The Behaviour Change Wheel: A New Method for Characterising and Designing Behaviour Change Interventions, by S Michie, MM van Stralen, and R West. Implementation Science. 2011;6:42.
Background and Purpose: The purpose of this research was to identify and evaluate the quality of existing frameworks of behavior change interventions, develop a new, comprehensive framework of behavior change interventions, and evaluate the reliability of using the new framework to characterize the components of interventions targeting tobacco use and obesity.
Method: A systematic review was conducted to identify and evaluate the quality of existing frameworks of behavior change interventions. Web of Science, Pubmed, and PsychInfo were searched using keywords such as "framework," "behaviour change," "intervention," and similar terms, and 8 international experts in behavior change were consulted. After screening titles and abstracts, full-text articles were reviewed. An article was included if it described a framework of behavior change interventions, the description provided enough detail to identify key characteristics, and the article was written in English. Two authors judged whether each framework met 3 usefulness criteria: comprehensiveness (applies to any intervention type), coherence (intervention categories capture the same type of entity), and linkage to a central model of behavior (a hypothetical description of behavior mechanisms). To develop the new framework, 2 reviewers independently abstracted and refined the categories of behavior change interventions and category definitions from each framework. The influence of each category on components of a behavior change model, developed by the authors, was determined. Two authors independently used the framework to classify components of a tobacco control strategy and an obesity guidance and reached consensus on a final "gold standard" classification. Two policy experts independently performed the same classification and their results were compared with the gold standard.
Results: Of the 1267 titles and abstracts screened, 37 articles/reports were evaluated, and 19 articles describing 19 frameworks were included in the review. Zero, 3, and 7 frameworks met the criteria for comprehensiveness, coherence, and linkage to a behavior model, respectively. Categories of behavior change were identified as either interventions ("activities aimed at changing behavior") or policies ("actions on the part of responsible authorities that enable or support interventions"). Policy categories were considered mutually exclusive and included communication/marketing, guidelines, fiscal, regulation, legislation, environmental/social planning, and service provision. Intervention categories were conceptualized as overlapping functions because a single behavior change intervention could serve multiple behavior change functions. Functions included education, persuasion, incentivization, coercion, training, restriction, environmental restructuring, modeling, and enablement. Authors also developed a model of behavior indicating capability, opportunity, and motivation as primary influences on behavior (ie, COM-B). The final framework was graphically represented by a 3-tiered behavior change wheel. The behavior model (COM-B) is at the center circle, with interventions functions and policies on the middle and outer rings, respectively. Percentage agreement between authors in classifying tobacco use and obesity strategies was 88% and 79%, respectively. Policy experts achieved 75% and 85% agreement with the gold standard classification.
Conclusions: The new framework is comprehensive, cohesive, and linked to a model of behavior. Preliminary findings suggest the framework can be reliably used to characterize behavior change interventions. Future research is needed to demonstrate whether use of the framework facilitates the efficient development of effective behavior change interventions.
Abstracted by:
Nancy M. Salbach, PT, PhD
Associate Professor
University of Toronto
Toronto, Ontario, Canada
Development of a Behaviour Change Intervention to Increase Upper Limb Exercise in Stroke Rehabilitation, by LA Connell, NE McMahon, J Redfern, CL Watkins, and JJ Eng. Implementation Science. 2015;10:34
Background and Purpose: Behavior change of stroke survivors has been a recent focus in the stroke rehabilitation literature. What if therapists' behavior also needs to change? A gap between the evidence for and implementation of complex interventions in clinical practice in stroke rehabilitation is becoming apparent because of lack of fidelity in clinical carryover of research findings. The purpose of this article is to describe the process undertaken to develop an intervention for upper limb repetitive task-oriented training for stroke survivors that would "maximise the 'potential fit' of the developed intervention, and intervention materials, with the context in which the intervention would be implemented."
Method: The authors used the Behavior Change Wheel (BCW; see preceding Abstract) model to loosely guide a process that emerged as they proceeded. This process first entailed the establishment of a collaborative development group to include the research team, key stakeholders, and end users. This group studied the problem and identified and analyzed target behaviors using the COM-B model that considers the caregiver's physical and psychological capability, physical and social opportunity, and reflective and automatic motivation in relation to the behaviors. Based on this they designed the intervention based on Behavior Change Technique taxonomy (BCT). Last they refined the intervention.
