at survivorship clinics offered by the University of Iowa Holden Comprehensive Cancer Centre in the United States; however, little detail of the current program and its integration can be found. None of the studies provided details on costs associated with maintenance, however Van Gerpen and colleagues [37] state that ongoing funding for the program was provided by the medical centre’s foundation, and Huether and colleagues report that after project completion of the ‘Energy Through Motion’ CRF intervention, “continuation of funding for patient supplies was obtained through requests from a regular benefactor of the cancer centre” [36].

Implementation research in CRF management is severely lacking, highlighting the need for focussed research in this area. In Table 2, we provide key findings and recommendations of our systematic scoping review. Although feasibility, acceptability and effectiveness outcomes are widely reported across CRF literature, a greater focus on other pertinent implementation outcomes such as adoption and program maintenance are paramount to translate CRF guidelines and interventions into real-world settings. While we acknowledge that these studies have different aims, and may not comprehensively cover all dimensions outlined in the RE-AIM framework, it is suggested that CRF implementation studies incorporate several stepwise iterative phases to provide opportunities to trial, assess and refine elements; determine resource needs and costs; and gather evidence of implementation impact [49].

Most studies included in this review described CRF implementation at the health professional level, or in acute health care settings. However, the physical, psychological, and psychosocial needs of cancer survivors after active treatment require continuous long-term support which is often only provided by primary and community health care teams [52]. As such, there is a need to extend CRF implementation and translation efforts to community and primary care settings where they will be more accessible to larger populations of cancer survivors in the community.

This review has two main limitations. First, the inconsistencies of what constitutes an implementation study (e.g., implementation processes, terminologies, definitions, intention to treat, inclusion/exclusion criteria) made it difficult to detect a distinct relationship between the use of implementation models and strategies, and implementation outcomes. However, this is not unique to our review with similar inconsistencies frequently reported across the implementation science literature [53‐55]. Second, our review was limited to studies in English, potentially resulting in some level of publication bias limiting the generalisability of results.