Results: The development group consisted of the researchers and physical and occupational therapists and rehabilitation assistants from an inpatient stroke rehabilitation unit. This group met approximately once per month for 7 months. They chose to prioritize therapist behaviors because they believed that this group was most amenable to change, and that their behavior change would have the most positive impact. The target behaviors were identifying suitable patients for exercises, provision of exercises and equipment, communicating exercises to family/visitors, and monitoring and reviewing exercises. The behavioral analysis of the target behaviors concluded that the areas most in need to change were the therapists' physical and social opportunity related to the key behaviors. They used the BCT taxonomy to develop interventions to address these areas. This process resulted in the creation of the Promoting Recovery of the Arm: Clinical Tools for Intensive Stroke Exercise (PRACTISE). It consists of team meetings and a toolkit that includes a screening tool to identify appropriate patients, exercise plan, exercise pack to facilitate communication about the rationale, the exercises, a way to monitor exercise, and an audit tool to help therapists monitor their behavior in implementing the intervention. The intervention was pilot tested by the development group clinicians to determine where refinements were needed. The intervention is now being pilot tested in 2 other clinical sites.
Conclusions: The authors concluded that it was difficult to describe their complex and iterative process. They believe that their process presents one, but not the only, way to develop a complex clinical intervention that can be implemented with fidelity. Because no systematic frameworks have been published to date, they suggest that their experience can serve as an example for others who are interested in changing health care provider behavior to bridge the gap between research findings and clinical care.
Abstracted by:
Lois Hedman, PT, MS, DScPT
Associate Professor
Northwestern University Medical School
Chicago, Illinois
Evaluation of a Large-Scale Weight Management Program Using the Consolidated Framework for Implementation Research (CFIR), by LJ Damschroder and JC Lowery. Implementation Science. 2013;8:51.
Background and Purpose: Overweight and obesity is negatively associated with health outcomes such as morbidity, mortality, and health care costs. A large proportion of adults in the United States are affected, and prevalence is particularly high in the Veteran (VA) population. There is an important need for effective preventive services. MOVE! is an evidence-based weight management program delivered across a US network of 155 medical centers and 872 community-based outpatient clinics. Implementation success varied widely. The purpose of this study was to identify factors influencing implementation, using the consolidated framework for implementation research (CFIR) as a conceptual framework. The CFIR offers a taxonomy of constructs that influence implementation and are organized into 5 domains: program characteristics, outer setting, inner setting, implementation process, and characteristics of the individuals.
Method: Semistructured interviews with 24 MOVE! staff (facility coordinators and staff involved in MOVE! implementation) were conducted 18 to 22 months after initial dissemination of the program. MOVE! staff recruited were from 5 VA facilities selected to maximize variation in implementation effectiveness, assessed using VA candidate participation rates and information on program components implemented. The interview guide was designed to explore the extent to which 4 of the 5 CFIR domains influenced implementation (the individual characteristics domain was excluded as individual-level behavior change was not the focus). Individual transcripts were coded using the CFIR framework. They were used to write a case memo for each VA facility outlining summary statements and quotes for each CFIR construct assessed. Using these 5 case memos, a rating was assigned to each CFIR construct according to the valence (positive or negative) and strength of its influence. In a final analysis phase, CFIR constructs qualitatively correlated with implementation effectiveness were identified by comparing facilities with high (n = 2) versus low (n = 2) implementation effectiveness.
Results: Of the 31 CFIR constructs assessed, 10 strongly distinguished between facilities with high versus low implementation effectiveness. Of these, 6 were related to the inner setting: networks and communications (quality of working relationships, team formation, communications about MOVE!), tension for change (ie, felt need for change), relative priority (compared with other initiatives or programs), goals and feedback (tracking and reporting of program data), learning climate, and leadership engagement (ie, program support); 2 were related to the implementation process: planning, and reflection/evaluation (on/toward potential improvements/changes); 1 was related to intervention characteristics: relative advantage; and 1 was related to the outer setting: patient needs and resources. The remaining 21 CFIR constructs distinguished weakly between facilities with high versus low implementation effectiveness (2 constructs), were mixed (16 constructs), had insufficient data to conclude (3 constructs), or were not applicable (1 construct).
Conclusions: This study identified factors explaining variation in the success of the implementation of the MOVE! program. As underlined by the authors, this article can be used as a methodological illustration of how to use the CFIR to identify determinants of implementation success and recommended actions for improvements. Authors refer to a website with additional information on the CFIR to promote its use (since the time of its publication that material has been moved to http://cfirguide.org).
Abstracted by:
Justine S. Baron, MSc, PhD
Clinical Epidemiology Program
Ottawa Hospital Research Institute
Ottawa, Ontario, Canada